APRIL, 1902. MONTHLY WEATHER REVIEW.
II. AVERAGE ANNUAL PRECmITATION IN THE UNITED
STATES FOR TBE PERZOD 1871-1901.
By Pro€. ALFRED J. HENRY.
The chart of average annual precipitation, accompanying
this paper, is based primarily upon the observations of the
United States Weather Bureau and of its imniedi:ite prede-
ceesor, the United States Signal Service. The work of vol-
'untary observers in cooperation with the Stnithsonian Institu-
tion, the Sigual Service, and the Weather Bureau has also
been utilized.
The total number of rainfall stations used in constructing
the chart was 734, classed, according to length of record. as
follows:
Number of stations haviug LL record-
of 30 yeam or more - ____ - ___ __ - __ ____. _____ - _____ _____ - __ - __ - 140
. Of 20 to 30 years ___ -. -. __ __ .- ___ .- - ___. ___ __ __ 1 M
Of 10 to 20 years ______ ___________ _.__________ ____.__________ 3%
Under 10 years 8:!
These stations are not unifoiiiily distributed throughout the
United States, there being more stations in the Nort.heastQrn
States than elsewhere. If the 734 stations were uniformly
distributed, there would he a single sbtioii to each 4,100 square
miles or an average distance apart of about ti5 miles. The
average distribution io the more, thic.kly ,settled pwtm of the
country is about one station to 9,500 square miles; in the
sparsely populated iagiotis, a single station sometimes repre-
sent an wen of as much as lti,OOO square miles, although the
$\wage is much less.
The observations used in tho preparation of the chart cover
the fundamental period 1871 to 1901, for which time con-
tinuous records have heen obtained from about 135 stations.
These stations fulfill most of the reyuirements.of the so-called
fundaniental stations except as regards exposure. In the great
majority of cases the exigencies of the service have necessi-
tated not one hut seveid reinovals from one building to anotber,
so that homogeneity of exposure has been out of the question.
The only check on the integrity of the observations is that
which is afforded by the internal evidence of the records. It
should be remembered that for each week from April to Sep-
tember and for each nionth of the year the weekly and nionthly
rainfalls are charted and studied both as to geogmphic distribu-
tion and as to the reldons which they sustain to the seasonal
average. I t is, therefore, a compai-atively easy matter to
detect a marked change in the amount of rain caught at nny
single station and to refer it back to the cause. Sniall changes,
due to altered exposures, can not, of course, be detected by
weekly and uionthly comparisons. It is prohthle that the
errors introduced by the several reniovals were not of uniform
sign, and that the excess of one period was otlset by the deficit
of another. A few cases have occurred where the new espo-
sure of the gauge gave less than SO per cent of the i-ainfall
proper to the stabion. In all such cases the esposure of the
gauge was changed and an appropriate correction applied t,a
the imperfect record.
The records of the short series stations in some cases have
been extended up to the full period by a process of extrapola-
tion, based upon the assumption that the ratio which subsists
___. - - -. - - _. ._ __ - __ - __ ___ .__ __. ___ - - __ - __
2Q7
between the rainfall at any Angle station and a near-by station,
or group of stations, having the same climatic characteristics,
is piwtically constant. Owiug to the sparseness of the observ-
ing stations it was not always possible to secure as many a~
three fiindamen@l stations for reduction purposes, and in a few
cases record;r of fifteen to twenty years in length were accepted
without correc tion.
The stations used in preparing the. acc?onipanying chart
XXX-41 with their geographical coordinates, length of rec-
ord, and altitude above sea level are given in detail in the fol-
lowing table.
I n drawing isohyetab for a single nionth the tidelity with
which the actual rainfall may he represented on the finished
chart is largely a question of the scale of the map. The origi-
nal nmnuscript maps from which the charh of rainfall, pub-
lished in the. Monthly Weather Review, are reproduced, is
drawn on a scale of 1-10,000,0~~0, or Tia of an inch to a mile.
This scale is not large enough to permit charting all of the
available rainfall dah. Thus, in Massachusetb, with an area
of 8,040 square miles, hut fj of the 39 stations which report
iiionthly can be charted. I n geiieial, not niore than one-
t.liird of the total number of rainfall reports are charted each
month.
Although the last ten years have been fruitful in extending
the network of rainfall stations and in improving the quality
of the observations, the richness of material so noticeable in
preparing the current monthly precipitation chart imniedi-
ately vanishes when we attempt to c.onstruct a chart of aver-
age precipitation for a period of thirty years. The total nnni-
her of stations availalde for New England, the Middle Atlau-
tic States, the Lake Itegion, the Ohio Valley, the middle and
upper Mississippi Valley, and in the lowlands of California is
sufficient; elsewhere, however, the number of stations is not
sufficient.
I n preparing the accompanying chart, isohyetde were drawn
for every 5 inches of rainfall, beginning with the isohyetal
of 10 inches and conc.liiding with the isohyetal of 60 inches.
The interval above the isohyetal of ti0 inches varies from 10
to 20 inches.
From the one hundred md first nieridian westward to t.he
eastern slope of the Carcade Range, in WashingtonandOregon,
and the Sierras in California. alniost. all of the available rec-
ords have been placed on the chart i n mall figures. I n the
niountain regions it will he noticed that the figures in smie
cases are greater t,hnn is indicated by the shading of the
region. I t i Colorado, for example, the main niountnin mass
in the central portion of the State has been shaded t,o corre-
spond to 15 inches of precipitation annually. Theke are six
widely separated points wit,hin the nren of 15 inches that have
over SO inches annual precipitation, viz:
Breckenridge, elevation 9,524 feet, 58 inches; 10 years record.
Clear View, elevation 9,500 feet, 24 inches; 12 yeam record.
Climax, elevat,ion 11,335 feet, 34 inches; 7 yeais record.
Pikes Peak, elevation 14,134 feet, 30 inches; 17 yearn record.
Hanta Clara, elevaDion 8,500 feet, 32 inchw; 7 years record.
Summit, elevation 11,300 feet, 31 inches; 6 yeare recurd.
These stations are not disposed around the main Rocky
Mountain chain so as to point to my simple relation between
208
Stations.
ALABAMA.
Auburn ................... Deratnr .................. Greensboro.. ___ __. . __. ._. Mobile .................... Muntgomerv Valleylied:. ............. .............
ARIZONA.
MONTHLY WEATHER BEVIEW.
Lntitudi
.-
0 1
32 4 34 Yt 32 41
30 41
$2 3..
34 3?
APRIL, 1902.
Reno, Nev ............................................................
Truckee. Cal.. ........................................................ Summit, C.1 .......................................................... Qsro. Cal ............................................................. Emigmnt Gfl Cal.. .................................................. Iowa Hilt. Ca?.. ....................................................... Colfax, Csl ............................................................
BoeS. Cal ................ -i ...........................................
orography and rainfall. Some of the heaviest rainfalls in tht
State occur with surface winds blowing from the nort,heasb 0 1
from the plains against the eastern face of the mountains,
Again, heavy snows occur in winter on the southern slope oj
the Arkansas divide with southeast to south winds, while i t
other portions of the State precipitation occurs with north,
west to north winds, in the rear of an area of lorn pressiire,
With these fack before us it did not aeeni ad\-isahle, in tht
absence of specific data, to attempt to follow, except in a verj
general way, the topographic featnres of t,he State, and sinct
the valleys and parks between the high-level shitions have :
rainfall of less than YO inches, it seenied best to let, the shad
ing show the niininitini rainfall for the region as a whole mi(
to place on the chart t.he values for thc higher level stations
The figures on the chart, stand approximately in the lorn.
tion of the station. The orogritphic features of the surround.
ing country niay be seen by LLI itispoctaion of the accompanging
hypsometric chart (XXX-39). reproduced by pariiiission ani
through t,he courtesy of the Director of the United State:
Geological Survey.
A portion of the State of Chlifornia has lwen left 11nsh:tdei
because of insufiicient data. No data are availnble for thc
crest of the Sierra Nevada in that State, escept along the linr
of the Union Pacific ktilwtty. Following are the observatioi
stations on that line in their order, crossing the range fron
the eastern or Nevada side to the western or California side
Fed. 4.4%
i, $18 r.015 5.939 5:LSll 3 ,m 2$2
6.536
Ststion.
JInricope ............... ..I Nntural Bridge ........... Phoenix .................. Frewott .................. Hail Carlos ................ Signal .................... lexfls Hill ................ Tucson ................... Williamw ................. Yumfl ....................
r ,
ARKANRAP.
Arkaits~s Cibv ............ Caniden ... .:. ............ Dttrdnnelle ............... Fayet.terille .............. Fort Smith ............... Helena ................... Keesees Ferry ............ Little Ruck ............... Mount Ids.. .............. Newport ................... Pine Bluff ................ 3tnttgart ................. Washington ..............
CALIFORNIA.
msheim ................. mtioch .................. Lpt.w ..................... hthlone .................. tubnrn ................... lakernfield ............... lenicifl Barracks.. ....... lishop .................... loca ...................... 3owmsn ..................
Annual
tion.
33 ot 34 3 33 2s :34 B 33 If 34 "?
32 4s
82 14
55 la S2 44
38 33 38 3: $5 13 36 OE
3S "2 34 33 31; 29 34 45
34 34 35 51
34 15 34 32 33 51
33 50
33 00
so 55
3i 15 33.54 35 22
S3 02 3i 20 39 35 39 2i
Iiichm. 5. '
20. I
27. I
46.1
49.1
52. '
5% : 46. I
The distance from Colfax, in the valley, to the suniiiiit ol
the Sierras is but 51 miles. The rainfall is practically thc
same at both stations. About 30 miles north of the line oi
milroad, on the western flank of the Sierras, there are t,wc
rainfall stations. Ednianton and La Porte, both in Plitntai
County. The average for seven gears at La Porte (corrected:
is 77 inches; at Ednianton, 70 inches. The evidence of thea
stations, in connection with that afforded by the line of sta.
tions along the railroad, would seem to indicate that the zone
of maximum precipitation on the Sierras lies, not on the suni.
mit of the range, but between the 3,500 and 5,500 foot levels,
respectively .
The relation between rainfall and topography is perfectly
plain when a niountin mass, as the Sierra Nevada, lies at
right angles to the rain-bearing winds. When, however, the
rain-bearing winds are divided b.v the intrusion of a wedge-
shaped mountain mass and the winds flow along its sides pnr-
allel with its general direction, but not over its crest, the
relation becomes somewhat obscure. This is true in part of
the northern rim of the basin of the Great Valley of Cali-
fornia. North of the mountain mass of which Shash is the
culniinating peak, the rainfall diminishes to less than Si, inches,
while to the southwestward, &s at Delta, in the shadow of the
mountains which rise to an elevation of 7,000 feet, in Trinity
County. ininiediately to the westward, the rainfall rises to 61
inches antinally (1s years' olwervat,ions). If a north and gonth
line he drawn from the forty-first to t.he forty-second parallel
about 20 miles west of Shash it would pass over a region
whose rainfall ranges froni 18 inches at the north end of tbe
line to 56 inches tit the south end. It is manifestly impossible
to portray such sharp ruriation on the acconipanying map
(XXS-41), whose scale is .l.z&VVT.
The purpose of n rihfall chart, as understood by the writer3
is not to furnish accurste detailed data to the engineer, hut
rat.her to serve as a graphic aid in qiiic4dy det,eiiiiining the
general geographic distrihntion of rainfnll.
For t.hc region enst of the Rocky Mountnins t,he present
chart, is pr~l~tibly as acwirste as c m he had for some time to
conie. From the Rocky Mountains to the Pucitic: the chart
presentma the hrond fentoren of distribution only, leaving tmhc
details t,o be worked out, when we din11 have more oh serration^,
both for niountirin and v:tlley st B t' 1011s.
Gtwpwphicrrl cwrtZiitde8, eltwitirin, leriyili qf rtvwrrl, ntttl
fw mij<itrIl tdrcliona 6 1 Ilcz liiiietl AYatea c r i i d Charla for tic@ period 1871-1901.
d T'ERA GE A .NhTUAl L I'RECIPITY TION
Flagstaff .................. 35 1; Fnrt Aptwhe .............. 33 4t
Fort Bowie ............... %2 05
Fort Defiance.. ........... I 55 G Fnrt Grant ................ Fort Hnarhoea.. ......... Fort hfohave Gila Bend ................ 3- :';
Holbrook ................. I I4 N
............. . .>
................... ..................
~~~~~~
3orden 3s 58 pliente 35 17 ..alixtoga ................. 38 38 Lmpo .................... 82 37
0 1
R5 30
Ao 5r
si 3u
E8 0:
(r6 i n 85 YO
111 3; 109 4s 109 23 109 :u IN -53
110 '20 114 36 11? 46 110 1u 112 u:! 111 M 112 uu 112 2s
110 27
113 35 113 40
110 54 112 02 114 3ti
91 a* Y2 4s
93 09
94 15 M 24
90% 92 45 92 06 I 3x
91 09 91 5Y
91 24
93 41
115 W 121 4s 121 54 120 26 121 50
119 00
122 OR
11s 19 120 0.5
1w 34 1'20 04
122 34 116 80
iia 41
Eleva tion.
Fed. 8% Mi
L ?
tx
21: 1,tW
6.90; 5.6M 4. it4
ti. 6lii
4, r8f 6uE i3i
5,Mi 1,lN
4, g?w:
1,lM 5.311; 3.4% 1, GV2
%5
2.404
6. iw 141
4, Fs(
14s 123 330 1.350 418 197 7.w 29s
"33 215 0.3
660
.....
l i 0
46
102 "10
1.w 394 ci.l
4.450 6.5(5
5,m 27.1 1,290
365 163
1 W 18i9 18% 1871 1873 lW5
1W 1872 1.W 1532 1866 lWti 18.W IS90 1W lXi5 1x90 l R i G
1Wi. l.Wl lM9 1Ri9 1.W ISM 1Si6
lW2 lW 1 %W lSi0 1,Si 1W lSRl 1 W 1x72 1xK5 1*i 1W 1%0
l8iS 1Si9 lSs5 1%. 1851 1S.W 1.W llws 1SiO 1871 Ni5 1876 lSi3 1877
Record.
TO-
1M
lsol 1901 1901 1901 1901
1901 1901 1.W l!Ml 1811 1w1 1901 1901 1900 1w1 1901 1901 1901 1w1 1901 19CQ 1901 lS9Y 1901
1Wl 1Wl 1901 1901 1901 1901 1w1 101 1s95 1901 1901 1901 1901
1900 lua, 1900 1S98 1899 1900 1592 1900
1900 P-7 1895 1900 1900 1894
Yearn (inclu rive). - ..
5 26 25 12
L9
I6
24
11 I?. 21 12 18 30
L!
12 li
34 6 35
15 14 1'3 11 35 25
20 22 13 13 10 13 33
!23 22 16 11 29 12 37 17 31 16 3 25 28 10
brertige
preyipi- tut1o11.
RIlllUUl
- .-
I~iclir#. 51.3 49. til 52.48 63.10 51. rr?
-56.93
* 24.3Y 18. i4 14.95 13.12 14. i 6
15.94 5. e2 5. os s. 06 5.3 *21. 12
ti. 92 15.30 11.96
6.98
3.17 11.37 *I!. 31
-. SY
* 49.19
* 32. 13 44.42 43.78 4u. H2 W. i R 47.73
*so. 63 51.68
.w. '3
*.so. 00
,so. 9 32.74
11. I 1'2. 57
* 24. I 11.64 33.58
* 4. 89 16. ?O
4. ?i 3.14 69.80 8. i o 10.6s
55.01 19.80
APRIL. 1902 . MONTHLY WEATHER REVIEW . 209
Geoqmphical coordinates. ukuuatioi.. length of record. and AVER AGE ANNUAL PRECIPITATION for raiqfaU etatiom in the United &&a and Canada
for the period 1871.1901-Continued .
Stations . Latitude
CALlPORNIA4ntinUed .
Camp Wright ............. Cedarville ................ Chemkee ................. Chim ..................... ColftlX .................... Colton .................... Corning .................. Cove!o ................... Cresant City ............. Crokers ................... Cuyauiaca ................ Davisville ................ Delano ................... Delta ..................... Dwauw ................. Dunnigan ................ Duosmuir ................ Eldonulo ................. Elmira ................... Emndido ................ Eureka ................... Fallhmok ................ Pkrminft4m .............. Fernant n ................ Folsom Citv .............. Fort BidwGll ............. Fort CRnJk ............... Fort Onatnil .............. Fort Niller ................ Fort Rom ................. Fort Tejon ............... Fresno .................... Frnto ..................... Galt ...................... Genrgetnwn .............. Gilrny .................... Goahen ................... Hollistcr ................. Hornbrook ............... Humboldt ................ Hydesville ............... Independence ............ Indio ..................... Ione ...................... John ..................... Keeler .................... lieene .................... Kennedy Gold Nine ...... Kingsburg ................ . Kingcity ................. Knights Lnnding ......... liono Tayee .............. Lagrange ................. h p o r t e ................... Lemnore .................. Lewis Creek .............. Livermore ................ Los Angeles .............. Iaa Banosl ................ Mslalioff Mine ............ Mammoth Tunk .......... Marvsville ................ Yendocino ............... l e r r e d ................... Modesto .................. Mojava ................... Monterey ................. Mount Hamllton ......... Napn ..................... Needles ................... Ncwinan ................. . Nevada ................... Newhall .................. Ogilhy .................... Orland ................... ps'aro .................... hi m Springs ............. Paso Robles .............. Pi eon Point ............. d i n t ~reiitw ............. Point Conception ........ Point Re ea Light ........ portersvi3;e .............. Rsvenna ................. Red Fluff ................. Redding .................. Paerninen to .............. Salton ..................... Ran Diego ................ Sail Francisco ............ Snn Luis Obispo .......... Snn Xateo ................ Santa Barbara ............ Santa Monica ............ Scott Valley .............. S i ~l l .................... Spadra ................... Summit .................. Sumner .................. Templeton ............... Traver ...................
0 1
39 45 41 30 39 42 39 43 3 9 8 3402 39 5.3
39 40 41 45 37 48
32.53 3833 35 5 i
41 00 32 Ml
3s51 41 12 3s 41 3s 27
3.4 1s 40 4s :33 3 37 w 34 16 3s 40
41 53 41 10
41 05 37 00 38% 3455
36 $3 39 21 3s 16 37 55 36 .w 36 21 36 51 41 .w 40 4ti
40 32 3660 33 4Y 38 81 36 00 I s 35 12 38 15 36 39 36 12 38 47 39 w2 3; 32
39 4s
36 li 36 12 35 40 31 03
33 0 i 39 OR
39 I8 37 19 3i Qx
s5 03
3ti 3 i s i 20 s5 1s 34 w 35 20 39 10 34 25
32 4.5
39 45 36 .53
33 40
35 38 37 1'2 35 w 34 26
&.! 00 Wi 04
3.4 26
40 10 40 36 35 35 33 29 3'2 43 37 'u3 35 18
37 34 34 '3 3400 11 46
41 19 34 03
39 19 35 44
3530 36 "7
;;
Inehc.8. $?.MI
*16.69
44.9:
Longi- tude .
CALIFORNIA--ContinUed .
Vlwlin ................... Volcano S rin gw .......... ~e a v e r v i ~i ' e ..............
0 1
1% 00
1'Lo M 121 32 121 51
1m 57
117 22 E2 1'2 E3 15 1w 1'2 119 .53
116 35 1'21 43 119 .x 1'" 23 116 40 E1 5s 1'22 16
1'a 51 121 5 i
lli Ox
124 11 lli 09
121 01 11s *%
I'2l 10 1'3 11 El 9 I23 15 119 40 1 3 0.5
118 4.4 119 4Y 1 2 2 i
121 17
E! 51 121 33 119 24
121 25 192 50
1% 10 123 58
114 10 116 1: 1'20 56 El 15 117 -50 118 40 ILW 45
119 33 121 06 1'21 41 1'13. 57 120 3q
El 00
119 51 114 58 121 45 118 a5 1'20 46 ILW .io 115 li
1'21 35
E3 48
Eo 30 1'20 61i 11s 11 121 E 1'11 38 Y22 15 114 35 121 00 El w 11s 33 114 60 122 12 121 44
39 30 l!! 41 122 21 1'23 36 120 24 r2.3 110
119 @2 11s li
122 14 122 27 121 30 115 53
117 10 E 2 26
1% 39 1% 19 119 40
11s B
123 02 1 2 18 117 120 27
119 00 120 41 119 30
.............. ................ ....................
$leva- tion .
29.w
1::;
19.10 .% . :w
iz: $ ~3 .~
*ft: $
&et . 1800
193
2.42
965 si
50 4.4M
4. 800 51 319 1.13s
3. MI0 65 2. LC5
Grx)
6.4 io0
111 1. 066 182 4 . til0
3.- 397 402 100 3 . *E
332 6'24 49 2. i50
193 296 284
2. 151 50
400 4.59s
.! 2s7 $60
3.022 2. iO5 1 . .500
301 332
45 1. 325 L93
5 . ooo 22i
4.w
4%
3.30 121 3. 200 257 tii
lil
90 9. i51 15 4. m m 491 92 2. 59 1. 'JOU w 3%
31
.w 723 1.50 fi 2%
6W 461 2,Q.W WJ 5i;5
il 2re 87 60
100
.w 2. 570
3.555
5. 017 $2" 773 291
4 : 67.5
......
......
1. y 1.1
......
;;
7a5
Brrrkenrk1.ge ............ ~~i e y e n n e \i ' e ~a .......... C:lenr yiew Clininx ................... hlornclo springs .......... Denv6.r .................... Duningo ................. Fort i.11I1ins .............. Fnrt i;arllincl ........... Fort k w i s ..............
...............
!
mm-
.
1861 lSb9 l8il 18il 1850 Mi7 1M 1R81
1859 IS96 I S 7 IS72 18i6 1Rn3 18% 1877
1 $59 1P89 1.w;
l8iIi
1XXi 1856
l8ii I S i S 1872 l<W 1 w 1861 JW1 1875 lS55 lSii lW9 18% 147s IS74 18i5
IS74 1Mi 18.54 1.W lW5 18% 1878 I w1
1W 18ii 1W2 18i9 1xw 187s 18i4
1W 1894 18i9 18i9
1871 18ii 18i3 1$54 1878 1871 1871 1.972 1Sil lSi7 1Si.i
lcrtll I U i i
I892 1 *59
IW I S i i
ISXI
11w3 18i3 lW9 lW7 lSi5
1SiS
lhi6 IS59 IPS9
Mi9 1872 I R L
1WY 1W 1Si1 1849 M 9 18i4
1.W 1Yi9 1959 lSS9 1874 18il Hi4 18% 1K%
2 1
y;: 2 . il
15.1'3.
4::
Recon? .
I'mwhute I'll#l*1a
Pi kt-x Peak .............. l'iichlo .................. Kungelp ................. ~a n t ~~~~a n i .............. Siiiutnit ................. l'llnln ...................
.................. ...............
TO-
.
1883 1901 1% 1899 1900 1900 1900 I494 1901 1901 1901 1900 1900 1900 1cw1 I!wxI 1XW lIwM 19ou 1'901 1901 1'YW I*W
1 'JCM
19Y) is<r3 1NiY 1s92 1?yi4
i s i m 1901 1'901 1XW 1!w 1900 1LHW
1 SYY
l*twI
1W IS% 1AW 1901 It*u l%W lS99 l!W 1 SY9
1900 lS99 1900 1YOO l*W 1s!W l~J00 189Y 1ScM 1'900 1901 1 <100 1499
1900 1900 1 ,w 1W 1W 1W l9lu 1W l!W IW 1900 lW 1XW 1900
I I M 1894 1900 IW lk!D 1h99 1x99 1901 1m 1x97 1901 1900 1CNl 1901 1901 1901 1901
1M 19uo 1 s99 1891 1900 IS94 1900 1W 1899 1894
Yearn
i n c h sive) . _ .
10 I 11 29 31 3 15
1 x 1ti 5
10
L!
25
1s
6 2.4 1'1
11
15 13
15 25
2.k 23 1 ?i 10 25
G 25 1'2 2.4 1'2 28 2%
2 i
1 Y
27
G 32 11 17 23
23 17
16
2u 9
19 14 23
19
19 1u 30
li "3
30 14 29
30 24 '35
P 24
!I
12 30 24 11 18 20
12 14 14 15 14 11 1'2 11 30 3 52 1'2 30 52
29 I 35 E 32 1'2 16 3) 13 E 9
31
I
2
Stations . Latitude
Hartfnrd ................ New Haiven ............. New Lon(1on ............
DISTRICT OF COLDURIA .
Wnshington .............
A rdier .................. Bronkurillr .............. Eustiu ................... Fawt Mmde .............. .ltil.ksonrille ............. Jnpiter .................. lie\- West ................ Lake (*it.y ................ Yerritt Irlnnd ........... Prnmcah ............... Puntarmu ............... 8t . Augustine ............ pi1 la h.msee .............. ranipi ..................
FI.ORIDA.
GEclKGIA .
Anwriciir ................ AtlnnttL ................. Anptiatai ................. Blnkely .................. Yorsyth .................. Point Petcr .............. khulnn .................. Kmue .................... Snvtinnali ............... Thninasrillc ............. Tiwcnfl .................. Wt~lthonrville ...........
IlUHO .
Boisc ......... .......... Fort SIieruian ........... IC cot enal ................ LRlie .................... Lcwiritnn ................ LIbKCow ................. Mlurrnv ..................
Paris .................... Pocatello ................ Soldier .................. Swan Valley .............
O n k k i ..................
ILLINOIS .
Aumrn .................. (.aim .................... Chicago ................. Gaileshnrg ............... iireenville ............... Griggsville .............. Marengo ................. Nnttoon ................. YcLeansboro ............
0 1
36 3 33 1 G 40 47 3354 38 $2 41 45
3i 30 3!J 30 $5 48
3 i 10 BY 55
?* 51 3Y 45 3 i 15 40 35 3i 25 3i 15 39 05 3!J 10 3i 15
3S 04 40 :u :<n :w
40 30 39 $2 P 541
3s 1s
40 w 3 i 25
3 i 2S 40 U!
41 $2 41 18 41 21
3s M
L ! :w 28 33
28 4.5
2 i 4Ii
21; .I.
24 %?
-32 1'2 2s IS
30 d I 4Ii
9 .%4
30 2 i 25 5i
3 >y
32 *A 33 45 33 3 31 '23 3300 S3 5 i
81 3 34 16 32 Ui 30 5.. z4 3 i 01 44
43 :+I
4i $2 4.5 22
44 w 41; 23
4li 40
45 40 $2 15
$2 13 -12 32 43 25 43 10
41 4i 3i (wl 41 32 40 ;M
m50 39 41 $2 15 39 :%I
3807
IAngi- tude .
0 1
11Y 17 115 34 1'" 55
116 39 1'21 59
l?? 32
IV2 30 106 00 1v2 22 104 55
loti ai W 4i 1o;i 00 l U i .w la5 v2 105 23
1Ui S i
1113 z> 103 25
10.4 20 Ill3 12 1iri 3;
1Ui 51;
lui 3U
10s a5
la5 v! 101 36
1w 45
Ill4 45 I(% 35
lW2 411
72 04
7: .st ;
72 v5
ii W
x:! WI s2 !!2
s1 45
S1 .w s1 39 MI O i
Sl 49 S2 40
PI1 41
S i 13
rc? 14 s1 1s H 16 v. ' l i .- -
S4 14 s4 23
s1 24 H4 55 s3 55
si 59
s3 55 ni os Sl IL5
s4 01 R3 21
41 %Y
llli 1x5
111; 3.!
111; :w 111 20 I l i 00 I17 00 111; 00 113 57 111 25 11'2 3 114 5 Y 112 00
h'( 011 s9 10 si :<.!
I@ "2
s9 25
110 40 s5 37
.w a 2 4
.% 35
Eleva- tion .
Fe€ L 34s 020
2. oou
(8
2. 635
3. 400
9.524
4. 359 9 . .m Ll.3'25
I.. lE2
5. 3 1
G . 5:u 5 . MW i . !Si
s . .w I
4. MA
5 . .w 5. 721 3 . SiW
4.3.w h . i!%
7. 0110 5.1ui
14. 1'2-4 4. iS3
S . 5UJ
ll.300 4 . r2s
......
......
Yl;
45
IKI
11"
ii
-0u
Lw 14 2s
10 2lIi
22 56
14 25
LW
30
z3l
1. 050 143 MI 735
1 . oou :xFI IEi si
:%O liIj5
..... _. lllh
2 .]!IS
".l?IS
ti . i U U i.57 2 .SI<!)
2. i.50
4 . 1Y1
ti. 01s 4.4*2
5 . LW
6. 434
6%
:w *23
is(;
ti.w B E 724 46'2
lis5
hm-
mi0
18u9
1511y
1577 1513 1872
1w7 lX!J 1559 1 .s!3 Y 1W Mil 1471 1W
1 572 1555 1KW 7 .% 5
1591 1591 1.w 1859 1rw5 1591 1 x*
158 1m19 15b1 15!h4
1 S i G
1 sw
15111 1w1 1x71
1511
l<W 18C lS*) 18.51 Nil 1 .w 1532 l.W 1113 1519 lsil lSiA 1512
llclll
1516 lKi!)
1x71
1 55y
1514 1 *.!!I
lS!lO
151x 1hb1
1515
1x19
lKiF
1461 18x1 15-
1.!!!1(1
1.W4 1 5y2
1 X!J3
18% 1593 14Y2 14!14 15n
151y
Nil 1510 l.%l 1- 1w2 1W
11y
152
Record .
TO-
1901 1900 1894 I N 1900 1900
1w1 1901 1901 1901
1896
1901 1901 1901 lyol 1 5 3 1591 1901
1tm1 1901 1901 1901
18y-1
1901 1901 1591 1901 1!w1 1901 1w2 1901
1w1 1901 1w1
1901
1901 1901 1Y01 1901 1901 1w1 1cWJ1 1901 1901 1901 lKsY
1901 1w1 1901
1901 1901 1901 1w1 1h0
1'901
lIJ(I1
1901 1901 1901 19u1 1W
1w1 1891 1901 1!W1 lWl 1901 1901 1'901
1y01
1901 1901 1901
1901 1901 1901 1578 1901 1901 1901 1901 1901
Yearn i n c h live) .
15 E 16
ti
"8 am
10 10 Y 12 7 li
30 10 17 22 11 12 10 10 15
12
11 11 '1:! I? 17
4
G
11
$2
M 31
31
15
10 11 15
311 14 M
14 "4 "I
' 1 2 23 15 29
15 32 31 11
22 1'3 11 '20
.%I
20 17 6
1
14 I? 11 1 Y 10 5
.!
5
1;
c
19
:uJ $1
1'2 15 19 50
21 12
.
aerage rnnnal brecipi- tatian .
Tm%?& 10.M *1.70 38.80 5.31 17.74 16.91
16 .%I
*5.11 16.29 23 . 16
* 33 . 15 14.35 14.07 16.w 14.00 1'2 .I 17.20 * x . .i 1
13.29 13.16 11.51 14 . '90
9 ..a 19.51
10. il
1.72 12 . 29
*O t ;
*W.% 17 . 1'2
* :&: q *
41; . !c3
46.44 46.65
44.10
55 . t17 57 . 72 49.44
.w ..50 .53 .32 -59.57
35 . 39
M.'3
51.M 57 . G1
43. .w 49.41
.w .19 53.13
4%31 32.01 -I.!. 06
M.15 53.24
50 . 51
.w . b i
-50.19 .% .17 5%14
*%.a,
50 .?ti
14.16 25.19 '2% 55
17.36 11;.Sr5 3.17
* 40.40 1 . !W 11.49 13.20
* 13.W 14.43
86.48
$2 . '9 %3 .64 33.00 a.93 35.36
34.M 39.72
41.00
910 MONTHLY WEATHER REVIEW. APRIL, 1902
Geographid mrdindta, &h, length of record, and A J'ERAGE ANhTLL4L PRECIPITATION fur rai,!fall stnfivna i i h the 1hitrd Stah and Onncrda
for Ute period 1871-1901-Continued.
Stations.
-
Longi- tude. Latitud'
-
Eleva tion.
-
f i t . 4%
47t 771 6.&
90(
1, &W
W L
37b 1.101 741
w s3c
665 625 %W *W 1.1oc 1,022 5%
540
1,575 1 .m 617 510 3. -w 1,LW 1,901 3lN 615
1.600 1,218 700
.....
1 .2 3
1.500 721 1,012
1.044 591 1. ?65 7oG 780 6w 600 1:%M
m 600 900 Wl .w 750 1, 512 1,'L.W LJOP
3. a 1 2. ??A 1,4w 1 .M 1.132
1.955 2 ,3 3 1,186 594
819 1,125 2, .539 1.216 1,Z'S 884 3 .9 5 3, -253 1,355 1,010
600 773 suo 514
370 .Si1 995 561 930
ILLrNoIs-continucd.
Mount Carmel ............ Peoria .................... Philo. .................... 8 ringlield ............... $innebago. ..............
INDIANA.
Angola. .................. Cot urnbus ................ Evansville ............... Farmland ................ Huntington.. ............ Indianapolis ............. Lawnia .................. hfayette.. ............... Lagnnsport.. ............. New Hlvmony ........... Richmond ............... Rockvllle ................ S Iceland ................ &+a;. ................... Wort ington .............
INDIAN TERRITORY AND
OKLAHOMA.
Arapaho.. ................ Burnett .................. E u f a u 1 a .................. Fort Gibson .............. Fort Reno. ............... Fort Sill ... I.. ............ Fort Supply .............. Fort Towson ............. Fort Wsshlta.. ........... Healdton ................. Man run ................. Okla%oms,. ............... Tulfa.. ...................
IOWA.
Afton.. ................... Algona ................... Amaim.. ................. Charles City .............. Clarinda ................. Clinton.. ................. crew0 .................... Davenport.. .............. Des Moines.. ............. Dubuque ................. Fort Mndiwn ............. Hampton.. ...............
Keokuk .................. Logan .................... Monticello ............... Yuscatine ................ Onknloosa ................ Siblev ....................
&IC City ..................
Independence. ...........
Sioux City ................
KAN8A8.
Coolidge ................. Dodge... ................ Downs.. .................. Ellinwood. ............... Emporia.. ................. Englewood ................ Eureku Ranch Q ........... Horton ................... I
-
Lnngi- tude.
I*rlm. 41.45 24.36 35.31 37-90 33. a 2 9
3R.Io
!:::,: 36.81 36.77 4 .4 0 43. LW
39.71 41. 16 a7.45 39.
. z::
jl: i!
-
$leva tion.
LOVISIANA.
Alexandrltr.. ............. Amite .................... Baton Rouge ............. Grund Cotmu ............
bfonrue ................... New Orleans ............. Paincourtvilie.. _._. . __. __ PortEadn ................. Hhwvrpurt ...............
MAINE.
Bar Harbor.. ............. Cornish ...................
Fairfield .................. Flapvtnff .................. Gardiner ................. Hoiiltm .................. Kinw .................... Mayfield.. ................ Omno ....................
EaStpOrt.. ................
Rewrd.
p98 -27.25
32. €a L!. 40
31.14
3.5.4.q
30. la 32.73 32.36
SL. 18 37. chq
3 1 .9 31. .W 3.5.2s
8. K5 :s. 18 -3s. MJ
1.44 24.97 30.47
po. 00
* 14.19
25.42
33. 33 31.67 21.09
F63
I ,
%5 *2
2.36 245.70
9 :9 0
!
I
!
viroc~lla .................. Wdlnce .................. Wichita .................. Yatw Center.. ............
KENTUCKY.
Bowlin Green ...........
29.70
31.30
47. 62
49.21 46.00 i
atitudl
Record. 4verage annual
L%i!-
-
Team i n c h sive).
-
Fean inch sive)
-
%urn-
-
1% Mi6 18% 18779 1857
1884 lrwl 1K% 1982 llW2 1R71 1866
1nu 1 W la82 1897 1863 la, l e 2
18%
1tW 189.2 lW7 1836 1 W lb70 1873 1836 lM3 lW9 1892
m 1887
1871 lM1 1875 1875 1872 1 W 1871 1871 1x7s 1-1 18w 1877 1.967 1x71 lW 1K5,
1846 18% 1875
1687 1870
1W2
1x7s
1874 1879 1- I& ISM 1872 1861 1958 lW7 lwxcl 1 W
1S71 lS70 1fM 1879
i n n
i n s
lRS0 1 W
1W7 1889 18\7 l8i2 1871 lW9
iaw
-
TO-
-
From. Stratioils. predpi- tation.
O I
38% 401; 39% 39 4I 42 1;
41 3l 39 16 38 02 40 11 40 rvi 39 46
3805 40 2F 400.15
39 51 39.w 39 &!3 46 39 09
38 l a
353c 35 10 3522 3550 3533
w 4u
w33 24 01 34 14
3109 31 49 35 26
3608
41 16 43 15 41 47 43 &5
40 43 41 51 43 32 41 30 41 56 $2 30 4u 37 4 3 0 42 I
40 3i
41 39 $2 13 41 "6 41 18 43 24 4" 3.5 .I!?. 25
5800 37 45 39 30
3?l 21 8 24 37 03 35 56
39 40 37 13
8 .w 39 12 39 48 37 OY 38 .w 39 03 37 00 38% 37 41 37 53
36 -5s
37 00 3650 3824 37 16 37 16 3s 02
3s 15 38 10
0 1
87 49 6 3 6
8808
89 39 89 12
$500 2666 *7 29 85 10
E530 8609 I 07 8654
$622 87 M
&.I53
87 10 85 18 M ,w 87 00
9855 97 10 95 35 $5 20 98 31 M23
9Y 3 95 12 9638 97%
w '3 '97 33 95 66
942.5 8105
91 55 92 43 95 18 90 10 9' 10
w 39 93 3,q
90 40 91 3
91 57 91 21 95 47 91 15 91 05
9? I 95 43 96 27 95.00
I am
102 00 100 01 98 25 9835 96 12 99 .w w33
95 31 95 41 $5 44
96 37 100 32 9 i %
9736 0 39 101 57 101 35 Y7 20 9543
86 25 H 37 $7 57
F2 27 87 28 R533 H38 85 45
s4 55
0 ,
31 1) 30 44 30 2t
30 2i
31 E
32% 29 FA
30 01 29 9
32x
gg
44 28
43 44 44 M 44 3.5 45 0:
44 14
4ti 07 45 32
45 04
.I4 56
43 1
39 17 39 39 39 3 l
38% 39 43
s!! 10 3s2i 39 .a 39 39
1'2 E
*2 11
-I'! %
41 17 41 39 $2 27 41 M $2 $2 42 16
41 %4
$3 24 1.5 UF,
47 24
45 13 a2 20 15 4s
13 05 44 40 43 51 .I!! 20 $2 44
46 Ofi 4ri 92 46 53 $2 .w 43 44 45 .io
46 2% 44 45
47 12 46 48 43 *2 47 21 46 00
44 .58
46 M
45 30 44 1Y 46 10 43 45 44 56
44 30
4s .m
30% 31 51 3236 33 31
O I
9' -3 90 2s 91 11 81 M
93 32 93 07 92 46
9' 02 90w 91 01
8Y 15 9340
6s 12 70 51 6659
69 35 70 20 69 48 67 49 69 40 6940 65 44 70 15 70 37
76 37
I l l A 75 39
I f -a it; '20 76 27 i D 27 76 U4
7! $5
c- *
F32 I1 01 71 .W 70 uci 70 .W 73 15 71 OG
73 13 71 49
x1 11 s4 37
a 30
Lis 1' si LW
so3 47 05
86 1 s 81 48 E 31
Si SS
81 26 S7 13 87 35
85% M .w KL 22
.% 40
;;
91 14 9.2 06 92 37
w w w59 93 15
96 M
95 s 9430 94 24 w35 9305
97 14 92 15
S8 56 90 LZ
9000 a* 3
Fed. 15
w[
41
%
......
1:
..... i
a 2 4 9
M 781 40
9i
1 ,m 7ti 47a 1. loo
1. ooo 13'1 95 6w
13 700
suo 25
720
i a 2i 2,44u
.....
337
7w 14 100
690 4R3
i n
1,
1.240 750
609 1.350 610 .597
6cH
620 1,134 630
'xx)
R U
1, la2 710 Q O
1. Olb ti30 6#2 5%
639
657 ti72
1.338 1. L !
1,L!
srs 935 1,170 S21
1,130 7.53
m Luu *50
24
250 i;
Inches. 56.72 61.68 69.41 '
63.43 45.60 54.02
50.38 *32.91
58.95 56. OLC
46.58
5.5. m
1901 1901 1901 1901 1901
1901 1901 1901 1901 1Wl 1901 1x95 1901 1901 1 W 1w1 1901 1SW 1Wl 1901
1901 1901 1897 1890 1901 1901 1 W 1S54 1 m 1901 1W1 1Wl 1901
1901 1901 1901 1901 19tIl 1901 1 M 1901 1901 1901 1901 1901 1901 1901 1901 1901 1 M 1901 1901 1901 1901
1901 1Wl 1897 1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1900
1901 1901 1900 1Wl 1901 1901 1901 1901 1901
17 46 16 22
2.3
16 17 1s 14 13 30
.25
?i,
24l 3
27 13
2.3 34 19
8 10 Y 27 I i 29 15 14 15 11 10 11 14
;; 25 17 17
n 28 30 -23 47 47 14 32 30 26 45
49 17 16 12 3
9 27 19 12 'n E a2
12 30
3t;
40 14 14 17 23 10 2.2
13 21
14 13 11 14 11 14 17
L ?
11
IW 1W2 1W3 1&%3 1836 1873 1 M l*
1836 1WU 1WI 1871
1 W 1856 1873 18Rli 18% IS37 183G 1895 lkS.5 1hiO 1871 1861
1817 1871 1870 1 M 1867 1W1 lW2 lV93 1865
I&% 1818 l X G l 1847 1813 lW4 1x54 1823 1x41
l 8 i l 1Wi lS72 XW7 1&97 18% 1871 1871 18117 1W7 lk76 1863
18% 1857 lM9 1,W 1877 1 w 1A6 18il
lKV3 1870 1Ks7 lS57 1 6 3 1 W 1Wl 1.%!5
lLuII 1- 1.387 l e 6 1 W lLS7
18Y3 1870 1&2 156.5
1901 1901 18S 1901 llc15 1901 1901 1901 1901 1901 1901 1901
1901 1WI 1901 lWl 1901 1901 lS96 1901 1'901 1901 1901 1901
1901 1901 1901 l?ms 1Wl lW7Y 1Wl 1901 1875
1900 1Wl 1w1 1901 1901 1Wl lVX4 1901 1901
1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1873 1901 1 W 1901 1901 1901
1874 1901 1901 1901 1Wl 1901 19M 1901 1901 18M 1901 1901 1895 1901
1961 1901 1901 1901
15 1:
z l! 1( 1: 11 1;
6(
1( 11
s(
1( 4t 2;
a: t
5(
16 '&
8(
11
1;
I
B 11
3 i 10 9 11
65 &.I 88 3
&q
7 24 80 '44
23 14 3 13 11 51 P
"8 1" 14 2i 38 9 38 12 35 9 12 42 30
15 31 12 13 9 34 21 16 18 30 14 53 1s 13
7 17 17 89
J e s u p .................... Lake Charles ............. Liberty Hill ..............
47. P 46.66 44.16 8.35 Y1.45 43.43 38.39 s39.02 45. I3 44. 81 1'2.79 43.49 27.07 32.12 36.96 3.5.10 27.36 29. :37
E. 46
50. w 3n. 30 d 25. I
30.94 33.48
41. w 33.46 46. TJ 44.74 44.37 46.70 -2.68
54.25 47.79
44. 51 45. '3
%.7b 43. w Y3.59
46. i 9
39.68 46.W
38.90 531.33 33. H6 32.00 31.66 31.95 32.99 34.32
* 33. 20 31.65 35.30 81.66 32.32 1.48 1.64 31.47
* 3.3. 70 30.8% 30. $2 36.30
Amhemt.. ................ Bwtnn .................... Fitchhurg ................ Nantucket.. .............. New Bedford ............. Pftiufleld ................. Taunton.. ................ Williauistown ............ Worcester ................
MICHIBAh'.
Adrian ................... Alma .....................
Cn unlet .................. C harlevolx ............... Detroit ................... &ranabn. ............... Grand Haven ............ Grayling ................. Harlmr Bench ............ lialamaeoo ............... Janning .................. h t h m p .................. Narquette ................ Ontonagon ............... Port Hiiron.. ......... _.: Reed City ................ Rt I nace ................ ~u p t Ste n!arie ...........
A1 y n ...................
Tmverw City .... :. .......
M I N N W T A .
Beaver Bay ............... Du1ut.h ................... Grand niendow ........... Lake W i n n l h i ~h i s h .... L-mg Prairie .............. nf iiineapolis.. ............ Noorhend ................. Morris.. .................. New Ulm ................. Ri lev. Fort ............. &iiw Green ............
9.90 30.16 '23.80
.x. 92
*25. zw 2% 75 24.12 22. $8
25.42 27.00
1.99 19.14 3 .7 9
F. 55 st. Pad.. ................. 8t.Vincent. .__ __. . __ ._ _. . Wahnsha .................
...........
Canton .................... Cstlettahurg ..............
EESIZi5:::::::::::::::
Lexln tan Lauis3lle 1.. ......... Mou& Sterling ............ i
........... i6.65 I Biloni .................... 44.01 I Brookhaven .............. 44.97 I Canton ................... 46.i6 . Columhuw ................
64.10 57.49 51.16 54.03
aEureks Rauch combined with Fort Hayes.
APW, 190% MONTHLY WEATHER REVIEW. 211
ivernge I annu+ preripi- tntion.
Stations.
- --
~
Geagraphiml coordinates, eletwiion, Iength of record, and A VERA QE d NNUAL PRECIPITATION for rainfnll eldona in the Vi&d State8 and cbnada
Latitude,
.for fhe period 1871-19014~ontinued.
nlssmwpI-eon tinued.
Qreenville.. .............. Meridian ................. Vicksbn ................
Waynesburo ..............
water V3ley.. ...........
MI880URI.
Birch Tree ............... Conception ............... Columbia.. ............... Qlasgow .................. Eannibal.. ............... Hermann ................ Ironton.. .................. Kansss City .............. Lebanon.. ................ Miami .................... Oregon ................... ,Princeton ................ St. Louis .................. Springfield ...............
XONTANh.
52. 75 51.28 *.55.93
* 46. i7
:*.m 37.52 34.55
*%.YO 38.06
* 44.11 36.09 44.:H 35.M 36.12
.............
h d S b U r f . .. .:. ..........
i
............. Mesilla Park .%nta Fe.. ................ RoJwell ..................
NEW YORK.
Alimnv ................... Angelira ................. Auburn .................. 37:E 1 40.70 44.12
Buffalo ....................
Elniira ...................
Cooperstown .............
*14.38 *21.83 21.21 13.40 13.21 16.31 14.?5 14.03
17.61 15.70
*?l. $7
Gouvenwir ............... I Hnueymead Brook .._ ___. Humphrey ............... Ithaca .................... K w w Valley.. ........... Lowville ................. New Tork ................ Oswego ................... Oxford ................... Penn Tan ................ l'lnttnfinrg. ...............
Kochester ................ . LTtica .....................
pCJt&lm ..................
13.00
15 93 *15:70 I 23.36 . 15.M 1
NORTH CAROLINA.
Abshers .................. Asheville .................
22.11 14.D
19.75 *14. 43
3l.M !
*IS. s!! 7.11 13.50
x. $2
3.65 5.57 *]A37 *12.41 5.42 8.08
*. 89
8. 54
Y O . 47 I . 99 Si. 13 $2.90 42.93 44. la $3.50
39.94 -9.99
49.18 48.09 47.30
**. 00 49.00 46.97
Berlin .................... Bismarck ................ Fort Abererombie .... ..-.I
Rerord. Record. iverage annual precipi- tation.
-
Years i n c h rive.)
Longi- tude. Elevn tion. Longi- tude. Elevs- tion. Stations. Ihtitudr 'rom- TO- 'mm- TO-
0 1
333 32 2: 32 Z! 34of 31 41
sa% 39 l e 39 41
38 41 37 3 39 05 37 42
39 lr! 3959 40 25 3x 3: 37 1"
45 42 48 25
45 41 47 50 4 i 28
47 06
4834 46 34
48 4u 47 01 46 2i 4fi 25 46 54
48m 13 45 10
42 w 41 25 4240 4009 40 31 40 49 4058 4Q 30 41 OR 41 69 41 16 41 02 42 49 41 09 .Lo 21 42 50 42 00
39 29
40 38 41 &q
$9 1u 4058 3833 39 49 -1038 37 24i
37 56 39 33 41 07
%?o
40 63
44 10 43 12 43 42 45 35 42 4B 44 01
u 16
39 22 40 M 40 15 40 44 40 30 41 01 40 14 39 L9
0 1
91 01 86 44 9053 8935
91 29 94 40 Y2 14 92 52 '91 20 91 -23 90 37 94 37 92 41 93 15 $5 09 93 %2 90 12 93 18
f i t . l2t 375 24i
3oE 19 I
9!W 9tE
i N 74% 5N 5%
925
w 1, w 8'22 1,llY 1,02C
565
1,324
3,010 2, m 4.m 2,730 3,350
1,900 2, w5 4.110 4,149 4,2Y2
'4,800 2.371 3, .a5 2. m ?,2% 5, cioo
3 761
3, $21 1 421
1, l8Y
1, $30 2,lQ 2,*21
1.532 1.105 2,022
4.090 1.114 2, .w 3. .w
1: M
3: 278
......
6.211 5,311 4,700 4. 7m 6.929
4,569
4.07'2 4. z3G 6, ,500
6. 110 4.4.M
G, 976
6. fm
4, 344
1.470 I550 603
1% 5i5 0,279
......
52 619 187 140 61 678 60 119
0 ,
3559 3505 8 2 8 32 46 3338 3329 3551 35 L3
35 12 32 a 2 0
35 41 33 28
E 2o
42 39 42 18 $2 55 $2 53 42 42 42' 05 44 I 41 60 42 14 42' 27 44 10 43 47 40 43 43 29 42 ?.$ 42 42 44 41 44 41 43 ox 43 06
3620 3533 3558 36 13 35 15 a505 3600 3000 36u5 3505 96 15 35 46 3365 36 24
M 14
4623 46 47 40 24
47 35
47 57 46 11 47 ?s
48 56
48 09
40 49 3906 41 3u a95R 41 00 41 16
a930 39 55 40 4s 39 28 40 12 38 42 41 3i
40 25 41 40
41 3i
44 %5 42 01 46 11 44 60 4305 43 00 15 40 4,457 42 41 4325 45 30 45 '3
0 1
103 M 106 39 101 u5 108 03
107 00
105 38 101 57 1oX 31 104 31 108 41 106 45 105 57 101 ?9
p 45
I8 03 76 I 78 63 74 57 76 M 75 35
7b 3 73 51 75 33 74 00
15 32 77 01 73 *?I3
75 13
;; ;
26 35
24 57 17 42
81 05
I8 M
15 40
83 25 75 42 81 2P9
81 50 tu02
78 01 77 30 77 57
?XI
p 51
%!R
Y830 1 w SR 96 46 101 30 98 57 100 34 %Yo0 97 10 103 35
81 9 8430 81 $2
e300 ,33 35 81 29
84 30 82 19 6230 81 ,243 s3 32' $253
$240 $0 41 8351 $407
1 9 02
122 33 124 00 11i 50
1?A 15 119 00 121 .50
T23 05 El 60 1?4 11 123 33 1'21 60
Feet. 4,700 5.200 3, 1% 6. 010 4,619 7.500 6. S35 6. M9 5,272 4,245
I , 013
3,570
2.873
97 1.510 715 767 1,250
$Go
400 450 1. $50 817
1. ooo LJOO
314
335 ail? 750 125 394 50s 473
...... ?,*SI 5uo 773 11 3, 817 R
1 1%
1.614 1,013 376 34 $1 78
3: 900
1.470 1.674 935 1,055 1,565 1,670
760 1.875
1,070 62n 762 s24 782 1,153 975 740 1.1M
650 1,095 52i 8'29 670 6% 800
......
214 1,940 '3 3.471 55
125 500 4,200 72
8,
......
......
Inchcs. 16.76 7.58 11.85 13. 11 10.89 18.26 19.96 13. t46 16.00 8.14 9.40 11.52 *18. 31
39.14 8 .9 7 35. !t8 37.64 39. I 33.19 31.15 41. $9 44.46 32.75 51.40 35.41 45.42 36.80- 38.97 23. 90 3 .4 5 30. '29 33.22 41. $9
*60.00 43.13 48.36 50.01 63.35 59.37 55.03 52. up *0.2.0 61.43 48.99 48.94
.w. 12 10.53
5% 01
*21.* 17.76 23. 78
16. Y! 17. lUl
16. ,5s 17.?2
w. 54 14. -2
39. $5
41.43 35.69 3i.63
d7.63 38.64 41.17 43. ,59 42.07 40.14
34.73 41.34 32.48 39. a,
9.44
?O. 75
44.60 Lo. 15 77.75
4 19.9- 66.69 10.80 0 .3 7 39. ti0
21. 90 SO. 17 135.81
* 103. si
1M7 1 W l.WO 1,Y6 lIw2
1893 18.93 1889 1872 1&54 1874 1378 1870 1875 1M7 1855 lW! 1S3G 1.932
lS79 18S3 lRix 1*9 lS92 la39 1 W 1W 1894 1861 lsgo 1R7i 1870 1882 l a 5 M71
18$3 1876
lppli
1 W I W 1875 1 W 1M9 1%7 1873 lf67 187s 1871 1872 1878 1 M 1890
1877 1870 ltctici 1875 lS70 1 W 1870 1870 1890 1877 1870 18iO 1 m 1870
1892 l%% 1831 1Ki7 lW5 1m7 I871
1873 1867 I874 l.F-43. 1.354
1879 l&65 1867
1901 1901 1901 1901 1901
1901 1901 1901 1901 1901 1901 1Wl 1901 1901 I901 1901 1Wl 1901 1901
1901 1901 189$ 1900 1901 1901 1901 1901 1901 1901 1901 1901 1900 1901 1901 1898
1901 1901 1901 1901 1901 1901 ls00 lW1 1901 1901 1901 1901 1901 IS94 1901 1901 1W
1900 1900 ISM1 1901 lua, 1LJOO 1900 1 W 1900 IS93
lua, 1900 19uo 1901
1901 1901 1'301 1595 1Wl 1901 1890
1901 Is00 1900 1900 1900 1900 1900 1900
14 9 4 i 16 9
8 12 r2 22 12 26
23 28 1"
4b
14 ti5
15
5';
19 8
llj
llj
10 1''
21 18 7 12 12 ?4 19 14
7 17
1s 26 13 14 7 19 20 33 34 11 33 24 10 15 P
14 7
10 25
P a 2 4 25 14 2fi
29 10 10 25
b 9 31
9 45
5'2
23 16 13 16
27 14 23 55 47 11
'18 29
1890 1850 1889 I S 7 1 W 1856 llr51 1%. 1.W llwl l8Y2 l&W lM9
1826
lY27 1%2 1854 185.2 1833 1881 11(83 l%a 1879 1 8 7
1844 l*-! 1l3-3 1840 1828 1830 1826
iaxi
18%
1897 1857 1666 1871 1874 1878 1875 1871 1895 1m2 1890 18% 1875 1872 1871
1891 1874 1860 1867 1869 1 W 1889 1871 186ti
1882 1836 l a 5 1878
lppli
1838 1Rcis 1 W 1887 1817 1852 1R30 1859 14330 1861 1871
18C6 1889 l&W lM9 1878 lLW7 1879 1870 1864 1$89 1892 1%
1901 1901 1901 1901 1 M 1$95 1901 1901 1901 1901 1901 1Wl 1901
1901 1901 1901 1Wl 1901 1901 1874 1901 1901 1901 1901 1901 1901 1901 1901 1901 1901 1895 1901 18%
1901 1901 1901 1901 1901 1901 1901 1901 1900 1901 1901 1901 1901 1901 1901
1901 1901 1901 1841 1890 1901 1901 1901 1901
1901 1Wl 1901 1901 1901 1901 1901 1900 1900 1901 1901 1901 1901 1871 1901 1901
1901 1901 1901 1901 1900 llwo 1901 1892 1889 1901 1901 1901
11 19 6 24 21 19 51
33 17 16 9 42
B
74 13 31 47 48 9 9 17 18 41 10 34 €6
46
35 56 311 25 72 41
6 21 16 25 -22
13 26 2Y 6
19 12 21
2F,
2Y
31
9
15 21 20 12 22 34
2
19 67 46
'23 1" E 33 16 11 74 49 69
41 39 40
2%
!XI 17 39 12 22 11 21 22 10 10 10 6
Carlshd .................. Fort Baywd ._ __. . _. __. . __ Fort Craig ................ Fort Stanton ............. Fort Union ............... Fort Wingute.. ........... Gtillinaw SDrinn . __ .__ ___.
107 34 114 15 111 03 110 41 111 15 101 31 109 40 112 01 112 E& 109 30 110 19 105 49 114 10 105 10 115 .W
112 00
Crow A$ney ............. Columb a Falls ........... Boreman ................. Fort Benton .............. Qreat Falls.. ............. Qlendive ................. Havre .................... Belena ................... Kipp ..................... Lewiston ................. Martindale .............. Miles City ................ Mimula.. ................ ..................
I
......................
Poplar.. Troy Virginia Cfty .............
NEBUASKA.
Fort Robinson ............ Qenoe .................... Ha Springs .............. Hegron. .................. Imperial ................. Llncoln .................. Ma uette ................ M d e n .................. North Platte ............. Norfolk.. ................. Omaha ................... Ravenna ................. Santee.. .................. Sidney Barracks .......... Temmseh ................ Vulentine ................ Whitman.. ...............
NEVADA.
la? s 9740 102 38 9734 101 31 96 45 9800 9s -56
IOU s5 9723 95 .56
w M 97 43 102 69 96 11 100 3!2 101 30
NORTH DAKOTA. 1
Fort Stephenson.. ........ Fort. Totten .............. Fort. Yntew ................ Onllatin .................. Prmbinn ................. Williston .................
' OHIO.
Canton ................... Cincinnati.. .............. Cleveland ................ Columbus ................ Findlay .................. Hudson .................. Jacksonboro.. ............ Logan .................... Mnnsfleld ................ Marietta.. ................ North Lewisburg ......... Portsmouth .............. Sandusky.. ............... Steubenville.. ............ Toledo. ................... Wauseon .................
OUEGON.
Albany ................... Ashland .................. Autoriu ................... Baker City ............... Bandon .................. Camp Harney ............ Cwnde Locks ............ Eolu ...................... Fort Klamath ............ Gardiner ................. Glenora .................. Government Camp.. .....
117 01 116 62 117 40 119 46 116 27 118 34 119 02 118 14 117 44l 114 I 119 4 i
114 26 116 2i
117 43
Austin.. .................. Battle Mountain ......... Camp MrDermitt ......... Carson City. .............. Halleck .................. Hawthorne. .............. Hot Sprinm .............. Humboldt ................ Palmetto ................. Pioche. ................... Reno ..................... Taano ....................
dnnemucca .............
T b o .....................
NEW HAPPSHIUE.
Bethlehem ............... Concord .................. Hanover.. ................ Lakeport ................. Nashua ................... North Conway.. .......... Mount Washington.. .....
NEW JERBEY.
21 35 ,1 L*
72 15 71 30 51 29 il 01 71 18
74 25 74 34 74 16 74 1u i 4 27
I 4 45 75 01
24 45
Atlantic City. ............ Dover .................... Freehold ................. Newark .................. New Brunawlck .......... Newton .................. Trenton .................. Vineland.. ...............
818 MONTHLY WEATHER REVIEW;
Stations.
APIUL, 1902
Lati tude
____~~ ~~~
Gmqraphkd coordinates, elevation, Lzg& of record, and A VERBGE ANNUAL PRECIPITATION for rainfall slnliona in tlre United &ate8 nnCE Cbnada
for the p +o d ISrl-1901-Continued.
Stations.
Record.
Latitude
Avernge annurrl preclpl- tation.
Texas--continned.
Dnllnx .................... El PW> ................... Fort Brown .____ __. . __._. . ForL Clark. ............... Fort Cnnchn. ............. Fnrt Davis ................ Fort McIntosh. ........... Furt Kin old ............ Fcwt Stoc%n ............ Fort Worth ............... Frtdericksburg.. ......... Giilvestun ................ Houston.. ................ Muniit Blscnco. __ __. . ____. Palestine ................. Man Antonio .............. Wac0 ..................... Wt.atherIord .............
UTAH.
Record.
I
Average annual precipi- tation.
-
Yenm inclu. sive).
-
Team i n c h Rive).
Longi- tude. Eleva- tion.
Longi- tude. Eleva- tion. kom- TO- prom- TO-
oREooN-continued.
Grants Pass.. ............. Happy Valley ............ Hep ner ................. HJ River .............. Jwph.. .................. Lakeview ................ Newgort.. ............... Pen leton ................ Portland ................. Prineville ................ Roeeburf ................. The Dal ea ................
PENNSYLVANIA.
Altoona .................. Bethlehem ............... Carlisle.. ................. Confinence ............... Dyberry ..................
Erie ...................... Franklin ................. Gettysburg ............... Ciirardville ............... Gramplan ................ Harrisburg ............... Lewlaburf.. .............. New Cant e ............... Philadelphia ............. Pittuburg. ................ Warren.. .................
Emporlum ...............
REODE ISLAND.
Block Inland.. ............ Providence.. .............
SOUTH CAROLINA.
Aiken .................... Beaufnrt.. ................ Charleston ............... Cheraw.. ................. Columbin.. ............... Camden ..................
Yorkville.. ...............
.............. ...............
WUTH DAKOTA.
Aberdeen.. ...............
0 1
42 Z !
43 00 45 20 45 $2 15 19 12 00
44 39 45 40 45 32
44 20 43 13 45 83
40 32
40 3ti
40 12 39 57 41 37 41 :w 42 07 11 24 39 49 10 47 41 00 40 14 40 -53
41 02
39 57
40 s" 41 5 i
41 10 40 -50
33 32 E '11;
32 47
a4 00 34 15 37 26 5; %5
3 4 3
y 52
45 27 45 45 -11 w2 -11 26 .I3 M 4.4 39 $3 48 -11 21 43 12
44 22 44 M 42 M
3630 35 XI
31; 1s 35 v4 w, 28
%5 .53
I 1 0
35 V i
%I 56
3509 36 3ti I
%!I 20 31: 30
s5 25
32 .23 w 1s 30 l l i
3u .% 2% 39
27 49
29 03
3u v<
O I
123 20 118 30 119 30 El 28 117 03
12! 1'2 1% u2
118 45 1 2 43 120 5i E3 'Jv E1 12
78 24
r7 14
I5 1x 79 15 xo (15
76 1s
ib 32 iti 55 rw 21
75 09
I9 14
7.5 "3
79 L!
79 .w
7! 35
17 15
p u2
il 3i
71 24
91 40
MI 41
I 9 57
Nl 31 79 NI
xo J
81 13
79 ,51j
xi a?
% 26 1IPJ Frl
9ti 3i
103 3
9H 52
100 as 98 .w
94 14
loo "1
livj E Y i 28
I& 111
Fd. 964
1,200
1,950 9-w
4.4w
5, oflo 68 1.074 154
3. Oou 518 106
1,181 s5n .m 1,324 1. 100 1.0.511 713 95.5 (24
377 -I.-*
no9
117 .ul2 1.13i
1,m.i 1,400
27 155
.%5
28 48
144
73 19
Gw
$2
1,300 3.192 1. ai5 3, *A 1.95 1,600 1.7- 1,:M 3.3:* 1.57" 3. "34 1:233
Inches. 31.35
* 3.51 *21.42 .E. $2
17. il li. 14 7% '20 14.55 45. 28 1l.M 35.26
1s. w
3s. 44
4 .2 9 41.84 45. 26
19.3i 44.25
39. P .E. 51 39. oi 82.91 45.11 41.60
41.51 36.54 41.47 35.64 43.18
45.09 4ci. 01
4s. 10 46. Y8 4%. T2
41;. 14
49. li5
. 44. $2
*51.21
4% i V
4i. !Ji
3%. rd
13.9s 23.22 1s. 91
2v. *i
l i . :33 1% 44
'10. tii
20. liq
*]ti. i 9 15.92 35. .w
49. 97 51.40 -7. "3 52.71
45. titi
4i. ?'2
w. 01
.w. 97 49.15 *45. s4 rw. SS
*:a. >w 47.63
:I. I . d., ?!
24. ?'2 21.55
33.51 3%. 31 2s. 1 2
8. o(;
2ti. 28 33. 7ti
0 1
s2 -55
31 47 35 .50 I 17 31 55
300.10 27 29
20 27 3050 s2 43 30 LW
33 5.5
31 4.5 29 27 31 35 32 57
E ::
41 34 39 0.5 40 %5 39 34 38% 37 51 41 14 Si 13 40 46 41 82
4.1 ?a 44 2s 44 57 -1109
43 w 42 47
36 59 a7 25
36 M 37 a5 38 40 37 44
37 25
3(; 48 36 51
37 32 37 (r2 3q M
24; 55
48 15 46 5.5
47 uo 46 lti 4s 57
46 30 4 i M 4ti 58
4n 35
41: 40 Jk ?2 47 a5 4s Iyi
47 P 4 i 40
47 16 45 40
415 G? 47 40
3s 45 Si 44 3s 25
:is ;w
19 39
39 115
3Y 2s 39 01 40 07
?7 .w
0 1
90 3s 106 30 97 57 100 25 100 17 1lU O i YY 31 b 47 1R2 35
97 15 9n 4s 9450
95 1 Y 101 01 95 40 98 28
97 57 w ar
112 06 110 ox lW M
111 63
109 29 1E 51 111 .w 113 50 111 54
113 33
73 12 71 44 72 18
72 h R '25 72 32
22 52
p '3
15 Y! 76 00
I8 I 5
81 26 71; 17 7i 27 76 -53
78 30
81 05
g 2
118 10 117 20
P2IJ 35 1M 03
117 5i
12% .w 1% 00
119 30 1'10 30 l'24 32 1'2 :XI
1% Mi
IT2 w 117 25
1 2 23 E 2 30 11s 20
Pa 0.5
i i n 40
.so 00 s1
nl 81
so -5-4
7 i 45
so 10
I9 .5Y 21 31;
I9 3'
w 27
<w 52
r1 w
Inch. 33.22 8.84 25.52
21. Si 23. io 18.10 19.06 19.80 16.10 34. 32 3 .3 2 48.13 45.20
15.59 44.14 ?8. 41 31.80 30. MI
I
11.75 7. 26
6. 46
lti. 00 7.43 E. I1 14.15 6.73 17.5 4. .%-
Feet 4b6
3,762 57 1.0.50 1,%W 4,700 460 230
4, Y52 670 1,742 -41
.is
510 701 424 scil
4,232 4,541
5, Oou 5.010
......
4:g 4,366 4, ,w
34ti
1.124 7,50 871 50
310
2, OOO 0.10 18
2, 1w 1.350 946
li81 2. PA 91 '39
15 927 2,D.i
3. ooo 2.m I, 5ii 14 1. si3
9x5 l .l i 3 1:30
1.WO
40
15 29 123 1.i93 11
.w LOW 2, G24
......
2,2w l..W 5's
6%;
srw
1.400 $94
638 l.W2 e 4
li37
......
1 S Y 1890 18x9 1891 1xR9 1.W3 1x91 18!W 1858 1 a92 1877 1850
lS59 l8i7 l%Y
1x74 1x67 l<W9 1873
l%9 IS% l d 4 l97i IWi
I S 0 li9S lS% 1*5
1.w
1SW 1831
1851 18-37 1738 1w2 1M IMY
1992
lxxl 1.57
1990 lSY2 1890
l X i Y
Mi lMi9 1WI 1%%1 1.W 1SYl IS31 la73
1 K33 Hi9 1!UB
l X i Y
lVl.4
1w:i 18G9 I,%5 165.4
1M11 1VSY lSS3 1W
1AY8 lW
1*%5 1 892
l%%i
lRsj
1MY 1MIi I S 3
iasy
l'M1 1900
1901 1901 1901 1901 lWl 1w1
1901 1901 1901
i n s
1901 1'500 1WI 1901 1M 1901 lC901 1901 1 w 5 1901 1901 1901 1901 1 R(U
1901 1Wl 1941
1901 1901
1894 1901 1901 1901 1901 1901 1901 1901 1901
1901
1tWl 1XW 1M
lXY4 1'Ml 1901 1'901 1w1 1901 1901
i y n i
1901 19nl 1911
l9Vl 1901 1901
19nl 19nl 1901 1901
1901 1w1 1901 1901
i!mi
1901 1901 1901 lW1 1900 1901 1901 1901
12 1v 9
10
E 9 8 11 $2 10 '?A
3:!
17 21 22 2.5
2s 13 23 24 35
11;
27 24
19 17 85 57 17
2' i 0
28 15 94
13 19 $2 9 1Y 15
9
Y
1'2 1ti E '15 15 20 Y 10 15
2ti
15 2
9 23 32 ]A
li lti
38 34 3i 1 d 1!1 4
It;
1ti
10 Si r2
Y
2n 14 10
1W 1K50
1W' l l i ? 18.55 18.49 1849 lW EM9 1857 1568 1W 1W IN2 1K19 1.967 1W2
iaw
1870 l8Yl 1% 18-3Y 1M!I 1890 1R70 lWi7 1K57 1870
18% 1848 1869 1887 1873 1.9-5
1891
ILW 1874 1K35 l&W 1869 1871
1851 1&5v lKsH IW9 LxriU
i n n
I R94 Ish1 1,W 1874 1%. 1X5i 1Wl 1R9i 1x94 1 w lW ISi7 l,WS
lXY0 1,Wl l&% lMY 1K57
18W
1s94 lX'22
1N1i
1W
lW7i 1890 1873
1,W lRs4 1%- lW2
inw
1901 LW1 1w1 1901
18R9
1991 1WO 1901 18W 1901 1w1 1901 1WI 1901 1901 1w1 1901 1W1
1901 1Wl 1901 1901 1w1 1901 1Wl l!W1 1Wl 1901
1901 1% 1689 1901 1897 1901
1901 1901 1901 1901 1901 IWl 1901 1901 lwll 1901 1901 1W1 1901
1901 lMll 1901
1 s9Y
lrw)
1999 1901 1901 IW1 1!D1 1Yo1 1901 inw 1901 1Wl 1901 1w1 1Wl 1901
1901
1'Ml 1901
IWI 1901 I S 9 1901 1901 1901 1141 1901 1901
9 3fl 28 28
15
'LO
34 38 14 8
17 33 12 1s 19
31 14
8
31 7 14 11 1"
11
L!
12
;:
59
44
15 14 24
16
I1 31
* 1'2 22
2%
30 17 31 24 11 8 29
R 1:! 13 21 18
11 11 4
6 9 18 24 I"
Y
2v 13 23 31 10
R
Y
14 14 11 1'2 11 1 Y lti
15 14
16
Corinne .................. Drxerrt ................... F ~r t IJiiclicnne ........... Lrvan. ............. __.:. . Mnnb.. ................... Pflrnw ail ................. I:lgrlm.. .................. Ht. George.. .............. Salt Lnke City _____. __.___ Terrace. ..................
VERMONT.
Burlinftnn :. ............. Luncn urg .............. New urt ..................
Btraffnrd.. ................ Vernon.. .................
N o r t h l d ................
VIRGINIA.
Bigstone Onp .... .:. ...... Birrlxnest ................. Cnpe Henrv .............. Christicrnsl;urg ........... 1)ak Entrrpriie ..........
I.)'ntth lllrg ............... Msrion ................... Norfolk.. ................. Kivhmund.. .............. Spcdtsrill e.. .............. Wna-dstncli ............... Wytlieville ...............
Lrxinqtnn.. ..............
\\'ASHINOTON.
32.92 39.60 12'. gci
34.02 39.97 4s. 77
hil. 70 49. 50. 80 3K L*J
$2.75
40. ?u 44. lli
43.10 50.40 44.41
.w. 53 Y5.3 $2.17
.._~~
~~~ ~
Aabrroft.. ................ Flandrcnu.. .............. Fort h l c d e ............... Fort Ban~lnll__ ........... Fort Sully ................ Kimbnll .................. Huron.. .................. 20. WI 3.61
8. PG
05. oi li. 30 E. rdJ
12. w 11. b *13. 49
8. ti2 n2. 40 .%5.37
L!. $5 35. K5
32.31 5. c3
li. $2 13.35
in. 10
Vrilcinin .................. (!ulfmx.. .................. Ellrnrbnrg ...............
Fort C!olville .............
Fnrt canby.. .............
Oelrirhs .................. Pierrc ....................
Yanktnn .................
Rapid City ...............
Fnrt Siincw .............. Lalieside ................. Lirid. ..................... TEXh'NEE.
Andcrsnnrille ............ Ash\volid ................. Elieabethtnn ............. Chathnwm ............. Clarksville ............... Florence ................. Greenevillt.. .............. Johu~mville ............. Knoxville ................ I
Lammis ................... Mt#xcc Vrllll~y ............ Nrarh Bay ................
I llylll~Jitl.. ................ h r t Angeles ............. HL.;LttlC.. .................. Spnkane .................. Taroma .................. Vrinrouver ............... ~Vllllfl Walls.. ............ Waterville.. ..............
\VEST VIRGINIA.
~~~~~~~ ~~ ~
Memphis ................. Nashville.. ............... Rogersville ............... &avannnh ................ c;ilverlake ................ Wryncabnro ..............
TEXAS.
Abilene .................. Amnrilio ................. Austin.. .................. Brenham __. __. ._ __. __ _._. Burnet ................... Camp Ea!Ae Pnw ......... Corpns Christi.. __. ____ ___ Cnero.. ...................
*re. 00
46.09
44.90
Y8. (il
35.31
M.6U S39.61 47.08
42.78 -1.83 48:2
40.00
Rtwrly .................. Elkhorn.. ................ 1 'hurlerton ............... Cilenville ................. Hrirperw Frrry ............ Helvctiu.. ................ Hiliton ................... MnrgRntnwn ............. ParkerslJnrg.. ............ Kowleabnrg .............. \Vrst1,n ................... Whreling.. ...............
9Y 40 101 50
97 43
Qlj w2 SY 01
100 30 Y i 25
9 i 09
1.73q 3,6iti
E50 w I,%
w 1 R
l i 7
- APRIL, 1902. , MONTHLY WEATHER REVIEW. ' 913
Geopaphical coordimtee, eletmfion, length of remrd, and d I'ERAGE dA7NU;IL PRECIPITATION for rainfall stcltio)Le in the W t e d &ate8 arid Canada
for the period 1871-1901-Cantinqd.
Indice. 30.60 31.m 29.31 30.66 30.92 30.66 3.11 29.7s 32.04 31.17
1 Record.
I
0 , 0 1 WPOHING.
Fort Bridger .. __ __. . __._. 41 14 110 32
, Fort Fred Stet+ _.__. . __. . 11 47 106 57
I Fort hromie.. ...........
~
Fort McKinnev .......... U 23 loti 46
. Four Bear ................ U 22 : .1W 49
i Sheridan ................. W 49 I 107 00
Cheyenne ............... 41 Ox
1
101 A9
I Fort'F"ettr.rmnn .......... J'L 50 I 105 ' ~9
42 19 I 101 31
: Fort Yellowstdne.. ....... 44 59 I 110 41
: Lander .................. $2 60 i 10s 45
. Lusk ..................... 42 47 j 104 27
I
- stations.
1
Lntitude
I
Record. kversgo annual
tation. precipi-
I,reRes. 12.96 9.54 13.90 10.10 15.14 sg. '99
19. 90 11.S5 13.45 E. 80 *11.14
-
Tenrs (inclu, sive),
.-
Years (inclu. sive).
-
TO-
Elevn- tion.
Lon@- tude. From- TO-
0 1
45 47 44 31 I 01 4523 43 49 43 05
4405 44 1" 45 10 4302
WIWONSIN.
Grantsburg. .............. Greenbn ................. Haywnra.. ............... Koe nlrk ............... LS &La,.. ............... Mndison.. ................ Mnnitowoc Meadow VHllry Mrdford.. Milwaukee ...............
............... .......... ................
0 1
9204 H800 91 30 89 11 91 16 89 24 87 s9 90 15 90 -20 87 54
1Rs9 1spg 18LPJ 1890 18T2 1865 1861 1891 1 W 1870
Rrt.
6. Wi 6613
7. &Io 4, 270 5, oou 6.370 G, .m 5,372 5 007
5: 250
a: 750
1901 1901 1901 1901 1901 1901 1901 1901 1901 1901
8
15 7 11 L%
%5
SI 9 11 31
1870 1858 1 W I 1%9 1K19 1W lW 1893 1M 1889 1891
1901 1890 lW2 1SS 1w1 1 W 1901 1901 1901 1901 1901
31 15 14 11 22 e 13 G 13 1'2 8
... 955 616 954
1, *20 6x1
h n g i - tude W.
-
0 1
M 42 80 10
63 39 . &4G
ci(; 02
63 10 6.5 20
11 13 73 %5
77 55 75 $2
76 %I
7Y 24
$0 3 81 13 Ul 21 1100
y? 19
........
Eleva- tion.
-
Pd. 125 M 11s
. 19 85 35 21 20 29G
187
47"
330 -3% 350 1.2.52 ,592
656 655
.......
.enel
3f T r c
Ord.
ITwe. '29
27 "7 17 11 26 27 24 "7
* 21
18 27 61 12 27
-2.5 25
-
......
Record- 11 I Eleva. tion.
--
F&. 64.4 760 1,690 2.116 2,161 2.439 :1,38y 4,M!
2.15s 1. JU?
1.6% 1,193 d 4,1110 770 1,191 ...... ...... ......
Lengt Of rec Ord.
--
l-mre 11 28
18 18 17 15 17 6 18
16 10 12 10 13 21 2.5 15 12 14
Record- Mean nn
cipiw tion.
I*CkCn. M. 63 .50.2.5
57.03 47.25 50.33 41. ti2 41.29
3'2.99 41.72 40.99 30. ui
32.80 E. 211
33.72 24.59 34.42
31.21 8. 27
lUHl PlV
.........
Uran nu iunl pre ci ita- tEn.
IllPIle8.
pr. 76 20.9x 16.45 16. w 13. w 15.47 14.87 21.91 1.5.83 14.91 13. Y3
11.63 3s. 14 33. .56
8.81 36.04 37.2q 22. -23.86
~nnHdimi stations. -
ends-
-
1900 1900 1 9 0 190 1900 1900 19W 190 1900 1900 1900 1900 1900 1900 1900 1900 1900 1900 1900
t%$.
0 1
89 IS 97 07
9948 lo3 42
110 37 1w 45 114 v2
115 35 113 30 106 0 108 3u 120 a P?3 19 1'21 35 111 20 81 0
7Y 40 x3 28 so .w
be- gin+ _-
1872 1874 1874 1XW lW 1U7.5 IS74 1477
lS74 1W7
1 W 1874 la0 ln!, 1874 1876 1876
in74
......
br-
gln9- -
lsso 1873 1883 1893
1 W 1M 1 W 1!% l W 3 1% 1x91 1W 1Ml 1#38 1872 1860 lspg 1 W 1887
-
0 ,
47 34 46 10 44 39 4.4 47 43 50
ui 14 47 03
4S 31 .Ici 4s 45 30
46 1:!
45 35
44 13 43 39 48l LW
$2 40 44 30 45 19
........
I-
--
I: -
0 1
48 27 4Y 53 50 10 .50 44
50 01 50 -20 .XI (I2 61 10 .53 3;:
53 10 5'2 41 .50 41 18 21
5302 Ml LQ
4323
41 8 48 21 46 40
pnrt. Arthur. Ont ......... I 8t. Johns. N. F ........... 8ydnev N. S ............. Hnlif&. N. 8 ............. Grnnd Muntin. N. B __.___ Ynrmouth. N. S .......... Charlottetown. P. E. I .... Chatham. N. B ........... Fnther Point, Quc. ....... Quebec, Qiic.. ............ MOntrrHl. Que.. .......... Riirtt. Ont .............. Rockliffe, Ont ..... .:. .... Ottawa. Ont. .............. Kingston. Ont ............ Tomnto. Ont ............. White River, 1:nt ......... Port Stanlev. Unt ........ &ween. O h .. ........... Pnrry Sound, Out ........
Winnipeg. aan.. ......... MinncdoPa. Man ......... Qu' Appelle. N. W. T ..... bf.e!iicine Hait, N. W. T. .. Swift Current, N. W. T ... Cslgary. N. W. T ......... Banff. N. W. T ............
!pence's Bridge. B. C ..... Lstmtford. Ont ............ Coldwater. Ont.. ......... Misannble. Ont.. ........ Curtirr, Ont ..............
914 ' .MONTHLY WEATHER REVIEW. APRJX. 1908
(8 ) REXARKS BY PROF. C. F. XAR-.
Accepting the precipitation data as being npproxiniatelj
.correct, the first consideration that presents itself in the mat
.ter of charting it is the well-known fact that all precipitatior
is charncterized hy great local variation, even within limited
'tiiGtis, and it can not. therefore be supposed thtit observationr
at relatively widely separated stations within any particulw
' region should he very closely accordant. On this account ii
is obvious that where stations are widely scattered bhe data nl
the best but imperfectly show the precipitation over the wholc
area. A marked difference in the topography may of itsell
cause still greater vnriatioiis i n the precipitation throughoui
any region.
In making a precipitation chart, thewfore, one is confronted
at once with a dilemma: (1) He miwt either distribute the ob.
-served precipitation proportionately over the entire regioti
'comprising that field of observation, or (2) he must indicntt
-the afea surrounding each station throughout which the pre.
cipitation niay he assumed to he the same as observed at thc
station. If these latter areas do not overlap, then blank area?
are left within which no records of precipitation tire availalde.
The second method is preferable for scattered stations. hnl
the first is almost universslly followed, fanlty though it
may be.
I n the case of niountainoiis c?ountrier., and regions marked by
conspicuous topographical features we are confronted with
an additional complication. It is geneidly believed, and iE
in fact true, that the precipitation on mountain aumniits and
elevated ridges .ie, with eonie except,ions, greater than on thc
valleys and plains. This knowledge may be gained even with-
out the. aid of rain-guage re.dings. I n fiwt it ge.nerally happen2
that mtual measiirenients of precipitation-are not available in
mountainous regions. In charting precipitation over such H
region it niay be argued that the excess of precipit,ation
believed to exist on the suiiiniits should appear in some way
on the charts, but t.he aiiioniit of this has rarely heen inens-
ured, nor is tmhe extent of territory covered sharply defined,
and the student preparing thecharts finds a most serious dia-
ciilty in deciding upon the amount of precipitation attxihihdde
to the summit.s and in determining or defining the limits of
the area over which any excess should appear. A chart em-
hodyiug hypothetical features is so largely a product of indi-
vidual iniagination that it will probably prove acceptable o n l ~
to the individual who niakes it.
Judgment in these iiiatters 'nay be guided by the following
considerations :
The total precipitation orer the entire land And water sur-
fnces of the globe. is obviously equal to the total evaporation
from the sanie surfaces. Mrhen we consider only land areas
no definite relation can he formulated between the precipih-
tion. evaporation, etc. It is obvious, however, that the entire
precipitation over land areas is disposed of in the three fol-
lowing methods:
(1) A portion of it soaks into the ground; (2) another por-
tion flows through the streams and rivers to the sea; (3) a
third portion evaporates direct from the soil and river sur-
faces and thus returns to the atmosphere. Throughout any
particular area the measiirenient of the precipitation m y
give only a part of that which actually falls. I n some casea,
however, the observed precipitation may be in excess of that
yhich actually falls over the region. I n the present atate of
our knowledge it appears vastly more difficult to nieasure
seepage and the evaporation going on over an area than to
measure the rainfall. In some cases, however, fairly accurate
measurements of the run-oif of the rivers can be and have
been niade so that it can, to a certain extent, supplement t,he
nteasureinent of rainfall i n the fornixtion of precipitation
charts. The relstion Mmeen the prccipi t,at,ion over a drain-
itge basin and the riin-otl from the rivers is that t.he latter
must he less than t,he precipitation. When, however, one
at,tenipts to const.ri1c.t a precipitation chart hy redistrihuting
thc water represented by the run-off in rivers throughout the
area upon which it fell as precipitation, he ninst. not only
allow for seepage and evaporation, hut he is confront,ed wit.h
all the difficulties already mentioned of distributing precipi-
tation over regions for which no observat,ions have h e n made,
and the stat,enient already made is repeated that a rainfall
chart in which unmeasured rainfall is a.rhitrarily assigned to
regions to which it is supposed to belong is so largely a prod-
uct of indiridnal imagination t,hat it will probably not he sat.-
isfactory to any cscept the incliviclual who niakes it.. It would
seein, in fact, that where engineering and scientific interests
demand greater detail and ticcuracy in charts of precipitation
than is now attainnhle by the proportional distrilmt.ion of the
prectipitation as observed over the region embraced. then the
only scientific. and possihle retiiedg' is t,he increase in t.he cun~-
ber of stations, especially in rcgioiis of niarked irregular
topography.
-~
* (4) NOTE BY CLEVELAND ABBE.
Every geneixl weather service now puklishes it,s rainfall
records with s11c.h fullness that t h q are available for t,htB use of
those who wish to study special proldenis, such as the occur-
rence of destruchve rainfall. the average qtiantity of raitifdll,
the nuniber of rains days, t,he duration of droughts. bhe depth
of snow, e k . Not only are the iiunierical data puhliwhdd for
each station. but all agree as to the desirability of presenting
the data graphicrally as niups of rainfall. Hut the principles
that should bc followed in constructing these niaps have, so far
3s we know, never been codified and confirmed by any decision
of the Interiiat,ional Meteorological Congress or .by any con-
ference between individud ineteorologists and statisticians.
Recent discussions in the llloiithlp Wetither Review and
qwcially the. art,ic!les nnd charts in the present niinilier hy
Mr. Henry Gannett and Prof. A. J . Henry draw renewed
ittention to the difficulties t,httt attend the measiireinent of
local rainfall and the construction of uiaps showing the total
zeneral rainfa11 for any region. We, therefore. niay presume
;o offer the following suggestions with regard hoth to the
zauges, the stations, and the treatment of the data:
(1) I t is well known that a11 ordincry accurate rain gauges
3xposed to the wind catch less rainfall, and especi!illy less
momfall than they should, because of the effect of the wind
APIUL, 1902. MONTHLY WEATHER REVIEW. Zl5
in forming eddies and deflections that carry away the rain-
drops and snowflakes that should be caught in the mouth oi
thegauge. The stronger the wind at the mouth of the gauge
ao much greater the deficit.
(2) Gauges whose niouths are shielded froiii the wind by
special construction of horizontal screens or shields, as in the
shielded gauge of Prof. Joseph Henry, or that of Prof. F. E.
Nipher, show a very small defibit, if any; so, also, do gauges
that are protected froiu %he wind by a snlall open fence or
vertical sciwn around the gauge, m in the methods of Wild and
Boernstein; and so, also, those that are located on a depressed
roof, as observed by Hellmann, or those that are pivtected hy
the parapets of the roofs, as at most Weather Bureau stations.
There is, therefore, 110 douht that gauges can be so constructed
or so placed as to counteract the ordinary effect of local winds
in diminishing tbe catch.
(3) It has heen shown (see Monthly Weather Review, 1S99,
p. 466) that! gauges at ditferent altitacles above the ground, but
otherwise similar and freely exposed to the wind, show deficits
that increase with t,he altitude, presumubly because the stronger
winds at the higher gangen produce larger deficits. A com-
parison of two or inore siiiiilar gauges, located near each other
but at different heights. nborcls a means of obtaining an
approiiimate correct.ion t.o the redings of the lower gange, by
which to obtain the rainfall approximately free f roni'the wind
effect. If this principle be applied t.o the beat fornis of
shielded or protected gauges it, should give us better results
than when applied t.o ordinary unprotected gauges.
(4). Gauges of a variety of forms exposed to the saine wind
give diderent resnlt.a, owing fo the fact that the wind effect
varies with the shape, t.he proport,ions, and the sixes of the
gauges. The differences bt4.wee.n t,mo suc.11 gauges sometimes
varies with the square of the wind velocity. (See G. E.
Curtis, Bainf all o I I Motin t W tis hi ag t on. )
(5) Gauges exposed in apparently nnolJjectionable localities
in the same field show variations froiii each other of 5 per
cent, apptwently owing to the viirint,ions in the strength of
the wind, produced by the ordinary irregularities of the
ground; therefore nietisurements agreeing within this limit
may he considered as having tlie same weight or as identical.
(See C;. Hellmann, "Berlin Regenfeld.")
(tj) In view of the preceding paragraphs, all gauges that are
unduly exposed to severe winds, whether on open plains, or
hilltops, or high biiildings and which therefore catch much
less rain than actually falls, should 1.10 excluded from use in
preparing n; rainfall niap, unless proper corrections can be
applied. In general every gauge should have its special pro-
tection, so that there be no doubt. as to its records. Fortu-
nately there is but a sinall percentage of bnclly exposed gauges.
Those that are so located as to be well protected from the direct
action of the wind may be used for rainfall maps, since,.even
though placed in exposed locat,ions, they do not necessarily
give deficient rai nf a1 1s . (7) Railifall is largely affected by the minute details of local
orography. The station may be too much sheltered by neigh-
boring buildings or trees, so that too little rain falls on the
ground. When the wind is forced upa wounhin side it is likely
to give more rain than when blowing over a horizontal plain,
but it gives correspondingly less rain over aome regions to the
leeward of the mountain. When the wind blowR from a lake
or ocean onto the shore of a flat country it gives more rain
over the land near the water than over the lake iBelf or over
the land distant from the lake. These are some of the natuwl
cases of actual variation in rainfall, so that the chart of
isohyetals is much more coniplicated than the chart of contour
lines. It is very rare that rainfall stations are close enough
together to enable us to draw isohyetals showing all these
details. The rainfall on Barbados is one of the few instances
where it is practicable to study i n detail the connection between
rainfall and topography.
An extreme mse of the relation hetween rainfall and altitude
is presented by the records for the island of Ascension,
which lies in the southeast trade-wind region. This island
consists essentially of a mountain (known as Green Mountain
from tlie days of the earliest naviga.tors) end a lowland
which is mostly to the west of the mountain and which,
although dotted by hillocks, has an average altitude of
scarcely a hundred feet. I t is very rare that rain falls on
the lower part of the island, partly because of it,s position in
the dry trade-wind region and partly because the lowlands are
to the leeward of Green Mountain. The southeast trade is
deflect.ed upward and sideways orer and around thi.9 great
mountain. As the air ascends and cools abundant cloud is
fornied around bhe suniniit of the mountain, but as t.he air
proceeds onward toward the west or west-northwest, it rolls
over and oyer on itself, forming a regular series of isolated
clouds, st,ret,ching in a htraight line for a hundred miles to
the westward. These clouds represent the tops of succes-
sive waves of air. They look like a series of cigar-shaped
rolls and even t,he largeest of t,heni rarely allows a drop of
rain to fall. Although litt,le or no rain, properly so called,
falls on the island, yet the suniniit of Green Mount.ain envel-
oped in clouds is a beautiful picture of green verdure, aud
the drip froin leaves and twigs is carefully collected and pre-
served for the use of the naval garrison at the landing on the
leeward side of the island. This case illustrates the general
principle that the excess of rainfall on the windward side over
that on the leeward side must depend upon the height .to
which the currents of air are forced up by the obstacle, and
the general laws of this relation mnst be very analogous to
those. deduced by Prof. F. Pockels (see Mont,hly Weather
Review, April, lWl), for the case of the forniation of clouds
011 mountain slopes.
(8) A chart showing details of the clintribution of rainfall
must be based on the total precipitation either for a definite
nionth or year or for the average of n number of similar
months or years. In any case all the rainfalls that are charted
should relate to the same fundanienhl group of nlonths or
gears. The variation of rainfall from month to month or
gear to year is just as great a.~ the variation of rainfall from
place to place. Our charts should present the geographical
distribution' of the average rainfall duiing the same period of
time for all parts of the chart. The chart should be a chrono-
logical as well as a geographical unit. To this end all the
rltinfall records of short periods must be reduced, as accu-
rately as possible, to. what they presumably would have been
216 MONTJXLY 'WEATHER REVIEW. APRIL, 1908
~ ~ ~ ~~~~ ~~~ ~~
if the seiies had extended throughout the whole of the funda-
mental period of the chart.
The fundamental stations 6f any chart a.re those for which
wehave complete records at gauges that have not been changed
as to pattern, exposure, methods of observation. and record
during the whole of the fundamental period of the chart. If
changes in the location of the gauges or in other important
matters have been made, the different parts of the series must
be reduced to uniformity with some one of the positions used,
presumably that which was occupied the longest time; if 110s-
sible, the whole record should be corrected for the wind effect.
These 'changes i n the average observed rainfall induced by
changes of site and by strength of wind are important, hut
liable to be less important than the changes introduced by the
omiasion of the reduction to the fundamental niiniber of yeam
Any short series of observations may be extended chronolog-
ically up the full fundamental length by a process of extra-
polation, and we may plot the result along with the funda-
mental data for the fundamental fitations.
Experience has shown that when stations are near each ot.her
three years of record may by extrapolation give a fair t.en-
year average and a ten-year record may hy estrapolation give
a thirty-year average. (See Angot, Rainfall of Europe.)
The following exainple is hwed on data taken from Professor
Henry's Bulletin B, p. 92. The avemge r:iinfalls at Amherst,,
Providence, Boston, and New Bedford are given hy Henry for
six successive decades, 1537+N, as in the following table. Let
us assume that the record (49.4) for Providence for the fourth
decade is unknown and that we require it before we can obtnin
the desired average for sixty years at that place. The numer-
ical process assumes simply that the total rainfall is distrib-
uted according to some simple law throughout the region.
We first take the average of the five known decades at all four
stations. We then for Amherst, Boston, and New Bedford,
respectively, express the rainfalls for the fourth decade :M
fractions or ratios of that for the remaining five decnde3.
Assuming that each of these fractions for Amherst, Boston,
and New Bedford has an equal chance of holding good for
Providence, we apply each to the Provideticbe record for the
five decades and compute the hypothetical value of the missing
fourth decade. The average of these three values, 47.1Y, 51.09.
and 51.51, gives the desired interpolated rainfall, 49.92. We
might have varied the process by plotting these three rat,ios
upon a chart and graphically interpolating, so as to tind the
best ratio for Providence, which when applied to t,he Provi-
dence average would have given almost the same result.
........................................... .......................................... ..........................................
1B7-45 41.5 1847-66. 45.6
185746. 45.5
la77-86 .......................................... I?.? 50.8 4b.8 44.6 lW-'% ........................................... 1 4 5 .4 I 5 1 .1 1 4 7 .1 I 49.6
1867-76 ........................................... 45.6 49.4 S<.6 kq.4
Average of six decades ..................... Average of flve decades ................... RaMo of fourth decade to average of fire 44.01
decndw ......................... ......... ..........
The mean, 49.92, aa vomputed fm the assmned inissing
'decade at Providence, agrees with the actual observations dur-
ing that decade, 49.4, within a half inch, and if it had been
used instead of the observed figure would have given a mean
for sixty years of 46.54 insteAd of the.46.45 that resulted from
actual observations. By this process we reduce all short
records up to the fundamental length of period and t h u
obtain all the actid rainfall measurenienta that are available
for our chart.
I n reducing the average of a short series at a minor station
up to the average of the longer fundamental period for which
the chart is constructed, we tacitly aseunie that the aveixge
rainfall during the missing years at the minor stations has
been proportional to the rainfall during those same years at
the fundamental stations. The above illurti.ation shows how
nearly this is true for Providence, R. I., as compared with
the other three stations in.its neighborhood. Stations that are
farther removed or have intervening obstacles would not
show such favorable results. A general consideration of the
similarity of records at more distant stations may be seen
from the exhaustive discussion by Hann (Vienna Sitzungsb.,
1909) of the rainfall from 1735 to 1900 at Padua (4V 95' N.,
11" 55' E.), from 1764 to 1900 at Milan (4 5 O 85' K., 9" 10' E.),
and from 1813 to 1900 at Klagenfurt (46O 40' N., 14'; 80' E.).
Klagenfurt is 145 miles northeast of Padua, and Padua is 140
miles east, of Milan. Hann show? that the absolute variability
of the a,vei.age rainfall for periods of 5, 10, PO, 3(:), or 40 years,
expressed in percentages of the annual rainfall, agrees very
closely at the three stations. Hann also shows that the rsimul-
hneous departures of the nienns of 30 years or 40 years
from the nieans for 100 years, pursue parallel courses at Padna
and Klngenfurt, but an opposite course at. Milan, the estreme
departure beiug S per cent negative and 6 per cent positive;
in other words, the mean of any group of 30 or 4.1 years
is still liable to ditfer either way from the mean for a century
I q 6 or S per cent of the average rainfall. Incidentally it
may be mentioned t,hat in these three long series of rainfall
iiieasurenients Httiin finds a clear demonstration of the exist-
ence of the 35-year period, announced by Briickner, but no
clear evidence of a dependence of the ixinfall upon the fre-
quency of the sun spots. Thwe is a very decided tendency
for dry or wet years to occnr in 'groups of two or three;
groups of four such gears occur twice as often for dry gears
~s for wet years. In the same way groups of two or three dry
~r wet months frequently oocur, and groups of four or tive
wet months are more frequent than such groiips of dry months.
The idea of a "fundamental interval of time'' to which all
data for shorter intei7rd.r should he reduced as accurately as
possible by a process of chronological and geographical inter-
polation is one that applies equally to charta of isohars, ioo-
therms, and a11 other meteorological data. I11 every case the
charts of data must represent the 8ame unit of time. Iao-
metric lines for various intervals, such as 30 years and 5 years,
must not be mixed up together. If we make this mistake, then
numerous details will appear on the chart that have been in-
troduced by special irregularities of short groups of years
RINZ which should disappear in a chart that truly represents
the whole fundamental interval.
(9) I n drawing isobars and isotherms, it has always been
recognized that our stations are not sufficiently numerous to
APBIL, 190% MONTHLY WEATHER XEVIEW. 217
enable us to present all the details of pressure and tenipeia-
ture near the surface of the ground, nor do meteorologist8
need such detail. If it is desired to use uieteorological
data to represent only the general dist,ributioii of pressurc
and temperature in the atmosphere, we accomplish this besl
when the observed data are reduced to one uniform plane
above or helow the network of stations; hence t,he isotherm$
and isobars belong specifically to that plane. If one know*
his elevation relative to that plane, he may then use the
same scale of reduction and ascertain his own local pres-
sure or temperature.
. A similar consideration pertains to the rainfall. The nieas-
ured rainfall represents the fall that would be caught at any
point in the vertical or the slanting columns of falling rain
between the station and the cloud; therefore the rainfiill
chart9 already belong, without further reduction, t,o a hori-
zontal plane that niay be located anywhere below the min
clouds. The irregularities of the distribution of rainfall 011
tbio plane are of two kinds-(a) local, depending on the mind
effect at the mouth of the gage, or on the bad location of t,hc
gage to leeward or to windward of trees, houses, walls, and
cliffs, none of which affect the original rainfall itself. and all
of which must be avoided or allowed for by the observer:
(h) geneid, such the influence of t,he att.empt of the wind
to blow over a mountain; the passage of the wind from the
ocean to t.he land, or vice versa; the distribut,ion of rain in B
river valley or around ti storm center, or to windward of B
mountain range. It is to these general influences that the Tain
itself is originally due, and in si) far as the i-aindrops fall in
parallel lines, either stirtight or curved, the rainfall nieasiire-
ment will be the same, no matter at what elevation the irtiii
gage may be placed above the sea or below the rain clouds.
10. The amount of detail that we shall be able to present
and the scale of our chart depends on the number of'station*
available per square iiiile and on the iiiore or less ooiiiples rela-
tions between t,he orography, the winds, and the oceans. In the
case of less than 3,OW stations available for the United States,
or ail areiage of 1 per 1,000 square milea, or an average (lis-
tance of 35 miles apart, we can only present the most general
features, and any atteiiipt at miniit.iW is bot,h premature and
misleading.
Over large areas of land there mill genemlly occur regions
for which me have no actid rainfall observations. and for
these sonie will desire to estimate the rili&ll, relying upon
analogy, extixpolation, or other sources of inforniation ; but
the meteorologist will generally decide that in the absence of
actual records it is his duty to leave these spaces blank on
the iainfall map, and let others make estimates atppropriat,e to
their own needs and knowledge.
I n const,ructing meteorological chart.$ it is coninion to di-aw
full lines where inforination is supposed to be reliable, hut
dotted ones where it i.r deridedly hypothetkd. This is not
pm.tiable in rainfall charts if we adopt shading instead of
isohyetals. I n such cases, also, it is best to leave those areas
blank for which we have only unsatisfactory or indifferent
dah.
11. I n order to eke out the +can$ fragments of lainfall
data, it will be necessary to dmw the isohyetals or to shade
.
the rainfall char& by calling to our aid whatever is known
with regard to the general laws of rainfall. Precipitation
depends upon certain atmospheric peculiarities, namely, the
direction, velocity. temperature and humidity of the wind, and
the evaporation from the ground or the ocean. Precipitation
also depends upon the following peculiarities of the location,
viz, t,he height and slope of the hills, the trend of the lines
of the mountain ranges and the nearness of the mountains to
the ocean, and the intrusion of cold northerly winds.
The induence of the elevation of the kind is theoretirally
espbined by Pockels (see Monthly Weather Review, April,
1901), but ordinarily it must be est,iniated enipirirally. Thus
S. A. Hill (Met. Zeit. 1Si9. XIV, p. 161) shows that in the
iiorthwestHimal~ytyas, as we rise above the plains at their base,
we have a distribution of relative rainfall, as shown in the
following tiible:
I
This study shows that in this region the maxiniuui rainfall
occurs at an elevation of about, 4,160 feet., and, in general, Hill
dates that t,he rtiiniest stations in India all lie &.an altitude of
nboiit 4,'WO feet, including even Cherrapunjee, on the Khasia
Hills. and the stations on the (f h a t .~ He also adds that it is
well known that the niaxininni zone of iainfall in the Ghats
is lower down on the windward or ocean side than it is on the
leeward or land side.
All our knowledge of the roliition between rninfall and alti-
tude shows that when we have determined the altitude of t,he
zone of niaxiniaui rainfall we niay blien utilize that knowledge
in extending our isohyetals from knowii stations a 1it.tle way
tip or down into the unknown country. But me can not do
this .safely unless we hnve ascertained the altit,ude of the
niaviniiim zone for any given locality.
18. If forest areas actmually contribute to increase the quan-
tity of ixin, as nlaintainecl by sonie, we Mhould niake our
isohgetah iwcominodat.e themselves to the forests; but we cau
iiot :wcept, this theory. At first thought it may seein plausible
t,hat the esistence of a forest or a. grass-covered plain may be
taken as ~i sure indi&ion that sonie definite quantity of rain
tnnually falls in that region. However, we are pkented
from utilizing this idea hg our knowledge of the fact that the
growth of plants is due quite as much to favorable soils,
Favorable temperatures, and other consideiations :w it is to
rain. so that it is not safe to use bhe plants as a definite index
;o the quantity of rain.
13. I n geneirtl a forest cover retttins on an average 95 per
:ent of the total aiiniial precipitation, especially, of course,
;he snow. Four-fifths of thk, or YO per cent, is evaporated
and never reaches the ground; the reiiiaining 5 per cent does
reach it by dripping or running down the larger hmnches
tnd trunks. The evaporation from the forest soil is about
me-fourth of that from siniihr soil in open fields. The vol-
218 MONTHLY WEATHER REVIEW. APRIL, 1902
the precipitation caused by tlie presence 07 high land experi- ences at first an increase with increasin akitude, but after- wards a decrease for higher altitudes. &ere is, therefore, an I altitude of inaxiniiim precipitation, and at greater altitiides
--
ume of water retained in the soil is larger in the forest thar
i n the open land et elevations up to 500 meters, but it ir
Bmaller at elevations above 800 or 900 meters. The growtk
of forest trees consumes only a small amount of water 81
compared with that consumed by crops on cultivated soil.
The permeability of the soil to water is increased hy the pene.
tmtion of the rootsof the trees,so that by ahsorbing water deeF
into its soil the forest has the same general influence on run-
off aa is produced by a reduction of the general slope of t.he
ground. Owing to the quantity of soil nioistiire and subter-
ranean water within a forest? tlie springs arid streanis are
larger and more constant. By virtually diniinisliing the
slope, forests teiid to repress floods and distribute water more
evenly.
14. Again, so many nierwurenients have been ni:ide of thc
flow of wat,er in rivers and so iiiany comparisons with t,lie
observed ixinfall that one niight hope to argue from this flow
or run off back to the original i-ainfall over the watershed.
It is found that, in general, the flow in the rivers is from one-
half to t,hree-fourths of the rainfall for those c.asw in which
both quan tities have been satisfactorily measurecl, Init t,he
mtio depends on soil, soil covering, and slope, and it does not
appear that it can he known heforehand with any great accu-
racy. Each value of the ratio of bhe rainfall to rim off ~J W -
tains to special regions ancl special circumstances, so that. it
is not allowable to apply it to other regions or to a general
calculation of the average rainfall froni the known flow of
water into the river.
In conclusion, then, we believe that the chart of rainfall
must be made to depend wholly upon rainfall iiieasureiiient,s
theniselves, and only when this is properly done can me
use the chart to study out tahe relat.ion between rainfdl and
the various matters that interest agriculturists ancl enginecrs.
It will not do to use imperfect rainfall measurenient,s to clrter-
mine the quantity needed for plant growth or for river flow
and then argue back in a circle from the foregts or run-otf to
the ixinfall.
The following extensive series of selections froni p i blica-
tions and correspondence will serve to substantiate the above
concliisions, and to show t.he details of iiiodern practice as to
the construction of rainfall charts.
.
(5) PROF. J. HANN.
I n his ’‘ Lehrbuch der Meteorologie,” 1901, pp. 350-353,
Prof. Dr. Julius Hann has the following i-emarks relative to
the increme of rainfall with altitude above sea level:
The cause of the increase of preci itation dnrin r the winter season at ordinerv altitudes is to Y >e sought in t B e fact that
150 cni. in the summer. When a low-lying basin is surrounded by Mittel ebirge, or at least on the &de from which the prevailing i-ainy winds
ascending mass of air, on its relative humidity, and the tem- erature at which condensation begins. In the winter time
Li e relatire humidity and low temperature act together to
time. with drier air and higher tempeixture, the zone is pushed up to a higher altitude. Observations show that the summit3 of the Gema.n “Mit- telgebirge” belong in geneid i n the winter time to the maui- mumzone of precipitation, but in summer titnethe zone extends far above them. Unfortunately observations in the Alps do not yet suffice to establish the altitude of the niasimiim zone in wniiiie,r. bnt Erk has shown that for the winter se:tson the ina~i~iiitii zone of precipitation on the north side of the Bava- rian .- A l p is frequently located at 600-1,OiNI meters ahove sea
re f we the altitude of the maximum zone; in the summer
level.
I-Iellmmn, in a comprehensive comparison of the monthly r~iitifall~,expresPe.d in percen t,ages of the annual sum for theGer-
tiim Mitt,elgebirge. has en jedally described the occnrrenve of the prevailing winter rain 1 all, ancl has shown that whereaw in
the ”Siicleten?’ at altitudes of more than 900 meters summer rains pre\.nil, yet in the ‘. schiefengebirge” of the Rhcnish provinces and in the Vosges, even at aMtudes of 300 or 400
iiiettw, iiiovt of the precipitation belongs to tlie colclcr half of
blie year. First, in central E~rope the winter recipitation is in general heavier as we go
the northwest m d , second, the annual distribution of rain is niore uniform in that snnie direction, hence, also, the excess of the auninier rain diniinishes in the lowlands. Therefore
we see that the level at which the inversion froni prevailing suiiinier iains to prevailing winter rains takes place, is lower
as we pass froni south to ti.orth and from east to west.
In fact, Supan deduces R second higher level of inversion from the results of rainfall measurenieuts in the Bel inn Ar-
l’he explanation of this is easily seen.
from south to nort P i or froni eatit to west,, or in genera1 toward
dennea, as compiled by Lancitster. I n this re ion t. fl e lower zone of prevailing suiiiiner rains in the lowlan a s extends up to
nn d titude of about 350 meters, which is the lower level of in-
version, and here begitis the middle region of prevailing win-
ter rains on the plateau lands. Hiit this region extends only
up to about 500 meters, where again t.he prewihg suiiiuier
I-sins begin, so t,liat t,liis latter altitude niay be considered as a
act-otic1 level of inversion. In fact, on the Pic rlu Midi there would seeiu to be three levels of inversion, as shown by the following result& of olaervation:
Prrcipitatim.
1 Altitude. ~-- Winter. Summer. Anniial.
I
I .m/rrs. c‘m. PttL I C‘eL
36 w
161
‘I
Locality.
TnrFN. ..________.__.___________..____.....______ I 3W B;igri~cws.. . . . . . . . -. . . . . . . . . . . . . . . . . . . . . . . . . . . . . _ j 555
Plunlarle.. . . . -. . . . . . . -. . . . . . . . . . . . . . . . . -. . . . . . . . . Pic dii lIic1i.. . . . . -. . . . . . - -. . . . . . . . . .. . . . . . . . ._. . . i 2,m
1intnedi:~telg above Bagnieres the minter rains begin; Plant- %de has siiiiinier riiins; the suiiiiuit again hns prevailing winter rains. The niaxiniuni zone of the winter rains-as shown by
tt graphic presentation-appears to lie at 1.300 metew in the winter season and 1,900 meters in the summer. Jn this1nasi-
niuiii zone there falls 103 cm. in the winter season and dniost
219
Place.
Plains of Snxonr.. .................................................... Mountains of Skxonv ................................................. Central rnd rw~urheriI Bohemia ................. .r. ................... Great Hungnrinn Plain ............................................... Trnnsylvania (Sieben biirgen ........................................
valleys, to which must be added also the local showers anc thundei-storms that are missing in the winter season. There fore, i n such regions, analogous to depressions, the sumniei rains exceed those .at the same elevation in the country out side. An example is shown in the following tahle:
Percentnge of rrin- fall.
Winter. ' *iimml*r
18. B I 3
1.5 ! 4
~-
~...I,
21 3
9 . :1: 15 41
I ..
0 0 l 3 1 .
d o -w ... ........................ 30
cSU-IO ........................... ! 61
............................. '
9-10 ........................... : 102
N. p 7 0 ........................... 38
w-50 ........................... I 5Y
wa. .......................... I 73
lo- u ........................... I 51"
The altitude of the zone of niaxinium precipitation is found
to be 1,3V(i nieters in the northwestern Himalayas; 1,4i1c
meters in the Ghats, and about 1,000 meters in .Java. In thc English lake district the inaximuiii rainfall is at 550 meters,
With reference to the distribution of ininfall over the eart.h'i
surface, Hann has the following reniarks on pages 351-:3CiO:
Strictly speriking, we know soniething alJout the clistrihution of tho quantity of rainfall on the land only, because on t.lic ocean the marine observat.ions record only t.he frequency 01
precipitation. Supan (Geog. Mitth., Heft VIII. 189s) has endeavored to till 11 3 this gap in our knowledge, at least foi
assuiuptions as to the relative depth of rainfall. Bot even on the continenB there are hrond regions i n whicli there is not a single rain gauge and where, therefore, in place of nieasurenients we must, substitute inore or less rat,ional estimates. No iiieteorological element is so dependent as is
the quantity of rainfall on local conditions for its occurrence. or so often shows unespected differences n t neighboring locrdi- ties. It is therefore easily understood that every effort to represent graphically the distribution of rainfdl over the earth's surface (as has been done successfully for pre qwre, .... temperature, and even cloudiness) must stunible upon the greatest difficulties and uncertain ties. A reduction of local rainfall to sea level is quite inipossihle. since i k variatmion~
with alt,itucle above t,he sea follow no general rule. It there- fore requires a certain boldness t.o puhlish a rainfall chart of the whole glohe, showing lines of equal quantity of rainfall, as WRS timt done by Elias Looinis first in the Anierican JOLW- nal of Science in 1889, and afterwards an improved edit,ion in
1889 in his Contributions to Meteorology. But the need of a perspicuous resentation of the distribu- tion of such an important nieteoro F ogical eletuent as the rainfall over the whole earth's surface is so pressing that Alexander
Supan decided to prepare a new rainfall chart, of the globe on the basis of the greatly increased ni~ss of ohservat.ions. hut
confining himself to a presentation of only six grnclations of
mi ti fa1 1 :
Millinirters snniially.
Slight rainfall ... ___ ___ . ______ . .:.. .... ___ ... ___ __ .. __ ...... 0 to 360 Mm1erat.e rainfilll - - - - - - - - - - - - - - - - - - - - - 250-500 Moderate rainfall - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 500-750
Moderate rainfall - - - - - -. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - i.50-1000
Abunilant. rainfall - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 1000-2OOO Abundant. rainfall - - - - - - - - - - - - - - - - - - - - - - - - - - -. - - - - - - - - - - - - - over 2000
By this means the arbitrary features are limited to defining the houndary of each irtinfall region? and the map attains more scientific precision and coni~)reheneivetiess. The new rainfall chart by Supan, which is re1)uhlishecl i n thi? present Lehrbuch,n will foiiii the basis of our few remarks on the general distribution of the quant,ity of rain.
the Atlantic and I n h. inn oceans, by nieans of some rezisotialile
II
~~ ~ ~~~~ ~~
0 0
N. 0-10 .......................... 10-28 .......................... 2o-N .......................... 3 M U .......................... , 75 5 9 4 0 -5 0 .......................... ; 113 50-60 .......................... 11" 60-70 .......................... 107
ogee also Supan, Pet. Geog., Mitth. Ergiinzungsheft 134, 1895, and AuguRt Heft, 1898: Buchan and Herlwrtaon bar tho lo mew'^ Phywical Atlas, Meteorology, 1899: A. J. Herbertson, The distribution of rainfall over
the land, London, 1891.
The iriost general features of rainfall distribution on the earth's surface are conditioned by the general circulation of the atmosphere.
In the tropical belts, where the ascending motion of the air is the most active and takes place on the largest scale, and where the air is also richest in aqueous vapor, in consequence of the high tempefature and the great extent of the warni
oc~ean, the average quantity of precipitation is also greatest. On the other hand. at the boundaries of the Tropics and in subtropical latitudes, where the air that has risen in the interior of the tropical zones has a ain sunk to the earth's surface, there falls, on the average, t % e least rain of all; indeed, there occur here large regions where regular precipitation is ent,i.rel~ want,ing. The great steppes and the arid belh in hoth hemi- spheres belong pritwipallp to these latitudes. I n the next higher latitudes, as has already been explained, the nunierous large and small atmospheric whirls cause more or less ahun- clan t, precipitation, a,nd therefore the annut~l quantitv of pre- cipihtion increases, but only afterwards to diminish in still higher latitudes in t,he neighborhood of the circum olar re- gions, in consequence of the low teniprature and t t: e slight capacit,y of the air for aqueous vapor. In the polar regions t,heiiiselres the quantity of precipitation is very sni~ill, hecnusc the air has too little moisture. especially in winber. From these points of view we easily understand t,he zonal clist,ribution of the rainfall.
John Murray has attempted to estimate the averiige qunn - tit,. of rain that falls on each zone of latituclo 011 the basis of Looinis's rainfall chart. The results to which he has attained
are, of course, only rough approximations, and relate only to
the continents. They are as follows:
MEAN RAINFALL ON ALL CONTINEWTS RY ZONES OF LATITUDE.
Preriyi- I Preripi- tation. Latitude wne. I ,I Latitude zone.
The lkrgest quantity of lain falls in the equatorial regions hetween 10': n0rt.h and 10'' south latitude. The rainfall dimin- ishes toward the sub-tropical latitudes, where it reaches a iiiininium; beyond t.his it again increases. The sonthern hemisphere beyond latitude 3iF has a larger rainfall than the iiort,hern, bemuse it has no great dry region of an int,erior :!oat.inental surface which in the northern heinisphere attains
% great est,ent, in the .neighborhood of the fiftieth degree of Iati tude.
f)*J.wtio)i (sf rtrir~f~r77.--Originally all the aqueous va or con- bdned in the ntniosphere conies froni the ocean, whic E covers two-thirds of theearth's surface, and proportionally even more than this i n the warin zones. The ever niovin atniosphei.e .wries the aqueous vapor to the very center of the f argcst of the ~ont.inents. The diffusion of aqueous vapor, although it goes
111 slowly, also contiil>utes i h part to a irtther uniform clistri- bution of vapor, at least so far as teniperht,ure per~iiit~s. Since ;he vapor condenses over the continents aiid falls as ixin or mow? therefore the moistened surfnce of the earth and the ivater that accumnlatcs in the depressions become a secondary source of atmospheric nioisture and preci itation. Large
sustained by lainfall, give back niuch aqueous vapor to the itniosphere and favor the formation of precipitation. But
tvery chart of rainfall shows how relatively slight is the
&2 1 #l ~d ~p(Alll(12si8 o)i tfir 7tjratr7 c(lli8r8 t2fCmtt~lit7tio~1P ill the ai8-
.akes, and especiallg regions covered with B ense vegetation
990 MONTRLY WEAT~ER REVIEW. APBIL, 1903;
influence of the lake upon the local increme of preci itation.
wedern biberia come froni the aqueous vapor t.hat is carried landwards from the Atlantic Ocean and the North Sea l y the prevailing west and northwest winds, where it is condensed in the wandering cyclones as general land rains in the cyclonea that enter these coiintries from Europe, or as thiinderstonn rains in the loail ascending currents of air. The moistened soil now spin gives u I aqueous vapor. and thus the sanie
over into the vei-ticnl circulation and appear again and again as precipitation. But since in the winter season and civer t,he continents of the hi her latitudes the low tenipertitniv reduces
therefore again the principal source of the summer ixins can
only be the aqueous vapor hrought into the interior from t.he ocean. Even the small quantity of snow that falls in winter is brought hither froiii the ocean by the cyclones. Again, after months of drought t.he tropical rains t,hat occur in the interior of the continent can be fed only hq' oceanic aqueous vapor. We see what cpnt,it.ies of moisture the weit breezes bring far into the interior of t,he coiint.ry by consid- ering the general land inins of central Europe, which often last niany da-ys, with northwest winds, and give rise to great floods-where, moreover, the saturated air at relatively low teni eratures re resses the lociil evapomtion." &me we finx that in general the quantity of wecipit,ation
of the continent,*. In fact. the great continents, espechlly where the iiiountainn interfere with the rain-ljearing winds from the ocean, are deficient in i-ainhll, even t,o the rharact,er of a desert, IS in the interior of Asia and the interior of North America. Where innges of monnt.ains along the coast form
a wall to. prevent the penetration of damp oceanic air, t,here the dry or ininless region uiay occur quite near the ocean. as?
for esample, in Australia, whose eastern coast has heavy pre-
cipitation, whereas t,he plains on the lee side of the coast range are quite dry. It is pai.ticiilarly to be remarked that in the winter season, nnd espechlly in t,he cold winters of high latibudes. mountain chains form R muoh more effective sc:reen against, the t,rans- portation of aqueous vapor from the ocean than they do in
suniiiier. The land on the lee side of a niountain range is therefore relatii-ely and also absolute1 7 drier in winter than
which the ocean winds can in summer time pass over the mountains in a saturated condition, t,hey bring to trhe land beyond the iiioutitains a reater or less quantity of aqueous
hetivier sumnier rains to fall there. Therefore, whereas in
siiniiiier time the rainfall diiiiinishes on the windward coast hemuse the level of the plane of condensation tiow lies higher than inwinter, the rainfall increases on the lee side, and there- fore the difference hetween the rainfalls on bhe two sides is
iuore or less diniinighed. . After these preliminary reniarks we mii consider the gen- em1 features of the distribution of the quantity of precipitn- tion over the continents. We reinark firnt that in the Tropics the east #ides of the continenks and the islands are in general iichext in rain. I t is the aqueous vapor that, is brought directly from the ocean
a Supan and Briickner have lately ,ointed out the ordinary unrlerewti- mate of local evaporation, and have skown that the eraporation from t.he su$ace of t.he land lays a very iniportant part in the origin of siininier rains. * * * On(, one-fourth of the precipitation on the surface of the land flows back to the ocean in t,he rivers. Therefore a large part of the precipitation must be derived from the evaporation from t,he land. That
1s to say, we measure the same qua.nt.ity of water t.hat came ori 'nally froiii the ocean inany times over in our rain gau a; it is often confenwd and immediate1 evaporates again. * * * Wren, therefore, we ascribe two-thirds of t i e precipitation to the aqueous vapor IJrought fmin t,he ocean t.his will appear to be a low estimate.
The heavy summer rains of the interior of Rirsia an Tp even of
vapor originally derive d from the ocean can enter many times
the water containe d in the [local] atiiiosphere to a niininiuni.
diminishes as we proceed from the coasts towar C I the interior
in siiiiinier. On account of the hig h er tenipeinture wit,h
vapor, depending on the a f titiide of the iiiountains, and allow
~____.
~
. .
by the prevailin ea& winds or the trades that is so abun-
rises rapidly as we go inward. The a undant rainfall is pro- longed into the temperate zones from these tropical regions along the ea.3t coast of Asia and North Amelia, and extends beyond the foi-tieth degree of north lat,itude, because in that region in the suninier time the ocean wind blows with a more
or less nionsoon chairtcter. The same occurs on the east coasts of Australia, Youth Africa, and South America, but. on the other hand, under the same latitudes the west coasts are much poorer in rainfall. - But in the higher latitudes, where on the polar side of 40°
the most winds begin to prevail, the comparative relations are entirely reversed. The most reinarkable and tvpicnl illustrat.ion is shown in South Anieiica. I n both bemie- pheres beyond 44" of latitude the west coasts are ver rich
in rain, ns is shown by the rainfall chart of northwest dirope and America, South America. and New Zertland. Indeed, such great quantities of rain fall on these coasts as in soiiie
places to equal the rainfall'of the tropical zone.
In general the heaviest rain falls wherever steady winds coniin from a warm ocean encounter elevated' lands. I n t. f e iiiiddle and higher latitudes on the continents, rains also fall abundantly over a flat countiv, where the great at- mospheric whirlwinds are frequent and, therefore. along the so-called paths of centers of low pressure, especially those t,hat are most frequented. Esamples of .this are found in North Anierica.. in the region of the Great, Lakes. in Denmark and southern Sweden, also in the Hungarian plains, which latter, in spite of their continentd position and their being surrounded by nionntainti, receive on an avera e more rain than the low
to the Atlantic Ocean. In fact, over &ungary there passes a
storm path that is much frequented hy the atiiiospheric whirls that are t.raveling from the Mediterranean and the Adriatic t.oward Poland. On the other hand, Momvia and Boheniia are rather far removed to one side from the paths of the At- lantic whirls that are passing over England and Denmark into the Nort,h Sea. But, everywhere the cluantity of precipitat,ion increwes on the slopes of thg mountains on account of t,he ascending motion of the air that more frequently occurs there and the ronse uent cooling that leads to the condensation of aqueous vapor. l%-ery special rainfall chart of any country has, therefore, a great similarity to the hypsometric chart itself. * * * It is certainly to be remembered that the ext.reme rainfltlls are almost, invariably confined to a small region and are low1 phenomena. That region of t8he globe that on the aveivlge receives the greatest qiinntitv of inin is certainly the archi- pela o of the East Indies and the northern coast of Australia
On the other hand, over broild regions of the $lobe, the an- niial rainfall is either entirely wanting or confined to a few centmimeters. Probably there is no region absolutely ivlinless in which no rain falls in the course of many years. Even in
the Sahma and on the west coast of Peru, rain occasionally falls in the course of a year, and often it is very heavy. The
Polar regions belong to t.he driest port.ions of the earth in ahso- lute measure, since often there are only 10 or 80 centiiueters of precipitation in the course of a year. But in spite of this,
there is no want of moisture because of the low temperature md the frozen soil. The same quantity of precipitation has
~i i estrnorclinarily different import according to the local climate.
\
dantly condense d on these shores, es ecially when the shore
plains of Moivzvin and Bohemia, alt a ou h these lie much nearer
and a ew Guinea.
(6 ) PROF. A. SUPAN.
I n lS98 Prof. Dr. Alesander Supan, t,he editor of Peter-
tii~nii's Mittheilungen, published an exhaustive memoir on the
distribution of precipitation on the islands and continents of
_. Amm, 1903. MONTHLY WEATHER REVIEW. as1
the globe. We suhmit the following translat.ion of some por-
tions of this memoir, bearing directly upon the present ques-
tion:
It is well known to me that a rainfall chart of the globe will meet wit,h the apl)roliation of geographers rather than meteorologists. Even Julius Ham, whom we geographers with pride refer to as the chief of c!liniatologists, allows that the chart of the qnantit,y of rainfall can have only n didactic and not n scientific. mlue. It is maintained t,hat the measure- menh are not yet sutliciently nuiner~~is to cxclude t.ho arbi- ti-ary features introdaced by the draftsman. Even to-day t.his reproach is jnstitied up to a cert.:iin degree. Although we now possess avei-age values for :$hotit 4,MN inirtfall sta-
tions, yet, unfort,unntely, t.liese tire distributed vcry unequally. More than half of thein helong to E L W O ~~. andeven here there are broad regions. such as northern, Russia and T~irliej~, that
have only sporadic series of ohscrvatioiis. But if we allow that the European shitions are a sant1)le of what, we ought to have, then for d l the lsncl :L~c?? of the earth we should need over 20,000 st,atioiis. or fire t.imes t w iiixny ax :ire now at our disposal. Arid even wit,h these we should not he :ihh to do much niore t.han t,o const,rnct a sc*heniatic~ presentation.
In no other of the climatologic.al elements do local influ- ences play so inip(.wt,:int a 16le xs in precipitation. The niore varied is the surface of the land and so niricli the inore rapidlj the relatire altitudes change. therefore so much t,he densc!r
must be t,he nctwork of st,ations. Two mints coine especially
the prevailin rain winds and its ahsolutr! altitutle. We know
and that the precipitation increases up to a certain altitude and
then again dnninishes, and i n such :t way bhat the altitude of the zone of niasiiiiuni precipitation decreases the newer we approach the polw. And now consider the complicated variation of t.hese condi- tions in a mountain range like that of the Alps or in a mountainous count.ry having the complex stroctiire of centml Germany. The distribution of rain in a relatively simple portion of the latter. namely, i n Silesia. has lately been gi~phic- ally studied by J. I’arbsch. This essay is of the highest interest. as a study of methods. Although 5% nbution records were at his disposal (for fire years only hit simultaneous), still this scarcely suficed to bring out the influence of the orogirtphy on a rainfull chart, on the scale of lil,WO,WO, and for sotile mountainous regions even this scale is too large. More than ‘i3O,O(JO stations would be necessary to enalde lis to draw a minfall chart for the whole land port.ion of the glohe with the accuracy with which Yartsch has done it for Silesia. We may. therefore. judge of what may be done with 4.000 stn.tions only. Consider further how unequal the inaterial is. Outside of the civilized countries we shonld IN very
happy if we had five or ten year arerages atid not infrequently are we reduced to single gears of observations. whose utiliza- tion of course requires the greatest circunispection. But even longer series of observations are with difficulty compara-
hle with each other when they helong to diflerent, epochs. Furthermore. we know the tlitticulties of t.he measurement of snow, especially of drifting snow, which fills iis wit,h some
distrust of the mean precipitation for the higher latitudes. Even the establishment and the constrtiotion of t,he instru- iuents are of importance, hut in only the rarest cases can these be controlled; from this point of view Hellmatin has discovered many errors in the rainfall charts of Germany. I n the tables
uublished bv Partsch for Yileriia. we find Oderherg with
iuto consideiation-t,he location of the p \ :ice wit,h referencr to
that the wit1 d ward side has more rain than the leeward side
other neighboring stations at once shows that the figure given for Oderberg is erroneorw; but what a confusion would pre- vail if we had no control at hand? It is in fact an unattractive problem to erect a bidding with such niaberial. But, still the problem must be solved. The chartographic method is the vital principle of coniparative climatology. Humboldt’s chart of isotherms tirst established this science, and yet how insuffi- cient was it3 basis-only 57 stations. If Buchan had been too cautious we should to-&iy still have had 110 isobaric: charts of the globe. The c:onstriiction of aminfall chart has, however. to contend with etill greater difficulties because the distribution of precipitation depends so niuch on local circumstances that oft,t,imes these coin letely oblit.erattte the intliience of the prin-
the locd character f roiu the measurements, and especially we
have no method of reducing total rainfall to sea level. The consequence of bhis is that we need a much more comprehen- rive collection of observations than for the gra hit: presentatmion
we keep this restriction in mind we can not see why such a rain- fall chart, shoiild be of less valrie than the ahtirts of isothertiis and isobars. Il’here gu s occur it. is the place of the awom-
an advantage of the chartographic met.hoc1 that can not be overestimated that, even in cloul~t~ful cases, it forces us to recog- nize shades of special intpurta.nc!e in the meaning of the m70rd.9,
whereas i n hb~lxtr presentations the defects of the network of observat,ions rem:tin uniinportnnt and the presentation
ran skillfully conceal these so that we either do not notice them or can not t~ring the author to task for having overlooked them. * * * It has been mentioned above that we can bring against rain- fall charts the objection t,hat they leave too much play for ar1Jitrw-y features introduced hy the draftsman. If we ac- knowledge that this danger exists, still it is diminished so
much the more in proportion as we enter less into the details. I consider that it is not allowal-,le to introduce more than six grades at, the resent time for all land portions of the globe except for &rope, the United States, and India, and the greater part of the islands. If these glades increase i n extent as we go upward (namely 880 mni. for t,he lowest stage
and 1,000 nim. for the upper stage), t.his has a twofold reason, the tirst because in the lower grades differences of 950 ~iini.
are practically much more important than in the higher stages, and second becnnse very consicleiable depths of rainfall are,
so far as our experience goes, not s read over large re ions.
selves to four grades, iartly bemuse of .the srualler scale of
season varies from place to place more than the quantity of annual rainfall.
cipal factors. An x we hwe 110 means by which to eliminate
uf the distribution of other meteorological e P ements. But if
panying text to justify t E e chartographic presentation. It is
In our chart of total seasonal irtinfa P 1 we have restricte f our-
the charts and partly a I so because the quantity of rain in any
* * * * * * *
Having shown that the winter rainfalls, in general, cover
broad areas of country, while the summer minfallu ‘aro local
showers, often acconipxnied by thunder, Supan gives, on
pages 4043, a study of the relation between altitude in the
interior of the continent or islands and the increase of rainfall,
especially winter ixins, front which we make the following
extract:
I n the region of prevailing winter rains, wherever we have
a sufficient amount of observational material, we find a rela- tive increase of t.he precipitation with the altitude in the colder half of the vear. For instance. in Scotlalid. on the coast. kill mni. and the neighl~oring Annaberg with 737 mn:; both
29--D
228 MONTHLY WEATHER REVIEW. APRIL, 1902
attains 60 per cent of the total annual; while the whole of the Met. Zeit., 1S87. p. $4.) I arrange these in three groups, as
western hinhlands has 61 or 68 Der cent. The same Dhenome- I follows:
-- -
from the preceding. The above tnble gives us t,he nlenns for the altitude of the reversing level for diiferent :z:rt$fe winter rains at an altitude of 100 meters ~~i d
non is rep&t.ed in the region ofprevailin summer h i s . In I
vidual zones of altitude in the ICingdom of Saxony is very
~
instructive, and the following values are deduced t.here,ft.oni: ;
~
Altitudeof Altitudeof the lowest the,loweat
St4LtlOll that has pmvail- Lora1itp. station that Locality. hss prevail- ing winter ing winter
this reepect the table compiled by Paul Sc % reiber for the indi- I
I rains. I Rainfall.
POIITHERN SERIm.
JTdfrn. .............. ..................... 354
................. 490
3%5 ..................... ................... .......................... NORTHERN SERIES.
Harz.. ......................... 1
570
............. 4w 630 SRiierlan ll..... ................ ........................... Westrhrininrhwl Schlcferge-
.......... birge ........................ ........................
___
CENTRAL SERIES.
Altitude.
Mdcrs. wni. nim. 100
__________-__ - --
1 Alti- 1 Rainfall. 1 y;;ffly0f
I tllnr. I
.
;
........
---
The winber i.ains of Ais-la-C!ha elle are also nnd6ubtedly controlled by the location, since t I: e altitude of the plane of
reversal is considerttbly higher in the Belgian Ardennes. I have computed the followin inenn values from the esccllent collection of Helgian rainfal 7 stations by Imicaster:
- -----
4% ....... ....... 43 ....... ..... 44 ....... ..... 45 ....... ....... ....... ....... ......... I attach no importance to the sbsolut8e values of these fig-
rate of increme of the precipitation with altitude is constanti but this much at least we may conclude. viz. that the level of
reversal is 1owe.r in proportion 8s the. recipitation in the low-
ures, for thej rest upon the unproven assumption that the
lands is iinifornilv distributed throue E out t,he seasons. or. in
other words, in l;roportion as the a t h a 1 ininfall variation is
smaller. When we consider that in (3erniany the annual vtwitition diminishes HB we pas^ froin the sonth toward the north and from the east toward the west, we must expect that the alti- tude of the plane of reversal mill diniinish in the same direc- tions, and, aonsequently, that in the northe,rii and western portions the land surface of any given altitude may have win- ter rains, while i n the south and esst, the snine altitude partic-
i This. in fact, is denionstrnted in
t r e collection of high stations that we owe to Hellniann. (See tes in the suninier iains.
~~~~ ~~ ~ ~ ~~~
another res ect: %he irtitifill increases with the altituie for a
1,300 meters in the Hiiualayns, 1.4~N in the Ghat,s, about 1,000
in Java, and about 500 meters alt.itude in the English Lake districts. Hann has shown from theoretical considerat,ions (see his C!limatology, Vol. I, p. 398) that the seasonal changes of this iuaxinium level proce.ed in such a way that it rises with incrmse of t,em >eratiire, as is also shown from its geoginphicnl distxibution. kow. our table for the Ardennes gives II nunier-
ical value for this increase. We nee the s~ininier rains increase from level to level, but the winter rains increase only up to
while and t Yl en again diniinishes. The inasinium zone lies at
A P R ~? isoa. MONTHLY WEATHER BEVIEW. 293
if. lr;
tis Y8
1WJ
174
the fifth level, and then again diminish. At the same tinie the annual period changes at the highest level. At a11 altitude ol 447 meters Lutrenmnge still has prevailing winter rains, bul at 467 meters Gouvry has suninier rains again. Therefore, in the Ardennes me learn to recognize two levels of revei~sal,
and we have here the following sttxtitications: A lower zone of suninier rains.
A zone of winter rains.
An upper zone of suiliiiier rains. We may conclude that the upper level of reversal rises toward the east in the same WHY a? does the lower level front the fact that Hollerath, which lies about 25 kilometers east of Malniedy, in spite of its altitude of 612 niet.ers, still has winter
rains. In the European region of our chart of precipitation, and i n all sewons of the year, the mountains appear as insular re-
sultr of winter rains in the midst of regions of summer rains. This holds good not only for the Geriiian central niountains, but. also for the French, where these islands of winter ribinfall
in all probability are tuore extended than we at present (w i
demonstrate. Outside of Europe in the region of the lowest ategory of annnd ranges such occurrelicen are certainly present, but. with one single exception,. ohservations are still wanting. Thid one case relat.es to the southern half of the Atlantic States of North A4nierica. which constitutes the most
' estensive of all hydro-meteoric! islands thus far known. We select here ILW an example five Mississippi stnt,ions between the Gulf and the mouth of the Mississippi. and utilize only the simultaneous observations of 1870-188S in order to anticipate all olijections :
A lower level of reversal at ahout 350 nieters.
An upper level of reversal at tibout 450 meters.
Mwb. Mm. I JIin. 730 pni2 1,.m 46 54 6 7M 706 1.1w 53 47 6 727 6'25 1,W2 64 46 ' 5 563 539 l.lW2 4 405 5.54
~
9(il
I % 1 9
-- ..
DistHIICf Stations. from the
New Orleans.. . . . . . . . Virklihurg _._.__.____ Memphis __.__.___._. csiro ..._._....-..... st. Louis. ............
hni . .w 310 ti10 s25
1,w
That the pre~ipiht~ion should he distributed in the siininier
season more unifor~nly than in the winter is quite normrll, but on the ot,lier hand i t is ahnormal that the winter rains should not only not, diminish toward the interior but should
increase. Froin Noreniber t,o April Vicksburg receives niore
rain than New Orleans, and in February this increase ext,ends
even to Cairo. This c w i only he attributed to the increasing altitude above sea level.f8 The factor L, that is to say. the aqueous vapor derived hv evapoiwtion from the surface of the land, tirnt has the upper hand in St. Louis in the suninier, hut over the eastern plateau the winter rain stretches up the Ohio River. This insular occurrence is combined wit,li a region of very sniall annual minfall. In countries with sharp1 defined siinimer iuaxinia we might
would occ~ir, hut the alternat,ions of the high Alpine stations
From the Wen- fielstein, 1,727 meters, up to the Sonnblick. 3,100 meters, all have decided r~uniiiier rains. The same is also true of Pikes Peak, in the Rocky Moiuitains, although winter rains oc!ciir 011
its western slo e at. 3,500 meters. An exception occurs on
the Pic du Mi$ in the Pvrenees, but we must not forget that on the northern and southern slopes t,he annual distribution of
rainfall is very uniform. The observations discussed by Klengel are of the highest interest in this respect, because
espect t,hat at least in hig I nioiintain ranges winter ixinfalls
ive a negative answer to this expect&ion.
11 Poasibly a different explanation may be found.*. A.
three layers of reversal are present here. to south we have the following stations: Going from north
Rninftill. Increfw with altitude. Alti- I stations.
1
tude.
, .
,
Wintcr Bummer. Annual. Winter. Summer. Annusl.
-_ I ! - I-
The winter rxins begin immediately ahove Bagneres: Plan- tack has suiniiier rains; the suniiiiit of the peak again has winter min. This depends upon t,he changes of the zone of maximunI rainfall. If we consider the curves showing the rate of
increase we shall see t.hat the niasiniiim zone ia at 1,300 meters a.ltitude in the winter and at 1,900 i n the summer. The winter curve rises at tirst rapidly to 1,030 mni. and then gr:idiially sinks; the sunimer zone of maxiniuni precipitation hasnearly l,500iiini., and thecomparison with the above figures shows t hut the rainfall diminishes rapidly in both directions.
(7) IUR. HENRY GANNETT.
With regard to bhe influence of forests, cultivated lands, or
arid lands, as such, on the aiiionnt~ of local 11~infal1, Mr. Henry
G:innet8t, 1x1s expressed hinisclf very clearly in the following
paragraphs which we qiiobe from jiage 375 of the .Monthly
Weather Review for August8. 1901:
An esample of the peraistenc.e of error is the belief that the presence or absence of forests has an intluence 11 on the
the fact that forested regions enjoy a heavier rainfall than those not forested, and jumped to the conclusion that rainfall is produced by forests, and as a corollary that the removal of forests diminishes the rainfall. * * * The situation is simply that the cart has been placed before the horse. Want of ixin prevents t,he growth of trees; want of trees does not prtwent rain. This position is generdly ac.oeptec1 aniong physical geo raphers, but the niajority of the
aniount of rainfall. Some keen observer long ago s etected
people still reverse mise :IIIC f effect.
(8 ) IUR. HENRY GIANNETT.
Under date of February 5,1909, Mr. Henry C+annett writes:
The relief of the earth's surface has a great influence upon rainfall. Theoretically it. should be so, and all precipitation nieasui-enientr and olJservations of stream flow and vegeta- tion sustain the theory. The fact that i.ainfal1 is greater upon
inountains than u$on the adjacent lower country is a matter of
c3oiiinion observation. Other things heing equal, the hi her
it receives. Vegetation is intlnenced by two elenien ts of climate-tern- peiature and precipitation. The fact t,hat a region is forested eneral, the best of evidence that it enjoys a greater rainfa B I than a neighboring region which is not forested. Fur- thennore, the species of tre.es indic.ate, and, indeed, roughly measure, the amount of precipitation. The Sierra Nevada, for instatice. bears upon its long western slope a succession of tree species which indicak very closely not only the temperature but also the aniount of precipitation. These timber belts are well recognized and have been traced over great areas. Under similar temperature conditions the bottom of the yellow-pine
the uioonbin above its base the great,er is the rainfall w a ich
224 MONTHLY WEATHER REVIEW. APRIL, 1908
belt indicates a certain isohyetal line which, in the southwest- ern United States, is very nearly that of 80 inches annually. Direct measurements of rainfall form therefore only a part. and a verv small p r t , of the data concerning the distribution of rainfalf; which IS available for preparin maps. The relief
erning precipitation, the extent and character of the tree vegetation, and the volume of streams relative to their drain- a e a r e q furnieh a large body of additional data. These are
which we.have few gnu e measurements.
are usex ignoring these additional sources of information, the map is not only imperCect but positively incorrect. It is, foi instance, certainly incorrect to show the area occupied by the Sierra Nevada with the same precipitation as measured a1
Fresno, Stockton, or Owens Lake; or.the area of the Wasatch range, from the nieasnremenh at Salt Lake City; or that ol the Colorado Mountains, f roni ~iieasiirenien~ at Denver, Lead- ville, or Grand Junction, yet that is what would be done if the gauge measurements only were used.
To what extent of detail the relief of the country should be recognized on a rainfall niap depends entirely on the scale of
the ma . If as small a scale as that of the United States Daily $eather Map“ he used, the influence of only a few of the greater ranges and platews can be espressed. Generali- zation is necessary here, just as in a topographic. niap dmmii
of the country, combined with our knowle 5 ge of the laws gov-
o s especial value, since they commonly relate to regions in
If, in reparing a rain f all niap, only the gauge measuremenk
. upon a small scale.
(9) PROF. B. E. FERNOW.
Under date of February ti, Prof. B. X. Fernow, of the
School of Forestry, Cornell University, Itham, N. T., writes
tts foll0w.s:
I agree with and share all the wishes of Professor Jeber- wn, bat I also admit all the reasons of Profe$sor Henry (Monthly Weather Review, November, 1901), why these wishes can as yet not he While as a student, of biology, and especially eco 4 ogy, I should, with Professor Jefferson, like to have correct rainfall data, I believe with Professor Henry that existing means-i. e., imperfect rainfsll gauges and deficient number of stations make it impracticable to secure them, and the desirable improvements hardly lie in
the direction of correcting admittedly poor data by an uncer- tain formula. Professor Henry does not perhaps s h t e as strongly as he might his objections to Professor .Jefferson’s proposition to correct rain-gauge records and to make allowance for moun- tain induence.s.
If I am suspicious now of rainfall dah, when they are rec-
ords of actual observations, such as the i-aiii gauges )ertiiit> I would certainly not look at them with iiiore conti 4 ewe if
I knew they represented “doctored” facts. I ,refer to do
the doctoring myself when I think that thereby t t ey may be itii roved. 8nder present conditions I believe the corrections mould
be impracticable, but I also believe that irn ~rovenient~s in rain
join with Professor Jefferson in asking for iniprbvemenh. Meanwhile let us have, SO far as mere records are concerned, nothing but facts, unadulterated-with the fi-actions of inches
“The daily map here referred to is on the scale of 1/1O,OOO,OOO; the rainfall maps of Messm. Gannett and Henry and the relief map, all in this current number of the Month1 Weather Review, are on the waleof 1/19 OOO,ooO; the r ular month& rainfall map of the Monthly Weather Re&w is OR the B s e of 1/25,000,000.-Ed.
ratified.
PJiges are possible and most desirable, an a in so far I woulcl
-~ -
of rainfall usefully omitted-and not hypothetical ones. When it conies to discussion of the observations the matter assumes
n very different aspect. I t has perhaps been overlooked that as far m the meteor- ologist is the recorder of observed facts, he is in an entirely different position from the geo ra her or climatologist, the
present the facts diflerently. Perhaps there is also a misunderstanding as to what the isotherniic or isohyetic lines on uieteorologioal maps mean, or, should I say, ought to niean with our present insufficient out- tit. To me they niean only a araphic method of presentin actual observations-in aiding &e eye to quickly see in w h d places (of observation) the, same conditions existed at t,he same time. To avoid conceiving these lines on the record ma s as
straight lines connect nierely the points of equality If this were done, nothing but the mere facts are represented and nobody will suppose that areas of equality are intended. When the cliniatologist or geographer proposes to use these facts for geneidizations, then. to be sure. A certain amount of philosophy-facts not observed in loco, but in geneial phy- sical IRWS, influences of contipuration, ctc.-must be adduced to make n reasonable interpretation and useful application. The meteorologist’s business is to make the hicks as accu- rttte and perfect as he niay, whatever the use to which they are to be put. The climnatologist, the biologist, the geogra- her, are the builders mho must have sense and knowledge ow to use theni. For the geographer nnd climatologist, then, I consider it right to hke into c*onsideration the probable intluencc of con- figuration. elev:ttion, etc., titid to niake his rainfall and tem- perature itre:is conform. to these prol>abiiities. wit.h the actual facts as B hasis. I repeat :&gain, it must not be overlooked that he is concerned only in generdizations, and not in such details as the meteorological recorder is. If the Weather Diire~ii, from time to time, issues cliniatic nnd not inet.eorological inops, then in these, to be sure, one woiilcl expect such we of their data as cotlforni to oitr generd knowledge of physicill facts and laws.
You ask “to what estent c;tn a rulc that applies to one mountain be extended to other niountains in distnt parts of
the country?” Who knowst I know this much, that nioun- tain peaks in lower Arizon:t, iietirly as high as Mount Wash- ington, say 5,000 feet, have no power of condensation ancl remain arid, while t,Iiose exceeding th:it hcight are covered :lown, to points far helow the 5.01~1.,-foot level, with :i forest wrowth-the result, of precipitation. I believe with Professor ketiry that local variations will be suficientr to prevent any
“Is the connection between foreet,s and rainfall so definite
is to warrttnt utt8rihutming henvier rninfall to :I forest region as
zoiiipared with an adjacent bare. rtlgion ‘i“ ”
As yon know, I ‘consider t81iat. :is far :is observed or meas- .ired data are concerned, nothing detinite is known regarding ;his connection. The philosophy of the caiises of rainf:dl would w:wrant us
.n aasaming that if an extensive area were covered with nn- ~roken dense forest cover-such cover as would protect the roil a ainht inso1:ttion. creating a large area of cooler, wore
;hose of an extensive area entirely devoid of this cooling .nfluence. But I take it that, in order to be appreciable and If practicable value, there would have t.0 be a certain propor- ;ion between these conditions of the soil cover and the other iieteorological and topographical factors that are involved in :rtusi ng precipitatiou. As a rule, at least in settled parts of the country, there is dternation of o n field and wooded area, an$ here, I should
be inclined to tgnk, the influence of the wooded part would.
interpreters and generalizers o f f t ese facts; the two must
atigthing else, it might be wise not to smooth them, hot g. five
E
dimhle geneidiz. ‘1 t’ 1011s.
iitnii d air-its rainhll conclit.ions would be different from
APRIL, 1908. MONTHLY WEATHER REVIEW. 885
counteract the influence of the open couutry and aveiqge con- ditions would prevail as compared with the estremev of unbroken forest or open plain. I believe common observation supports this philosophy, although measurements of any value are entirely absent i t i
spite of the attemph to obtain thein by the German and Ans- trian forest meteorological stations. The trouble here lies in the difficulty of excluding or discounting all other influencen except the one to be meamred, namely, t.he forest corer, and
in the general inability of the rain gauge to measure. I should therefore consider it a risky advance beyond our knowledge
to “doctor” the facts in this direction. Let me close this very hasty letter with the wish th:tt. a
broad- u e policy may permit the Weather Bureau to advance
are so much in need. rnetho rf s o meteorologicd inquiry and statement of which we
(10) PROF. B. E. FERNOW.
Again, under date of March 91, lSWS, Prof. B. E. Fernow
says:
I n reply t.o the query, “To what extent dow the premwe of a forest clenionatirtte the fact that a cei.tain amount of rain is being received annually ’! ” I can only answer that we have no data on which to bane even a guess; iiioreover the ahsence or existence of a forest growth is always dependent on iiiore
phcnomentt in combination thnn the one fact.or of rainfall, which may even be the least important, although a necesawy
one. That there is a relation between moisture conditions and
the existence of certain floras, hence of forest growths, is well known, and the moistmure Conditiolis may become so pre- carious that, it). c~nbi;ntrtion with ofhr cod;ft‘tj)>a, thep lead to forestless conditions. Rut. it. irc not, the aniount of rainfall, at least not directly and primai ilg, t h t produces the distri- bution of forests. No one factor can be so segregated froni
all others as to account for the coniplex phenoniena of dis- tribution of life. Rainfall a d relative humidity niust be at onc’e placed in relation to tenipeixture and all other factors influencing life; in tree groa7t.h especially. what I call the timspiration factor, is det,erniinativs, namely. the relation of
available water supplies at the root (which also depend on soil conditions) to the consuiiiption at the crown by the tixns- piration of the leaves, due to tempeniture, relative humidity, charnct.er and velocity of winds, especially during the period
of active vegetation. Above timber line, dthongh precipitation is aniple and
relative humidity is high, forest growth is absent, because of t,liin soil, combined with steep slope, whic i can
not retain sufficient inoieture. partly becmise of low tempern- tiire and, still more, liabilit,y to frequent frosts, and because of several other causes. I n the desert and plain, forest growth
is absent, because relative humidity and high t.eniperature, and especiallv dry winds. niake more demands u on the trans- piration current thnn the roots can supply. gate thut, the rninfall during the period of vegetation is about the mule
nt Dodge, Kilns., as in Philadelphia. but the dissipating influences, the evaporation or transpiration factors are very different. It has been asserted that tree growth or forest growth can not exist where the relative humiditv sinks below 5@ per cent during the period of vegetation and the precipitation below
50 mni. Yet, while we may accept these lowest limits. in combination with the factors of evaporation :tnd soil condi- tions usually found in such places, as generally true, we must also admit that, if the latt’er factors are modified, for instance by protecting,mountain ranges or improved water storage in
$rt’lY
--
the soil, forests may still be absent., because the mechanical nieany of estahlishing them in coni Jetition with other form of
could not be rown there. and, in fiwt, forests have been
our own country and elsewhere. I would conclude that rainfall is not the most important or controlling factor in forest distribution: that it is almost im- possible and futile. ctvtainly impracticable. to segregate any
one factor M cont,rolling: that the conihination of factors, which ’r call the transairation factor (whivh I aclmit is intrtngi- blc. escept in conception), controls in most. cases the existence of forest. growth. The distrihtion of our species and the possibility, piwti- c:tll-y attest.ed, of t~r:insporting them succesfully from one set of rainfall conditions to regions of entirely different, rainfall woelrl go a long way to niake the esistence of a very direct relation doubtful.
vegetation were deficient. but it a oes not argue that forests
grown where t I ey were not found by nature in Russia and in
(11) M R . F. H. NEWELL.
Under date of February 4,19(!9, Mr. F. H. Newell. hydrog-
rapher, United States Geological Survey, writes as follows:
For at least twelve gears I have l.)ecti studying nonie of the rainfall statistics gathered by the We;nther Bureau, and have given eapecitil attention to your rainfall inape, and have fre-
c uently talked with Geneid Greely, Professors Httrringtm,
the dah. I iave coiiie to regard your norninl ininfall ilia 1,
not as abaolutely worthless, hut rather us misleading in te - ing only a part of the truth and in ignorin itiatters of coni- nion knowledge. I have never republishe f your map in any of the reports or papers I have prepsred privately, but, when- ever I have had occasion to use such niap, have rechwn it, using Mr. Gannett’s sketch as a guide. The Weather Bureau rainfall ninp ia undoubledly fairly good for the 1110re thickly setrtled part of the United Shtes,
where there is not a great diversity of topography. but for the western two-fifths of the country it is very misleading, Iwcause it ignores the great. mountain ranges and the great differenc?es in recipitation due to their presence. For
storm npoii the. higher summits of the Rockies, and those of us who have snrveyed and camped through these mountains
are well aware of the heavy precipitation, and yet t.he Weather
BLL~WU apparently ignores the fact: The same is true of the Wasatch. Cascade. Sierra Nevada, and other great mountain masses of the continent. There are few, if any, stations fop
observation of rainfall within these niountains -from which come the imporbtit, streams vital to the development of the West. If we measure the aiiiount of water flowing out froin sonic
of these iiiountain ~tremiis and find the total volume delivered during the ent,ire year. and then compare this with the pre- cipitation a.sr shown by t,he Weather Bure:tu iiiap, we reach the astonishing conclusion that in sonie instances over 1oij Ier cent
of the rainfall reaches the streanis. At one time $ spent several iiiontha studying these unonialies and cinie to the con-
clusion that the Weather Rure:tu stmatistics of irtinfall do nof go far enough to be of any value whatever in the study of river flow for the more iniportant streams. I t seems to me that, eit.he.r t,he niap should hr left. blank for these great. moun- tain masses, thiis frankly confessing that no inforniation is ttvailable, or t,liat careful approsiniation or intelli ent guesses ~houltl be inade hy uien who are thoroughly faniikir with t,he iiiountaina and know froni personal experience soniething of the distribution of ininfall.
h i
,loore, It Henr 7. and others about8 the method of presentation of
exam ,le, in loo tz -ing out of the window of the otiice of the Weat 1 er Bureau in Denver we can see snow storms and rain
’
226 MONTHLY WEATHER REVIEW. APRIL, 1909
Reference has been made to the desirability of correcting rain-pi e recorh for height, etc., but this seenis to me to be splitting%ain, for, as pointed out, it is useless to discuss deci- mals when the wits are in error. The rain gauges at the city stations are probablv well cared for. hut those i n the country and i n charge of vor11nher observers are frequently in deplor- able condition. I n m y studies of river flow, I. have tnken pain# to visit a great many Weather Bureau gauges iti the hands of
voliinteer ohservers, and the more of these I see thd Jess con-
fidence I have in figures of rainfall obtained in this way, except- in aa the results of one observer are contiriiied hy nuiiierous
graphic in formn tion. It is stated by Professor Henry that “each section director is expected to visit stations in his district and c0rrec.t fitulty exposures, when such are suspected to esist.” 1 understand, however, that such inspection exist,$ in theory rather th:tn ia fat, since the funds availalJle for such inspectmion are ex-
tremely limited, and unless railroad passes are obtained it is iuipossihle for the section director t,o travel over his district. Moreover, his duties are so varied that tis it matter of fact he rarelp is able to gain :t personal knowledge of the country for many years. Ln fact, I questtion whet,her in the west,e.rn half of the United States more than 1 or 8 per cent of t.he rain
e u es have ever been inspected hy a competent, person.
Tfe result of the.cle erroneous reports a pears in the anoilin-
lous figures occasionally puhlished. I E v e fitrely visikd a
volunteer observer and found the gauge in good condition. Freqtient.le there is water stnnding in it from the last min storm.” bonietiuies it in near st shed or tree. Even at arnig
posts, where the meteorological instrunients have h e n con- hded to the care of the hospital assistants, t,he conditions are sometimes faulty. I might write pages of incidents and coiiiplnints of thc in-
accuracy of rainfall observations, hut I appreciate the sur- rounding circumstances and know that at, best the observntions of irtinfell represent not so u1uc.h absolute as relative ditfer- eiices, and that it is only by coni aring aonsicleirtble numbers of observations with each other t\at we can arrive at general
conclusions as to the distribution of the precipitation. To es- hibit our full knowledge of the dist,ribotion of r:iinfall we must hke into account other factors beyond the rain gauge, taking what lawyers woiilcl term udicial cognizance!’ of
there are inore frequent and longer storms upon tlie nioun- tain masses than over the broad valleys wh‘ere people live. This is attested by the luxuriant growth of vegetittion and hy the rivers which drain the uplands, and vet, ignoring this fact, maps of precipitation are niade based wholly upon the single fact of rain-gauge mensnrenients made in the valleys. It is not possihle, on it stnnll map of the Uaited Stat.es, t.0 show the minor elevations of one or two thousand feet,. such
aa those of New England, but the Appalachian Range should always he given, and also the principal mnge of the liocky,
the aso ode, the Sierri Nevada, etc.. Oniiv&ons of this and disre ard of other climatic etFects seeiii to me to re,nder the
than the broad valleys o t%e country.
ot % ers and agree with the evidence deduced from geneirtl topo-
matters of wmuion knowledge. We “i mow, for exaiuple, that
F
Weat 5 er Bureau nictps ractical1.y vitlueless as r e g d s ot.lier
(12) HR. F. H. WEWELL.
Again, under date of Februar? 17, Mr. F. H. Newell writes
I have been giving particular notice to sonie of t,he maps of
It seems
@Of wume the rain water will always remain in the gauge until me=-
as follows:
the monthly reports of the State weather services.
U r e d .4 . A.
to me that i n conipiling these too little attention is paid to local topography. oval hnee being sketched with respect solely to the points of olmrvation when they tnight be deflected to skirt! along t.he nioiintain mxsscs, in aacordance with the known facta of precipit,rrtion. A tiinst incongruous condition also results when t,he rainfall m:qs of two States are placed side hy side, each having 1)cen niade independent of the other
and without niiy regard t.0 t,he great. mountain masses. The result,! it! seem to nIc, is not, creditable to t,he Weather Bureau, becauiscb itr seems t,o indicate :I blind following of scat.tered data
without- regard to other well-known phenoinena. Take t,he last maps of annual previpittibion, just received, those for New Mesic.o and (’olorado, place them adjacent, and notme the est,renie lack of coincidence and disregard of a11
mountain masses and prohahle direction of stortiir. Sonie of the niost, rugged iuount,aitis in t,he coirntry lying in the soutb- western portion of Colorado are shown as having less than 10
inches of rainfall, while on the other hand the broad desert. of the Monte.zuim Valley is given 15 inches or iiiore. There is a lit,tle c!ircle of 85 itic!hw mound Pikes Peak and around one or two iiioiinttiins, but! the rtmitining great mountains upon which there is 1~r(-~h:thly 1x11 equal or greater precipitation are
le,ft. out. Taking t,he sime data, I have asked one of our men, who is faiiii1i:xr with the t,opography. to sket,oh n rainfall iiiap? with the resiilt that his drawing. while equally true to t,he dabs ohtained, brings what seems to nie to be L far bett.er showing,
and one which e m not he open t.o the. chrge of neglectin well-known geogriLphicd htcts. The lines of eqaal rain faJ& instead of Iieitig ineani ngltw oi-als, w e deflected along the mountain inasses and sqmrti,te the hrcmd valleys, upon which we know there litis been littlch prec!ipit,ntion, froin the elevated re ions where, from bhe iip ~ettirtnce of blie snow and from the
has hem a heavy prei4pitation.
vo 7 time of water discharge d hy the rivers, we know that there
(13) PROF. QEORGE L. QOODALE.
Under c1ut.e of March 20, 1902, Prof. George L. Goodale,
Hotunic? Gardens, H:trvnrd University, Cumbridge, Mass.,
writes as follows:
In answer to the question ‘* Do special types of forests ire any de,linite inforiiintion its to tlie annual rainfall or nie 7 t.ecl snow 0 ” I a m coinpelled to snv that this must he answered now
in the negat,ive. Wit.11 :to idsition of new data, it may be and
The optimum conditions of vegetable Itct,ivit,y are well known for iiiany p1ant.a. and t.he functional rel:it.ions are so well under- stood that from two known quant,ities one can estimate an iinkno~n quantity. For instance. given the heat i d nioistiire
availahle, one r:in stnt,e what tlie luxuriance proh1d-y is. Or, given the lusuriance and one of the two factors just referred to, t,he other can he judged. Thus. if one knows the lusuriance and the available heat. t,he sniount of nioist,ure necessary to IJriiig about, this lusuriance is easily conjectured or, perhaps it is better to sag, calculated. If one could know ,whether a given forest of white pine is stunted or t,lirifty, or, to be inore precise. could know its rate
of gmwt.li, and at the mnie time should know the mean an- nual teuipeniture, the aiiiount of ntinfall could he fairly deter- mined. Now, it is questionahlt? whet,her enough is yet known in regard to the rate of growth of our mountain forests, and in regard t,o t,he nitfan annual teuiperature in those districts, to make it. worth while to gueds (i;s to t,he rninf:ill. With more
acouixte data as to rat,e of growth and either of the two fat!- tors referred to, it will be piwt.icahle to deduce tlie other factor.
l~rohttl~ly will be atimerec P afirninbivelp.
A P ~, 1908. MONTHCY WEATHER REVIEW. 527
But, on the whole, it appears as if, in the forest districh, especially in the mountains in question, it might be wise to establish shtions in which meteorological and forestq prob- lems could be worked out together.
(14) PROF. C. S. SAEGENT.
Under date of March 20, 1902, Prof. C. S. Sargent, Arnold
Arboretimi, ,Jamaica Plain. Harvard University, writes :is
follows :
The rainfall, particularly on high nioiintnins. appears to be so greatly influenced by local cawes ot.lier than forest. growth that I should not think it would be sde t.0 iiiake any siich generalization as you suggest. 1 should suppose that the character of a mountain forest would be 1arp4y influenced by
tem emture-that is, hy altitiide--but t,hat the size of the trees
of large trees would indiciit.e a larger rainfall t,hmi a forest of snialler trees of the same kind.
wou P d be dependent on moistnre, aad that a mount:tin forest
( 16) PROF. JULIUS HANN.
I n the matter of utilizing the presence of forests as R guide
in drawing isohyetal lines, t,he opinion of Prof. Dr. Julius
Hann was recently expressed in convertintion with Dr. C. Ahbe,
jr., in Vienna, as follows:
Wheu there are but few ratinfall stations it is proper to dinw isohyetal lines provided it, be diatiuctly stated t,hat t,he loca- tions of these are estimnted. and provided thtit the basis of
estimation be distiuctly stiit,ed, and t,hat, fnrthwniore, the rain- fall wtually obse,rved he clearly entered on the, imp and properly distingnished froiii the supposed rainfnll. A nittp
whose ieohyetals are based principally upon the esistence or
n0nexistenc.e of forest.* slid the dtitiides or orography of the country can only hare v:tlut? as a pedagogical aid to the pres- entation of the laws nccording t.0 which it. WRH coiistructed, but has no scientific value 1:teraiiee it h:ia no suficient number
of actual nieasiirements of rainfall to support it. It is doubt- ful whether the estimat,ed rrit,io of run oil’, or the discharge of rivers, to the precipitation over the river hnsin is known with suficient nccuincy or can in any cwe be estimat.ec1 so as t,o enahle us t.0 reconstruct, itn accurate syst.eni of isohyetal linerl even when the measurenient~s of the streams are s:ttisfactory. When there is no nieasiircment of run off or of rainfall it would not be justifiable to draw and puhlish cren hyrotheticnl isohyetes. In Aastrirt, where there nre long series of obserm- tions at nuinerous stations representing mountain tops t ~n d valleys, the isohyetes are drswn in accordance with obscrrtt-
tions aud on lnrge scale maps :itid without, inbroducing hypoth- eses. Of course it, is easier to draw niaps on a very sniall scale, as is done in t,liose of the United States that are sent to
Vienna. In regions not, otherwise provided for it, is desirahle to establish gages that will hold a week’s or :I. niotrth’s supply of rain and read them as often as it. is prncticnhle for a uian
to visit them, SO that wc m;i? at least get t,he iiiont,hly and
annual sums. In Aiist&i. such gages usuttllv give higher readings than the sum of the ohservations niahe in the usmtl niauner at 9 a. in. daily.
( 16 ) PROF. JULIUS HANN.
Again, under date of Vienna, April 6, Prof. Dr. ,Julius
Hann writes na follows:
The question that you suhmit to me is ixther dificnlt to answer. It is so much t,he more dificult because in your case
:of the Monthly Weather Review charts) you have to do with isohyetal charts for individual months and not the uvei’ft e
yeneiltl rules for the increase of ininfall with altitude and the local rules with reference to the prevailing direction of the wind are to be applied approximately to the determination of t,he act,iial distribution of rain; for a specin.1 month the intlu-
ince of the direction of the wind can scarce1 7 be estimated, and still less so the iiifluence of altitude. Tie presence of forests (for exam )le, in t-he western portion of the United
M an indication that from certain altitudes upward i n that region more rain falls than below, hut there is absolutely no Imis for any e.stimste as to the.anioant of this surplus.;, and any isohyet,als thus drawn on the chart would he purely arbi- trary. The quantities of run-off from the rivers are such also that they c.an scarcely he used; the rat,ios for the reduction from run-off to rainfall vary extraordinarily in a region like that of the United States. especially in the dry Rest.. I n those portions of Euro e for which the best class of
list in in)’ Lehrbuch. pages 359-360) the conditions w e more hiple. h i t here also even the chnrbs of rainfall distribution For the average of niany years are very unreliable for reall- iiountainous countries (see, for esaiq.de, the IsohgeLals of
Switzerland, by Billwiller), hut they are hetber in the region
3f t,he 1Clitte;lgebir e (in Bohemin, near Saxony).
lrawing isohyetal charts for individunl months-having regard
to altit,iide, wind diredon, or the distribution of forests--that hall represent anything 1110re than a siihjective work of the imagination. In so far as bhe distribution of min thus pre- sented depart.s from the quantities actmrlly nientiured it is xrbitrary, and every drsftsmrtn will produce a different picture. Such charts may even be hurt.fu1 and productive of error, if it
is forgotten that in niany places they are purely hypothebical. When one looks at something that is well drawn. he generally
ttssutnes it t,o be correct. Even the espert can not, of coiirse,
without some reserve, divest hiniself of the impression tiiade
by such a chart. Therefore I can not advise yoa to present wch purely hypothetical chnrtu t,o the public. But one circnmntitnce to which yoti allude can lend to a inaterial iruprovenient of t,hQ monthly chnrts of rainfall, viz, the introduction of the isohj sen or contour lines for each thousand feet. of elevation. :ontrihute to the. rtccuiwp with which you originally draw the lines of equal rainfall, dnce the student mill not be likely bo .Imw his isohyettils in very iniprolmble Htyle right over nioun-
hin heights, but will more likely follow the contour lines.
isohyetaln for a long period of yeltrs. I n the latter caqe t 6 e
3htes, in the Roe 1 -y Mountains, etc.), is indeed to be regarde,d
:harts of niean aanual rainfal Y have been constructed (we the
My opinion is t I at there is no well-established hsis for
? his addition to the charts will
_-
I
For instaim, without being guided by the cont,our lines, one night, in Fig. 1, di.nw the isohyetal 3.5 right, over t.hc siini.
nit at “8.” whereas by taking this siiiiiuiit into consideintion .t would seem better to draw the isoliyetd aronnd as at C C! C ind merely inscribe the words ‘I inore than 3.5” on the sum-
228 .‘MONTHLY WEATHER REVIEW. APRIL, 1902
If a rain ma 1 utilizes different shades of it certain color, I do not think it advisable that., on the same inap, forest,s and
mit, which in fact also comes nearer the truth, One mi ht, nioreover, think that he should take amount of the prevxifing direction of the wind; but in thiR case one would have to draw a special chart for each i.aiufall in the month and for each direction of the wind and then compile the monthly chart from the summation of these individual charts. But this is not practicable. Moreover, as. is shown by the distribution of rainfall in the lake district of Cumberland i n northwestern
En land, mountains of 9,000 or 3,000 feet in height, or even hig a er, do not form a irtinfall divide, espec!itilly not in the warmer regions and the warmer seasons. I n northwestern England [and similtir locations] the greatest rainfall occurs beyond the watershed and there is no a )recial)le ditference between the rainfall on the windward ani8ee sides. The very
great influence of the details of t8he configuration of the earth’s surfwe contributes to this. These dilferences, which are often quite unsuspected, are only to be recognized I J ~ direct nieasure- men&. The only improvement that 1 can recommend is the intro- duction of the contour lines into the rainfall chart.s, so that they
ma be taken int.0 considernt.ion in drawing t,lie isohyetds
of rain falls at t. e mnie level and t,hiit illore falls at greater altitudes? which niay be expressed by the words bLabove - - - -
inches.” According to this principle the charts of aveixge
isohyetals in Europe are constructed, in doing which definite quantitiea of rain are assmned for the higher elevations. t,lie estimates being made proportional bo t,he increase wit;h alti- tude, a8 determined hyv measurements iiiacle elsewhere, a
process which seems allowable in chart,ing the general aver-
itge distribution of rainfall for iiiaiig yews.
I n the eastern portion of the United Stntes where there are
no high mountains coold not. contour lines of 500 feet he drawn’! For even this sli ht, elevation has sonie influence on
across such elevations.
E
un dy er the most robable assuiiiption that the saine quantity
the rainfall, and one oiig R t not to draw isu1iyetaI:tlr;l directly
(17) PROF. H. GRAVELIUS.
Under dnte of April 9, 1!)03, Prof.- Dr. H. Gravelins. of
Dresden, editor of the Zeitrchrift f iir Gewiisserkmide, writ.eF
as follows:
PrOt,ess of Forrectibn c,,oLlld ollly lead to hopeless collfusi;n. ”
I The impossil,ilitv of a,.rivillg at, allownllce for ,,levation
of the land as determined by contour lines or isohypsen. This fact ip, for instance, well demonstrated by Dr. Hellmann’s map of eastern Prussia, where even moderate elevations are precise1 characteiixed by the lines of equal amount of rain.
Kingdom of Sasony that I published two years ago.
and hydrographer shotild always be published only on a rela- tively large scale and therefore only for a somewhat restricted area.
(3) If, €or any special cause, it is desired to show on the map the forests and the relief of the land, simultaneously with the distribution of rain, the latter should be given only by isohy- etal lines without using any kind of shading. This latter met-hod is used by the ‘’ Centralbureau” of Baden. 1 think that from imps drawn according to this mode i t may be seen that the area of a mountain top havin rainfall is uot cr77(vryt? negligible by compwison wit,h t e aieas of loalanda having an average rainfall.
An ana r ogous resalt iiiay be drawn from a rain map of the
Finally, I come to the followin conclusions:
(1) Ratnfall maps designed for t % epru,c&d of engineer
% he:vy
Abnve 3.OOO feet (3 records) ....................................................... ?..%XI to 2.W fret (3 records).. ..................................................... 2 .W t r ~ 2A9Y f6et 11 rerorrl). ....................................................... 1..W tat 1.W feet (3 records) ....................................................... 1.OOO to 1.4W feet (X recnrrls) ....................................................... 5(ro to *WY feet. 12 records). .........................................................
(18) H. SOWERBY WALLIS, ESQ.
LTnder date of March 1. 1909, Mr. H. Sowerby Wallis,
director of the British rainfall system, writer ax follows:
41 77 1%
112 145 83
I have read with much interest. the discussion on the reduc- bion of records of rain gauges in t,he Monthly Weather Review for Noveinher, 1901, and should like to indorse Professor Henry’s protest qpinst the proposal that lain records should he corrected. Chtuges iii had positions can not yield satisfactory results by correction, and it mnst be i.ecogni,zed that the records of any individual gauge ma.y a )ply to an extremely liiiiited area. But
iiie-aft,er 30 gears devoted entirely to rainfall-equivalent to saying that we should determine the precipitation by the scientific use of t,he imagination. That in conxtriicting hysiographical maps, allowance mist
is obvious, and it is probable t.hat for the centid regions of contincn ts considerable accuracy might he attained in correct-
ing records, hut for coast regions where there is it well-marked ixin-bearing wind. as for iiisbnce in the Hiitisli Isles. anv
to siig.pt the wholesti t e correction of rain record9 appetm to
he niade for dewtion w \ iere there is absence of inforluation
For, firstly, if +e cw4der the matter oily from a purely meteorological point of view, there is no need to distinguish especially the regions of forest!, a# will he seen, for instance, from Dr. Hamherg’rj iiiquiriw into t,he influence of forests 011
the climate of Sweden. On the other hand, in connection wit,li hydrographical researches. it limy indeed he desirable to dis- tinguish forests, but this should be done by some other sign than a shade of the color used in showing the rain areas. No doubt, in all questions regarding waber supply. the fore& will be of considerable importance, but froiii ht,e European researches it seems not to he so in respect to the pirely climatological question of t,he geo raphical dist.rihution of
expected to be one of g o d forest growth, but we are in no
way allowed to consider t i p G )r i a well-forested district as one of greater amount of rainfall than any adjacent region con
trolled by the same geographical condit.ions (i. e., elevation distance of the ocean. windward or leeward ex mure. etc.).
atiou takes laee. In this respect, no clouht, the scale on which the map is 3 rawn becomes a matter of controlling im~ortance.
If this scale is large enoi~#i, then the isohyet.al lines will. sometimes in quite an astonishing manner, rcproduce the relief
rainfall. A region of sufficient pear s y amount of rain iiiay be
Further, as to the influence of elevation, mot h er consider-
purpose; it seem to us that the irkhieke of position with re-
s ect to hills and vnlleys is far greater than that of altitude. l t e have tried grouping the redricecl iiieanr; accorcling to alti- tude, and l!ere is the result:
.... ........ - I lean.
A corresponding relative variation would be found in most parts of Britain.
APRIL, 1902. MONTHLY WEATHER REVIEW. 829
(19) DR. C. HABT mRRIAM.
Under date of March 20, 1902, Dr. C. Hart Merriwi, Chief
of the Biological Survey, Department of Agriculture, writen
as follows:
Tour question as to the possibility of basing iltiiifall maps
on the distribution of foresb or of a.nimal life is an interest- ing one. and one to which I have given iiiuch attention. In my opinion it must he answered in the negative. Humidity hm more to do with the distribution of life than has rainfall. The great dificulty in studying such problenis as this ib the insufliiciency of the cliiuatic data. As a rule the chta are not plotted with sufficient detail for our purposes. For inst,ance, we now know very exactly the distribution of large nuinhers of aninials and plants in the far West, particularly in Cali- fornia, but we have no ilia s which pretend to show t,he dis-
any way approaching the necessities of the case. Fnrther- more, when we come to temperature, nearly all the riblished niaps show its distribution by arbitiwy periods, wfich have absolutely no relation to agriculture or zone distribution, and which appear to be worthless for scientific biology. If we rould have niaps of the West coast region, taking: in, say, Washington, Oregon. California, and Nevada, showing the distri1,ution of hnniiditF by nionths, and isotherms giving
the total quantity of heat for the period of growth and repro- duction. and also ibot,herms showing the niean arerage tem- eratnre of t.he .hottest six weeks of summer, p,lot,ted on n
Erge scale niap. we should have the foundation for an intel- ligent dudy of the climatic control of agriculture and of the geographic dist.ribution of our nstive plants and animals.
trihution of temperature, R ainidity, or i~tinfall on a scale in
(20) PROF. R..F. STUPART.
Under the date, of April 11, Prof. It. F. Stupart, director
of the nieteorologioal service, Toronto, Canada, says:
With re aird t,o t,lie prcpaixtion of rainfall charts. we have nearly finis % ed tabulating the iiinterial for raiilfnll charts of the Dominion, :ind I hare been quite in doubt as to how the isohyetnls should he c1~a~v1.n. However, I am inclined to think if stations are far apart, as in our northwest territories, it mill be h t t e r to give simply the rainfall at stations, as on the South African and Aust,r:ilinn charts. In other cases where stations are not very numerous, I ani alniost inclined to inter- olate strictly hetween them, making no allowance for surface reaturcs, unless. perhaps, where higher lands lie imnediately to the eastward of water aurhces. Is not the'foreat influence too doubtful a qiiantit,y to allow fort
(21 1 PROF. W. H. BREWER.
Under dat.e of April 11, 1909, Prof. M7. H. Brewer, of Tale
University, New Haven, Conn. writes as follows:
I do not think that special types of forest give any definite inforniation as to the annual (amount of) rainfall and snow.
It is rare indeed that forests occur where the amount of an- nual rainfall is less than 18 or YO inches, unless it way be in regions in which the rain falls at t,he ri ht time to be most
that the aveix e tinnual rainfall may ire very misleading in-
portant factor. Trees are long-lived, and the exceptional dry years (or periods of several successive dry years) may, and probably
available. The r n q e of annual rainfall a as so much influence
ferences, the c f roughts of esceptiona K years are such an im-
30-10
often do, prevent forests in some regions where the average rainfall would be sufficient if more uniform. It seems to me to be a general rule that the range of annual rainfall is niuch greater in dry climates than in wetter ones; apparently the less the average rainfall the greater the rela- tive iltnge, and this, taken with the allied factors involved, would vitiate conclusions as to the actual amount the forests mi ht indicate. lome years ago I looked up the matter of the actual imge (with another object, however, tha.n its relation to forests), and take iny illustilttion of what I mean from iny old figures, which, however, were in most cases deduced from unsatisfac- tory data. Here in New Haven, from lS71 to 1889, or 18 years, the niininimn rainfall was 39.46 inches; t.he maxinioni 60.26 inches (the greatest recorded here). If we state the ratio after the fashion of the old arithnietics of niy " district school " days
we have
30.46 : 60.86 : : 1 : x=1.53.
According to the tables compiled by Professor Looniis for New Haven, iiud extending over about a hundred years, the driest, year of the whole period had two-thirds as great rain- fall as t,he wettest. At Wallin ford, near here, during 18
data for 28 years. 1 : 1.59: Cambridge, Mass., for the sanie 28
gears (lY49-1876), 1 : 1.48. I think that in that great originally forest-clad re ion, from
from Maiiie to Alabama, we have no place where t e rainfall
of the driest year will not be more than half that of the wettest one. Probably the same is about brue of central and northern Europe. Observations in Yaris from 1688 to 1871 or 183 years (according to a statenlent I have seen)-gave the ratio 1 : 1.44
for this whole period. which is much less than the lifetime of
our oldest trees. When I was in Denver last suninier I could but notice t,he
spread of trees with ciiltivatioii there, at Coloirido Springs, etc., and where (as I was told) trees did not flourish unless irrigated or within reach of irrigating waters, I found trees where thirty-odd years eiirlier it was treeless. When I got honie I looked iip the rainfall froni such dakta as I hrtve. I
had on1.v a few years-Denver, 1891-190~:4 inclusive (lacking
1893). The range was: Minininm, 8.48 (in 1PHSj; niaxinium,
In California, and about which I was foriiierly more inter- ested, and using the rain year (from July 1 to June W), and nccording to the figures I had: At Sacra.mento, thirty years (1849-1880), niinimum, 4.7; niaxiniuq 36.4; ratio. 1: i.75. In 1896
I was told by one of the oficinls that at the reservoir south (where their great dain is Imilding) the amount of water falling Itt that place in the wettest years was ten times as great as in the driest years.
I saw a statement in Nature .a few years ago as to rainfall
in Sydney, Australia, and from the figures the ratio was
1: 3.58.
No rainfall appears to be too great for forests to flourish. The uncertainty would be in a climate where the average was less than 80 inches, but might extend to where it was consid-
erably above this. Shrubs will grow and forin dense chappara1 where the dry years are very d y , and some species will flourish where there
may be a succes~ion of several " rainless" years. Row forests reyqire for their existence a iltther complicated set of conditions, of which average lainfall is one, not too excessive droughta another, certain conditions as to soil another, and other plant corn etitors. Much of the rairie region east
gears the ratio was 1 : 1.40. At F roriclence, R. I., from other
91.43 (lS91); ratlo, 1:Y.53.
At ,%n Francisco, twenty-four years, ratio, 1: 6.65.
of the Mississippi €F iver has abundant rain P all for forests, but
root at all, in others for the young seedlings to long sui.vive. Then there are mechanical conditions of soil; and even other
factors, as mountains, come in to aid in the preservation of
trees, and so on. So that I think that in regions near the limits of rainfall natural forests the existence or absence of forests,
been prepred. This matter has heen referred to by the writer in the die-
cussion of Mr. Noble’s paper on Gagings of Cedar River, Washin@on,“ and the statement was made by me that in the State of New York both conditions obtain. The Hudson River catchnlent area shows a higher precipitation at the mouth
It eeenis to me that it would be perniissihle, in conipiling vour rainfall charts, for regions where instrumental data nre racking, to draw the isohyetes according to your best judg-
inent from a11 the indications, and xiat some nrbitmry sign (w for instance (a ) ), with the state CI inforniation that it indi-
mtes &’ inferred rainfall” or b‘ani~~int inferred from the vege- tation,” or by some other statement. The compiler of the chad is suppolred to have used a11 the data and information at
higher precipitation at its source than at its mouth. I n Eng- land it IS almost universally true that -precipitation increases with altitude,b but in this country it is by no n~ea~is a universal
‘rule. Indeed, the opposite freyuentJg occurs. Acc,ording to a table of average monthly, annual, and sea- sonal precipitation in Mr. Tiirner’s monograph on the Cllimate of New York State,‘ it appears that the const region, which includes the followin stations: I3loc.k Island, E. I., East
sent that P’’tiOn of the State of New Pork which especially Pertains to dee and which? indeed, illclUdes
the whole of t E e State. The facts fihowll t,hereon are the
anllual plf,cipitation of 35-68 incshes, while the Hudson Vallpy, which includes stations in Piltna111, 01x11 e, I)utc.)less, Ulster, ~o~l~lllbia, Albany, ~~~~~~l and wits illvtoli col,lit,ies, has
h u , 1902. MONTHLY WEATHER REVIEW. 231
years; Fairfield Academy, 1828-1849, twenty-one years, Ginn
ville Academy, 1835-1849, fifteen years. Assuniing the North ern Plateau m a unit (i. e., groupin6 several locations as on( station), the total number of years IS 1994, and the mean oj
all is 37.4 incheb. A reference to the rainfall niap in tht Report to the United-States Board of Engineers 011 Deer Waterways,n will show that this is necessarily an approxima. tion, because of the great lack of stations in the interior oi this region.
As regards the catchment area of the I T per Genesee River.
the source of the river. For the years 1889-1898, inclusive! the rainfall in the upper area of this stream was 49.19 inches! while at Rochester for the same years it was 35.64 inches. This stateiiient is especially interesting. hecause there seems
to be a well-marked line dividin the smaller izlinfalls of the lower area from the higher rain f alls of the upper. At Hem. lock Lake, Avon, and Mount Morris the rainfalls are all low! theavera e at Henilock Lake from 1876-189.5, inclusive, being
inches, and i n 1851, only 94.36 inches. We have here three ears of 'exceedingly low rainfall, in which the run-off musi
gave also been exceedingly low. In 1895 the rainfall a1
Hemlock Lake was only 18.58 inches. The average precipi- tation at Avon and Mount Morris froiu 1891-1896. inclusive. was 30.19 inches. The fol- lowing are stat,ions at which it was much higher for the gears 1891-1895, inc!lusive: Le Roy, 45.95 inches, and Arcade, 41.6C
inches. These d.ateinents of precipitation in the Cfenesee Rirer
"water year," or the gear from December to November, inclusive. An interesting example of decrease of precipitation with increase of altitude is that found as we go west of the Mis- souri Biver. At Lincoln,. Nebr.. with an elevation of 1,647
feet above tide water, the average annual precipitation for eight years, from 1891-1598. inclusive, wa9 86.31 inches, the ran e being froin 40.08 inches in 1891 to 14.3s inches in 1895.
At f o r t Collins, Colo., with an elevation of 4,995 feet above tide water, the average annual precipitation froin 1891-1898.
inclusive, was 14.11 inches; in IS91 it was 17.50 inches and in
1893 there was 8: niininiiini of 7.U6 inches.
The figures for average precipitation in Nehmska and Coloraclo are based on a calendar year f roni January-Decem- ber, inclusive. The following are from Russell's Meteorology,* illust.rating Atlantic Coast, rainfalls, and are the averages derived from
eighteen years' observations-from 18'70-1H8P. The rainfalls are stated to be fairly representative for large districh of country around the laces.
tion above tide of 43 feet, while the afeinge annual rainfall is 5'7.1 inches. At Norfolk elevation of Weather Hureau is 57 feet above tide, and average rainfall is 51.7 inches. At Ros-
ton, Weather Bureau ofice is 135 feet nlove tide, while the aveinge rainfall is 46.8 inches.
. The following illuatinte the change as one goes north through
the Mississippi Valley. At New Orleans the Weather B U ~C ~U office is 54 feet abore tide, thd avera e rainfall 62.6 inches; at St. Louin Weather Bureau office 567 9 eet above tide, aveinge ininfall 3P.S inches: at St. Paul Weather Bureau office 850
feet above tide, average rainfall 98.9 inches.
The following illustizlte the Rocky Mountain Region. At Fort Girtnt, Ariz., elevation of M7eather Bureau is 4,833 feet;
average rainfall 15.8 inches; at Denver elevation of Weather
there is a very decided illcrease in ininfa B as one goes toward
27.56 inc TI es. In 1880 it was 91.99 inches; in 1x79,' 99.16
In 1895 it was only 85.05 inches.
. catchment area are all bmed on
At Jacksonville t Fl e Weather Bureau office is at an eleva-
'4Publiahed by United Statea Congm in 1901.
aMete6rology. By Thoinaa Russell, U. 8. Asst. Engineer, New York, 1895.
~ ~ ~ ~~~~~
Bureau 5,$00 feet; average rainfall 14.7 inches; at Fort Ben- ton, Mont., elevation 2,565 feet; average rainfall 13.2 inches.
The following illustrate the Yacitic Coast region. At Port- land elevation of Weather Bureau office is 157 feet; averae rainfall 50.3 inches; at Snn Francisco, elevation 153 feet; average. i-ainfall 83 inches; at San Diego, elevation 69 feet; average rainfall 10.2 inches. These figures abundantly support the Proposition that in the United States the rule of 'Increased precipitation, with higher altitude, is by no means universal. The writer can not say
positively, because he has not esamined the vast number of records with reference to this point, but he thinks it quite possible that the reverse is iiiore nearly true; that is, owing to distance from the ocean, prevailin direction of the wind, and other causes, it is quite probabe 9 that for the entire country precipitation decreases rather than increases with higher altitude. The decision of this question will de end to some extent upon the steepness of ascent. Thus on oiint Washin ton- summit, 6,8*3 feet-which isprojectedinto the air consi erahly above the surrounding uioiintains, the rainfall is about 83
inches. I n other cases, where the ascent is gradual, no illcrease
is ap arent. The same is also frequently true of sharp ascents. On %,g% Peak, in Colorado (elevation, 14,871 feet), the rain- fall in 1899 was 16.7 inches. Moreover, t,he writer has mostly avoided couipaintively eniall differences in rainfall-those not esceedin 2 to 9.5
inches. In such case8 the difference is too smal K to be any certain guide. Especially is thir true in the case of the North-
ern Plateau, where there is still a vreat lack of stations. 'The diflerences between high and low &itodes should be as much
ns 5 or 6 inches. Again, whether the excess of rainfall occurs in the winter or sunimer nionths must be taken into a,ccount. If it occurs in the suninier. even 8 inches of ininfall may not
iiiake niore than 0.1 or 0.9 inch in the stream. Rainfall and run-off observations are not yet, nor are they likely to ever be,
definite enough to take into account an annual difference of much less than about 1 to 1.5 inches. The writer has ceased to be excessively particular about the total of the annual rain- fall. Assuming souie considei-able length of record, mail errors have relatively slight effect. This matt.er is referred to here becmse nearly all rainfall recordsat any rate, in the United States-have iiiore or less error in them, and while it is desirable to have records as reliable as possible, a few errors
do not ngect a record very serionsly. It is nevertheless very desirable to know the history of a record in order to insure the
:le ree of confidence to be placed in it. ft seeiiis very clear to the writer that the substantial t1~t.h )I this question of increase of rainfall with increase of eleva- tion is contained in an article by Mr. Alexander in the Monthly Weather Review for January, 1 9 ~1 .' ~ According to this article,
I iiiouutaiii range does not. per se imply an incrense of rainfall. 3nly when other conditioiis are favoinble will such a result Follow from the presence of a mountain range. Mr. Alex- tnder says:
" Iu regions of high hunIidity comparatively low uiountaiiis
nay be important agents in bringing about rainfall, whereas
in regions of low hmiiidity very high- mountains may have little influence."
As to whether the writer would shade niountain areas to show higher precipitation than is given by the data for lower tdjacent areas will depend entirely upon conditions. At Mount Washington it would ol)viously be entirely proper to make such liatinction, while at Long's Peak it eeenis evident that the rain- $11 near the summit is no greater thaii in the adjacent plains.
It sometimes happens that the ininfall is greater on one side
d K
nThe Relation of Rainfall to Mountains. By W. H. Alexander, Ob- lerver, Weather Bureau. M. W. R., Jan., 1901, SXIX, pp. 6-8.
a39 MONTHLY WEATHER BEVIEW. APRIL, I909
Melci-8.
.300
f ~1 .1 0 0 ............................................................ l..~l.liOo.. 1,W2.o00 ...........................................................
................................................................ .............................................................. ?.
.........................................................
.. -- . .
of a mountain range than on the other. Where a fact of this character is well estctblished-and provided the scale of the map is large enough to do it intelligently-there would seem to be no reason why the rainfall should not be shaded on the side where it is greatest. As to whether a better method of presenting the data iN possible than the usual one of drawing isohyetals end shading the a r m between, the writer is not very clear escept on one point, namely. that the topographical elevation should a.p ear upon the map in connection with every rainfall, as wit B out this, from an engineering point of view n t any rate, the niap ha8 very little significance.
roposition, proximity to the ocean and direc-
heavy rainfalls without specinl reference to the elevation. Proximity to the oc.ean, however, is not a universal cause, as
may be seen on reference to the rainfall records at Sail Fran- cisco and San Diego, Cel. I n England, where on the west coast these >henomena are more uniform, the winds are largely
the ocean to he condense$ on the mountain ranges of from
8,500 to 3,0(@ feet in height, and here the precipitation is from SO to 150 inches. On the east coast, however, even quite near the sea, there is a flat, level rountry, and the rainfall averages from 35 to 30 inches.
As a broad tion of prevai F ing wind8 are the more important reasons for
from t h e southwest, brin i n g the moist warni air currents of
I
i t W i i i i c l i w .
3?Y
'234
mi
: :w3
; 1 ss9
....
(2s) PROF. A. WOEIKOFF.
............................................ Amboina.
Do ................................................ Bernm ................................................ Do ................................................
In the Meteorologische Zeitschrift for 1P85, Volume SS,
pp. 113-138; 801-811; 85ib-363, Prof. A. Woeikoff paldishes
an elaborate stucly of the rainfall in the Malayan archipelago.
We extract the following items relative to the relation of the
rainfall to prevailing winds and elevations.
Page 116: The rainfall on the north and south coash of Amboina and Seram differs ac!cording to the direction of the wind. The! prevailing winds for the four months Decenilwr- March are northeast or northerly, but during the four months May-August they are southwest orsoutherly. The distribution of rain is shown by the following table, according to which the south coast has inore rain than the north coast during south- erly winds but less than the north const during northerly winds. This illustrates ~t general principle that holds good the world over, namely, that the sex breezes bring more rain than the land breezes.
1-
; 111111.
South.. __I 821 North ..... I 6'26 Sooth.. ._.I ,589
North I 1.095
.
.....
Suutlierly rind. Mag tU AllgllNt.
Il1911.
?.lm 1.209
1, Go5 .I*
--
Page 119: The windward side of the Celebes receives more rain than the leeward side in the ixtio of lti5 to 100.
Page 147: The north coast of Java in the region of Batiavia represents a comparatively flat country whose ascent in quite gentle up to the mountains of the interior, which rise rather
suddenly. The increase of rainfall as one goes from the const inward (viz. southward) must therefore be the result of a
very gentle geneirtl rise of the currents of air. This increase
is shown by the records of stations at inwearing distances from the coast, as shown in the following table:
Name of station. Distance Altitudeof Annual
I from cosst.1 stution. I rainfall.
Onrust.. ..............................................
219
Ratavis.. Meester Cornelia.. ...................................
301 116 367
........................................... '265 4s
P w w Ningo De k Bd;ong Gedeh.. Bul tenmrg
.........................................
.....................................
Page 131: If we group the stations of long eriod by alti-
aspects, we get no decided increase or decrease with altitude, or, more exactly, the annual rainfall diniinishes up to l,%NJ
meters Wncl then again above that increases; of course these annual aveirrges include in general both the local thunder- storm rains nnd the geneinl rains due to broad and deep mon- soon winds.
tude, wit,hout special regard to the windwar cr and leeward
The tigures are as follows:
Zones of zaltitude.
( 24 ) GIFFORD PINCHOT.
In a letter of February 14,1909, Mr. Gitford Pinchot, For ester, U. 8. L)epartnient of Agricnlture. says:
I find myself unable to reach an opinion as to the ext,ent to which mountains and ridges from which no observations are available should be slpded to indicate heavier rainfall than from adjac.ent lowlands. While, in niy opinion, such heaviw rainfall undoubtedly exists, it, mould be extreiiiely difficult to reach a definite opinion tis to its amount.
Tour third question with regard t.0 the estent bo which ti
rule that applies to one momit.ain can be estended to other mountains 111 different parts of the count,ry is likewise ex-
tremely ditlicult to answer. Local conditions would govern to such an estent that perhaps not even a very general rule could be reached. Altit,iide and ~noisture of the climate, together with prevniling winds, would be among the factors to be considered. Pour fourth question is one which, in my opinion, should be answered a5rmativsly. I believe that there is a sufficient connection between forests and rainfall to warrant attributing heavier rainfall to a forest region, tw com ared to an adjacent hare region in either one of two cases. &st, when the irrin-
APBIL. 1908. MONTHLY WEATHER REVIEW. 33s
fall causes the presence of the forest, and,.iecond, when the forest, as I believe it does to some extent, increases the pre- cipitation. I am well aware that observations are contraclic- tory upon this last pint. There are, however, as 1 think, strong a p7Gw1: reasons for believing that forests influence rainfall, and there are cases, ap rently well estublished his-
of greet regions have changed coincidently with t le destruc- tion of great forests. This whole question being as yet, how- ever, i n t.he realm of coiitrowrsy, I have rather avoided it,s discussion hitherto, believing that the nrgunient for forest preservation should be hased, for tlie present at least, on less controversial grounds.
P
torimlly, where the agricultura p" conditions and ossibilities
(25 ) PROF. VICTOR KREIUSER.
The German engineers officially charged wit,h the study of
the regimen of the larger rivers, the prediction of floods, &.,
have published elaborate voliinies of text and charts cleRling
with iiinfall and whatever can nfect river phenomena. From
pages 54 to 56 of the first volume on the Elbe, 13erlin, 186M. we
quote the following, which is believed to have been written
by Prof. V. Kremser:
On t.he windward side of n niountain range the ninsses of air must rise, cool, and deposit the excess of nioist.ure on the land. Thus the mountain ranges and even slight. elevations of the earth have heavier rainfall than the plains. and blius in
general with increasing altitude the quantity of preci itation
far froni the ocem we find that a representation of t.he vnrt.ica1 elevations in the valley of the Elhe is in peneiril also a picbure of the distribution of rainfall. But this is t,rue only in t,hc niost eneixl sense. since a niore accurate considcrat,ion lwings
to lig f it many niodific~tions. Since the nioist air coming from the ocean loses its wut,er either gradually or suddenly, but. does not obtain niuch in return from the evaporation from blie soil, therefore t,he rainfall diminishes as the air penet,rate.s into tlie interior and at amy given altitude the iiioiiiitaiii ranges hare less i~iinfall wit.11 increasing oontinentdity. Fiirtherniore, tlie precipitation iniist be heavier 011 the iiioist windward. iisunlly the west side of the mountain range, than on the lee side, sinco after passing over the suniniit of tlie niountain t,here is no
further caiise for condensation. On the lee side, where the air descends in the shadow of the rain, a drier region niuat form. This region must, est,end h r beyond the nioiint~sdn range, even though the tendency to a cle~cending niot,ion of the air giwlimlly disappears. These general remarks nre abnnc1ant.lj- veritied 1~5' t.he charts.
The greatest rainfall occurs on the to s of the mountains; the least rainfdl at the foot on the lee H i x e, n41ence it. iucrea.iles slowly with increasing distance t,o leeward. On the average of ad our mountain ranges the annual rainfall increases with altitude a t the ratme of 70-80 iiini. of rain per 100 meters of
ascent. That the windward aide has more iltin tllan tlie lee side i n
the I>asin of the Elbe can be seen best in the Erxgehirge, where t,he sout.heast side is to leeward and the northwest, side
is to windward, as shown in the following tnble:
increase$. If, therefore, me consider :i neighborhooc / that is
~. .
1 xorttirvest slope. I ~o u t ~i e a s t slope.
I- I- lk-' I- l-l- Xctcr8. Mdri-8. Mtit. .Iff .Irra .Iftll.
...................... ?00-400 4oo-fm ...................... i E l 6w-m ...................... 700 8% 708 730
The fact that for the same altitiide the precipitation dimin- ishes tls we ao inward from the ocean is shown in the following
table, in wkch, in order to compare si~~iilar conditions, the windmaid and leeward sides are tabulated separately:
Leeward nidc. altitude
2MHGQ metem.
Mountain mnge. cr;; I $zl2
stu- precipita
1
tiona. i I
tion.
Mrttw. Mm. Hnrz ...................... 7
1
243
1
M6
Tliiirin iun forM ........ 5 L75 610 Erzgebfige (s o u t h ens t side). ................... ..........
_-
Windward side. altitude
Leeward side. altitude
mm metem. I Distances
---I-l-
For greater altitudes the number of stations is so sniall that tlie result* are too greatly nioditied hy local c.ondit.ions and a
fair comparison is not possible.
[NOTE.-we have added a last colunin in thc above table, showing approsinlately t,he distances of the respective moun- tnins from bhe mouth of the Elbe, or the nearest point on the North Sea coast, ahove which t,he northwest winds may be supposed to be slowly rising as t.hey flow toward the southeast, carrying the ocean vapor into the highlands of Moravia nnd Hohemi~.-Ed.]
(26) PROF. =TON WHITNEY.
Under clnte of May 17, 1909. Prof. Milton Whitney, Chief
of Bureau of Soils, Del~rhient. of Agriculture, writes
follows:
Your inquiry of March 1s. relatii?g to the possibility of mapping the rainfall over ;ireas in which there are no rainfdl &tioils t.)y the character of t,he foreat wr0wt.h has heen received and htis been under consideration For some time. I
do not, Mieve that the chairacter of such growfh is any safe indicstion, except in the n~ost~ general may, to the aniouiit of rainfall. nor can I claim any iiiore npecitic distrihution of for- e& in trccordnace with soil types. As a rule we espeect to find hard-wood forests 011 heavy well-drained loam and clay. Where this occurs in level or gently sloping areas, and even
on niountnin slopes, hard-wood trees seein to prefer the heavier tvpes of soil. We expect to find the pine on the light, sandy soils of the Atlantic and Gulf coasts as well as in the lake regions and the lighter soils of t'he mount.ain areas. R e find t,he chest,tint estensively developed on the sandstone or shale ridges with oak, hickory, and maple in the heavier valley Innds of the north A t h t i c States. We find pine trees on
niany heavy intractable clay soils, such as the Potoniac clays of Maryland and Virginia, and also on many swamp areas. There are, however, so niany departures from the general rules that have Ireen laid down that i t is not safe to use the forest rowth in mapping soils any iiiore than I believe that it wou B d be safe to use the characteristics of the forest growth iii ma )ping the distribution of rninfall. A few examples will
forcib \ y illustmte IUS position.
-
It is a well-known fact that (on the Piedmont areas of the Atlantic coast States) on what is known ai Cecil clay, which is the heaviest and most productive soil, the native forest growth is of hard wood. After this growth has been removed and the soih cultivated for a few. years under superticia1 methods, the soils deteriorate and are allowed to grow up in old field pine with a spfinkling of stunted oaks, so that now, as a mat- ter of fact, much of this land which was formerly in hard wood ir now in pine. We also have the admiinhle records of the German foresters who have pointed out the succession of trees. This is mentioned quite fully by Storer. In our soil survey in Allegan County, Mich., last season, we mapped I large area of Alle an sand. We could see no differ-
yet in this case we were conipelled to recognize differences on
account of the differeme in the native vegetation and in the agricultural value of the land. These differeaces were not indicated in our soil map, but were il1usti:ited on a colored plate which will appear in our next re ort. C!ertclin rea as of
wood forests. These are now recognized as some of the best peach soils of the State. In the northern art of the area
was ori inally covered with a mixed growth of hard wood and pine. Fhese a r m are now recognized RX fairly good peach soils. Both of thebe areas have the sanie relative position as
regards the lake and are presumal>lg under identically the same characteristic conditions. Another portion of the snnd upon which white pine originally grew has not been developcd up to the present time. Some few clearings are found i n
which corn? rye, buckwheat, and n few peaches :ire grown with very inditferent success. On the pine plains prQper, for- merly covered by a sparse growth of white ine and at yres-
entirely undeveloped and only occasionally sniall crops of corn and rye are produced, with 110 tinanc!ial success whatever. I n the Southern States there are many areas in fine hard-
wood forests where it would be expected from the soil to find pine, and vice versa. The most striking example, however, of the apparent acci- dental distribution of forest rowth is in the native and un-
cleared foreah of Florida. $here is no difference, so far as
can be determined, between the soils and climatic conditions of the hainnlock hnds, which support a vigorous hard-wood growth. and the first quality of high pine land, which sup- por~q a very dense growth of long leaf pine timber. 1llusti-a- tions are given of these two characters of growth in Bulletin 13 of this Bureau, A Preliniinarv Report on the Soils of Florida, Washington, 1898. Furthermore, there is no differ- ence, so far as can be determined, in bhe soil or cliiiiritk con- ditions of the second quality of high pine land and what is known as the "scrub." Pet the contrast between t.hcse two kinds of native growth is very sharp and exceedingly striking. An illustration is iven of this in Plate 5 of Bulletin 13. The soil is B light smf, and on the high pine. land supports a vig- orous and generous growth of large pine trees, which, when cleared, makes fine truck soils, and produces the tender vege- tables in the greatest luxuriance for shipment t.o the Northern
markets. I n the scrub there is a dense growth of scrub oaks and low bushes and plants, irrrelg exceedin the height of LL nian's head,
evaporation by the pro ierty that desert plsnts have of cover-
the leaves up edgewise, to ex ose as little of the surface as possible to the sun's rays. & giws is found, and only the most hardy desert plants grow. When inw grow, it is an
ence in the mechanical or c % emical character of the soil, and
this sand near the lake were origina 7 ly covered with liard-
surveyed-that is, north of the Kalamazoo !i iver-this sand
ent with ti sparse growth of scrub oaks, t E e soil is alniost
all having thick leaves. protecte f finom the loss of water by
ing the surface of the 3. eaves with an enaniel and of turning
occasional dwarf spruce pine, and not t ! e long-leafed pine
found on high pine land. A few efforts to grow truck and oranges have been failures.
I would also call your attention to the curious fact of the. occurrence of deadenings which are occasionally found, and to prairie. conditions and barrens which are found in many places in the Atlantic coast and Middle West where small or large areas, surrounded by luxuriant forests, have been with- out tree growth within historic times, with no reason which can be ascribed either to soil or cliniatic conditions. Many other illustrations of the kind here recorded could be cit.ed to oaition that the native forest growth can not be safely prove use as a busis for mapping rainfall conditions, although in a general way such conditions may be useful in climatic conditions as they are in soil investigations, but the limitations to this inust be clcarly recognized. Professor Hilgard has called attention in the Tenth Census to the possible use of forest growth in the classification of lands accordin to their chemical composition and agricultural
general application. Our own investigations appear to indi- cate that by the constant growth for inany years of certain native phiits (or cultivated plants, for that matter), changes occur in the relative uiiiount of plant foods in the soil, so that the pro er hlnnce or ration, to borrow a phrase from nninml hushan (r ry, is no longer iuaintained' and the soil is suited to other crops and to other forest growth. This, however, is yet but an hypothesis, as our methods of chemical and physical investigation are not siificiently refined to enal)le us to inves-
t.ipte the matter at the present t h e .
value, but sui? B rules as are Inid down by him are of the most
(27r GEORGE E. CURTIS.
In :LSM, in order to study the relation hetwe.en ascending
currenk and rainfall and to investigate the large ininfall on
the sunitnit of Mount Washingt,on, t,he editor requested thdt
a large number of rain gauges might be distrilmted to rolun-
hry observers at as inany poin'ts as possilde in the neighbor-
hood of that niountnin. A pear Inter, when we came to study
the records, it was disc0rere.d that a 3-inch gnuge of objection-
nhle construction had been issued instead of the standard 8-inch
gauge that was originally contemplated, and this change
operated to discouinge further investigation. But with the
hope of settling some points of interest, iny nssiatnnt. Mr.
George E. Curtis, made a study of the effect of the wind nnd
the distribution of win 'at the very summit of the mountain
and enibodied his results in Signal Service Noh No. XVI,
The Effect of Wind Currents on Rainfall, Wrdhigton, 1PS4.
I quote or siiniinarize the following paragraphs relating to the
ohservations iuade between Septeniher 1,18YJ, and October 1,
1883, considering only the minf:dl, excluding t,he snow that
fell during that yew.
After a slight historical su~iimary Mr. Curtis quotes the
following recapitulstion by Synions (see his Meteorological
Mapdine for 1878):
The terribly hard-fought battle rerspecti?g the reason that a snialler qnantity of inin is collected hy rain $uge.l elevated above the surface of the ground is newly sett ed, for the fol- lowing poinh seem proved:
The greater part of the decrease is due to wind. The stronger the wind the greater tho decrmwe with eleva- tion. The less the diameter of the elevated gauge the less will it indicate.
APB~L, 1902. MONTHLY WEATHER REVIEW. 235
ICE ....................................................... E ......................................................... SE ....................................................... B ......................................................... B W ....................................................... \v .. 1. .................................................... N W ...................................................... N ........................................................
Totail.. .............................................
A gauge on the leeward side of a tower may collect as much rain as one on the g.roiind. A gauge in the iuiddle of a large roof may, notwithstanding its hei ht;collect very nearly the same as the one upon the grounf We now come to the second and larger qiiebtion, namely, What variation in the distribution of i-ainfall due to wind is produced by the to o!qmpyhy of the country B The ptactical problems have alredy been stated, namely, (1) to find over what area the recoid of a single gauge gives the avei-age pre- cipitation; and (8) to deteriiiine from gauges at ditkrent ski-
tions a law of variation, so that, given the observed precipita- tion at a few stations, the precipit,ation for any other localitmy to which the law applies can be computed. This is piwticahle only for those countries and districts where, the iainftrll being dependent on regular and constant winds, the variations in precipitation are progressive and miiforiii.
Among the effects of the topography of the count,ry on the distribution of rainfall, the variations in niount:titi districts are edpecially noticeable, and it is to these we shall contine our
attention. In general, the amount of rain increases with thc elevation above sea level up to a iunsiiiiuiii plane, aft,er which
a decrease takes place. 6. A. Hill has shown (see Zeitschrift fiir Meteorologie, Vol. S I V , 1Si9) t.hat in the northwest Himalayas, where the ixinfall is most reniarkablo in amount
and inte of variation. t.he observed re1:rtive annual anio1int.a can he represented by the following empirical forniiila:
R = 1 + 1.Nh - 0.40hS + 0.08hJ
in which /c is the absolute height in units of l ,W O feet above an assumed plane, which is itself 1,000 feet above sea level. From this formula the height of masiniuiii rainfall is com- puted to be 3,160 feet above the plane, or 4,160 feet above sea level. It is further shown that this elevation is that at
which, according to the observed law of decrease of teiiipera- ture, the southwest monsoon is cooled just below its dew mint.
expect the muxiniuui precipitation to take place. The result obt,ained from the enipirical foriiiula thus receives a theoret- ical and deduvtive contirniation. In the following ~oliiiiir. of tho Zeitschrift (Vol. SV, IFYO, p. 37'3) Hatin has collected ohservations .from stations on the Arlberg, in the western Tyrol, by which the )lane of niasiiiiuin ilainfall is shown bo
The highest station, a t an elewtion of 5,9W feet, has the niaxiniuni irtinfall. A very rapid diminution takes place on
the leeward side. where the stations record only about half the aiiiount of rainfall given by stations of eqiial altitude on the windward side. The Report 11 on the Rainfall of Barbados, by Governor
tihows a siniilur variation wit.11 altitude; the aniount of lain
. increases regularly with the height of the station, except in a few localities where the law is masked I>y ot.her locd causes
of variation. * * * The 1:rrge rainfall recorded atb the signal service sbation on
' .the sunimit of Mount 1Tashingt.on has seemed. particularly
worthy of careful study and E ecinl observations have been
, . undertaken. The first point orinvestigation has been relative to the distribution of rain on the autiitiiit itself, since it seemed probable that the enormous wind velocit,y experienced there might produce sensible inequalities, both in the gage reaclinge and in the actual rainfall, wherefore the records of a single .gage would not present avei-age results for the whole summit. From September 1,18P9, to October 1.1883, comparative oh- servations were made daily with four extra gages. placed 75 feet north, south, east, and west of the station gauge. These estra gauges were cylindrical. 14 inches deep. and surmounted by a small receiving cup 8 inches in diameter. The station gauge was 8 inches in diameter and S feet deep.
This p i n t will be that at which, in the mean, we s b ould
be at a soliiewhitt hig h er level in t,h:rt region thsti in India.
Rawsou, giving t YI e results of observations from 1547 to 1S74,
1.4s 0 18 1.54 4.33 3 .9 4.92
20.15
1.23
45. S? --
As the observer reported that the meamrement of snow is altogether unreliable for comparative purposes, owing to the velocity of the wind, the observations used in the following discussion are those of rainfall only. The result of this series of observations is given in the following Table I, containing the rainfall only during the thirteen months of observation; the records of each iige are first tabulated according to the direction of the winf?
TABLE I .-h i n f d , rxclztd~irg $itour, &ptenikr, 1882-&plentkr, 1888, inclu-
8iw, on Mount Wiihiiiylm.
-~
I A. I B. I C. I D.
1.31 ' 1.47 0.15 0.18 2.69 1.87 a.% 6.91
4.20 , 3 .~5
5.13 4 .E
'%.?3 26.31 l.'W 1 0.79
1.43 0.16 1.65 6.m
3.71 3.70 'A.%3
0.65
NE ......................... 0.03 E ____.._____________________ .oo eE .......................... .04 8 __________.____..______..____ .14
51.19 I 48.40 1 42.63
S W ......................... 0.08
w ___..-: ___________________ .IO N W ........................ .59 N .__________________________ .@2
836 MONTHLY WEATHER REVIEW. APRIL, 1902
as to be as nearly as possible under the mnie influences. I n this series of obserrations the differences due to location and environment are as far ltrs possible eliminated, and a direct com iison of the two ga es in collecting rain can be secured.
the 3-inch gage readings is l+O.O15 x 2, where 12 is the wind velocity in ani& of 10 miles er hour.
varies, therefore, directly as the square of the velocity of the
. wind, and is due to an insufficient collection by the snialler
y means of this formula the totala of the small ages given
the station gage.
, The average velocity of the wind during rain was, tis nearly as can be determined, about 45 miles per hour. For this wind velocity the formula ves 30 per cent as the factor by which
correspond to the station record. The totals of the gauges
will then be: A 59.57, B 66.55, C 59.02, D 55.43, stat.ion 58.70. * * *
* p* The ratio of t % e %inch gage readings divicled Lq'
The discrepancy between t K e meawrements of the two gages
g a p .
in Table I can now be rendered coniprable wit,h t % e t,otals of
the totnls of the 3-inc c gauges niust be increased in order to
(28) PROF. HARK W. HARRINOTON.
In his memoir on '' Forest influences," published 11s Bulle-
tin No. 7 of the Forestry Division of the Depart,nient of Agri-
culture, Washington, 1893, Prof. M. W. ,Hnrrington cliscurses
a variety of forest influences. Pages 111-11s are devoted to
the question of precipihtion over wooded and treeless clis-
tricts. He says:
When a dift'erence of rainfall corresponding to a differenc*e of forest conditions has been found, there is still occlusion for doubt as to which is cause and which effect. There is every reason to believe that with increased rainfall. other things being favorable, there will be an increased growth of trees. The facts at hand do not prove with entire conviction thnt forests increase the rainfall. ' The historical uiethocl is Incking generally in the character of t,he data for the heginning of the comparison. Besides, where a change of rainfall is itctnally shown to be coincident with a change in the foreNt growth it, is not entirelg ccrtnin that the former is due to the latter; it, may be due to what are called secular changes of the -rainfall, the reasons for which lie beyond our hornledge. The geo- graphical method is not entirely satisfactory. for the reasons The entirely convincing method depends
on observations above forests and with systems of radix1 stit-
tions, arJ proposed by Dr. Lorenz-Liburnan. and froni these there is not yet a sufficient atmount of published results.
'In a subsequent portion of bulletin No. 7 the present. editor
was able to *how that, owing to-the large effect of the wind a t
the mouth of the gauge, in causing a deficiency in the cat'oh
of rain, no ohservations have as yet been ninde sufficiently free
from this error to allow of their giviiig accurate results when
comparing the rainfall above tlie forests wit,li that beside them.
After reviewing all the accessible and best literature on the
subject Professor Harrington leaves the niat,ter entirely
unsettled, showing that every attempt bo demonstmte t.li:tt
forests tend to increase the rainfall has been unsuccessful and
that their influence is entirely inappreciable. Of c:ourse,
every portion of the land and ocean, by evaporation, contri-
butes something to the moisture and, therefore, to the rain,
. but evaporation from forests is less than from an equal area
of cultivated fields, and is scarcely distinguishahle from the
numerous other sources of moisture.
* already mentioned.
As to the influence of rainfall in deciding the esistence and
the character of the forests and the possibility of inferring
froni the forest,# the quantity of rain it is evident that the
growt,h of the forests, like that of any other vegetat,ion,
depends upon the supply of water accessible to the roots of
the plants. The surface run-off quickly becomes inaccessible.
It is the quantity retained by the soil, and, therefore, the
character of the soil itself, that is the most important factor.
With reference to this point Prof. B. E. Fernow, in the
nbo\-e-mentioned Bulletin No. 7 on " Forest, influences,?' on
page 144, Bays:
It is the water retained in the capillary interstices of the soil that deteniiines the designation of the soil as nioist, wet.,
or dry. Any siirplurs al>ove the greatest water capacity is I,onnd to drain off. The measurements of the c unntity of
waiter citpwities, both of which vary rently wibh the depth
laries show t,he greatest wster cn acity. The least retentive
of its Wright in mater: loarr~y soil, 40 per rent; pure c ny soil,
'in per cent. The sandy soils of the nort,h Cicrnitin plain can not hy capillarity raise the ground water higher t,han 1.5 feet above the wat.er plane below t,he ground. so t.hat the siir-
face strata over regions where the water plane is 2 feet below t,he siirfttco do not show an? greater amount of water than t.he surface soils in other re ions. Mr. F. H. King has shown
enough water, and t.he same principles mist apply to the dcterniination of t,he possibility of the forest growth. When a forest is once well started, it accumulates a forest litter t,hat is retent,i\.e of the litt.le water that reaches it.
water absorbed by soils will show the iiiasinium an a niininiiini
of t.he soil. Hiiiii~is and garden ill01 8 with their fine capil-
soil is a coarse quartz nand. Sanc 7 y soil may hold 55 ~e r cent
\ -"
tlie conditions under whic K I surface soils are unable to supply
(2s) PROF. O. HELLHANN.
In coinpiling his rninftill chart for East Prussia (Berlin,
~:WO), Prof. Dr. G. Hellninnii (one of the highest authorities
in relation to rai1ifiill and ii% nis:isurement) deals with the
record for the ten years 1859-1898 and with :I region of
Rhotit 500 square miles, or six times the size of the District
of Colunibia. Within t,his area he has 17s rainfall stations,
67 of which have complete records for the whole pericJc1 and
the 111 others have generally completed nine years of record.
The latter records are all reduced to lioinogeneity with the
complete decennium. 18d9-1S$+S, by means of factors derived
from the yeax- for which they have records in common. Dr.
Hellniann says:
The precipitation for Pillau is too small, on account of the too free espoaure of the gauge to the wind, for in proportion
nr a i-ain gangc is exposed to the disturbing influence of the wind SO much the less does it catch.
The highest of the East Prussian stations has nn elevation
of 935 meters. It is of course at some dishnce east of the
shore of the Baltic, but it is not the one that shows the heavi-
eat rainfall, as there are quite a number of low-lying stat.ions
that receive more rain.
The accompanying rainfall chart for the province of Emt Prussia is prepared in a1vord:inc.e with the tnbulated valuee
of the rainfall and with continued reference to the topographic condit,ione. It shows the distribution of the mean annual precipitation in five different gradations of 50 mm. each. The
'
On this point Hellniann says:
h, 1902. MONTHLY WEATHER REVIEW. 237
cipal detail features; the present.ation of orogra h i c cGnrli- tions on a chart of this sr:ile woiild have only very I ittle v:ilue. I Even all the rainfall stations van not be charted 011 this sllliill -- ---..
close dependence of the rainfall on the orography would, oj
course, he best brought out if a contour map could have beer
used in exhibiting the rainfall, but the technical difficulties ir accomplishing this are especially great in chails drawn on 2
Notwithstanding, anyone acquainted with tht relief of the country will at once recognize that even in t
region that has such slight variations of altitude as Eas~ Prussia, still the rainfall chart is to a certain degree a reflec. tion of the relief chart. The present chart of rainfall there,
fore acquires special interest, bec‘ause it brings out clearly tht fact that, even in a flat country, small elevations exert a de cided influence on the quantity and the distribution of t8ht rainfall.
Hellinann shows that, according to his rainfall chart, t h
precipitation increases rapidly from the coast inward for ti
distance of thirty miles: then comes u relatively dry region,
after which rainfall again increases along a region about c
hundred miles from the coast. So that, in general. ”the nioisl
air coming from the Baltic is forced to rise, whereby it, i:
cooled and gives more abundant rainfall tlian it would :it tlit
same elevation farther hack from the ocean.”
As regards the question how long a series of ohservatiotir
is required in order to ohtsin nornial vuliies of the rainfsll,
Hellmann studies t,he records for fifty-one years. 184s-1898, ;if
Tilsit, Kfinigsherg, and Klaussen,:ind the records for t~heshortt~i
periods at twenty other stations near l q ~. In geneid. thk
data seeni to show that the average for the last ten year?
of the fifty is uniformly a deficit as coiiipared with t,he wholt
fifty years. The greatest deficit of 10 per cent is in the south
eastern part of the province, and t,he least. deficit, zero. i s ai
Tilait and Memel, in the northexti portion. The annual values
of the rainfall depart from the mean values hy 45 per cenl
above and below. In a hundred gears of recorda 10 or 12
per cent will be very clry, namely. having only 50 or 55 per
cent of the nornial riinfall, and 3 or 4 per cent of the years will
he very wet, having the mine per cent in excess of the nortiial.“
. small scale.
the scale of the maps ordinarily puhlished in the Monthly
IVe,at.hpr He.view. These, ul;ips sl1ow the. river systems, wlie,nce
(80) PROF. PAUL SCHREIBER.
An elaborate study of the i-ainfall in Saxony liw heen made
by Prof. Paul Schreilier, of Chemnit~z, and from R suniaiar~
of his work by Gravelius (Zeitschrift fiir Gewiiaserkunde, Hd.
111, 1900, p. 48) we iiiake the following smii~iiary :
One hutidred and sixty-nine stations are available for t h e rainfall chart, of which 7X, or #2 per cent. represent, fng-
ments of the fundanlental ten year^, lSSt$-lS$& nncl innst, therefore be reduced to homogeneity with the remaining $11.
This proportion may seein large, hiit is leas than the CjS per cent of corrected stations used by Hehiatin in his chart of
ewt Prussia. . * * * The isohyet& were first, drawn on a chart of Sasony, on the scale of 1 to 600,00!~ and the lines thus
obtained were subsequently reduced to one-half of this scale.
On this chart only the network of rivers is shown in i t s iirin-
scale (the area of Sasonv is about 9,700 square miles). But in the higher regions of the nioiintains, or the sources of our rivers, relatively more stations can be shown then in the low- lands. The charts and the tables show how the distribution of the annual rainfall follow? the orographic condit.ions of the surface and that the higher regions of land in general corre- s iond with greater rainfall. This connection is especially
heavy rainfall within regions of smaller precipitation. * * * The study of the vhart suggests that i n certain regions where the soil is wet and tends to he swanipv the evaporation might play an imporhnt. rijle in producing these islands of rainfall,
hiit t.his assuniption is not, justified, for the rainfalls by
montrhs show t,liat the heavy mitifalls i n these islalids do not
orciir i n the nionths of greatest evaporation, viz. June to Angust, so that ti antmisftictory esp1an:it~ion of their occurrence
m i not now he given. Although the chart does in geneid show t,he correctness of bhe genein.1 rule that up to a certain height. which is not at- t,ainecl in dasony. the general rainfall increases with altitude, still it also ahown t,hat tliere are ini )ortatit departures from the
t,be only factor that regnlat,en the ilist,rihution of lain. An equally in~port~ant considerntiort is the location of t.he stat.ion 1vit.h reference to the ririn wind. Of t,wo stations having the
saine tiltmitilde bhnt on t,he winc1m:rrd side will have the heavier nnd that on the leeward side the .ciinaller rainfall. Moreover. sninll elevtitions that incwnse the attniial rainfall npon them- selves t,hrow a rain shado\v over the region to the leeward. The west wind is i n $osony t,he preniilitig rain wind, and the southwest wind conies next, to it, and nest to that the north- west,. But if we consider special niont,hs, such as the sunitner time, we find that the iniport.:ince of the wind direction changes, and that the northwest is often most important but geneially second in iniport,ance,, and that the north wind often beconies important.
This retninds us of the experience st Mount TVashington,
where t,he soiibheast and southwest winds bring most rain to
the lowlands, but the nort.hmest wind to the suriiniit of the,
tiioun tnin.
Finally, Graveliua and Schreiber show thttt isolated high
;rt.ationr! behave like rainhll ganges that are locat,ed too high or
qiosed too freely to t,he wind, viz. they give a relative deficit
D f rain.
c I early shown in those portions that stand out as islands of
rule. We see, in fmt. that the :I I titucle ahove the sea is not
(81) PROF. ALFRED ANGOT.
The distribution of rainfall in France and western Europe
ias heen especidly studied ~J J - Alfred Angot. His iiieuioirs in
.he Annii1e.a c h i Bnre:iu C‘cwtral M6t6orologique; “On the
riinfnll of the Tlwriiiti Peiiinsiihi” and “On thta rainfall of
western Europe” give full consiclerntion to the niethods of
ireparing rainfall charts. The monthly, annual, and seasonal
+arts for wedern Europe. sc~vent~een in nll, tire on about the
ianie scale as that of the United States daily niorning weather
nap and as that of the set of cliniatologicnl maps published
3y the Vnited States Weather B~irea~i. tind. therefore, twice
31-11
'438 MONTHLY WEATHER REVIEW. Amn, 1909
- -
Biiiuollid raliua wlcttitu to &iii ,WJarlicrir.
Beyonne ........................................ Amgurri.. ....................................... Bilhao ...........................................
arms ruinJ-c11l. lMl-1.wn.
Bayonne ........................................ Aragorri.. ....................................... Bilbtio ...........................................
Prn bddr dqitli qf rainfdl at Sail LWmliwt a#
~ulllplllnl pw 1.w-1s.w.
Xayonne ........................................ Amgorri.. ....................................... BiltJtao ...........................................
rainfall and altitude is to have the relief chart or contour chart
printed on transparent paper on precisely the same projection
as the rainfall chart, and so superposed that one n i y , at, n
glance, appreciate the relationships.
Beginning with Spain and Yoi%ugd, Angot proposed to col-
lectdata for the fundamental interval of thirty years.lM1-1890.
for each European state, as he recognized that it is inipossilde
to prepare monthly iainfall charts with satisfactory precision
unless all the data relate to the same years. For t,he Iberian
Peninsula he gives 98 statione, but 11 of these mere for periods
of less than three years and were not used; 37 mere reduced
froni periods of twenty-five years or more up to the funda-
mental thirty.
In his memoir, On the Rainfall of the Iberian Peuinsuln,
(Annals of Bur. Cent. 1893) Angot says:
It is necessary that over the whole region the observations should relate tc) precisely the saiiie years, so that olie may avoid comparing a series of relatively dry years observed at one station with relatively wet yetirs at another, therehy fal- sifying the relations between the lilvionietric regimes of
the two conditions of uniformity as to years of observation and .freedom from breaks due to the changes in the location of the apparatus and changes or irregularities in the distribution of the rainfall froni yerir to year and month to mont.h. I n order to compute the iiionthly and anuual totals of rain- fall for the eriod 1581-1890 at stations where the observations
When tge gaps relat,e only to isolated months the probable amount of rainfall during these months has been conipubed from that at neighboring stations. For this purpose we have assumed that the aniounts Qf water collected at these stations preserve the same relation among themselves. This 1aw.which is frequent1 applied, appears to have been first foniiulated by Mr. V. Journie, engineer of roads and bridges. in 1864. If the gaps relate only to a single decennial period,lY61-18TO. 1871-1580, or 18S1-15HO, the two complete periods have keen retained and on1 the mean of the incomplete period has been
Finally, if the observations coinprise less t,linn twenty years, the general mean of the whole period 1861-18DO has been com- puted by means of a cornparison with at least three statmion* in
the neighborhood of the station under consider:ttion. An cs-
am le will illustrate this method of calculation. $e will take the dation of San Sehastiaii; the first i n geo graphical order in our tables. This station fui.nishes us with only thirteen years of ol>servat.ions. viz, from 1S7H to 1SW. In order to obtain the means of the period Mtil-lSHc~ t,he three stations, Rayonne and Aragorri, in France. and Uilhao,
in Spain. whose observations coniprise thirty years, have IJecn com ared. R e begin by calcidating for the four stations t,he
eriod of thirteen years, and then fiiid t,lie iwt,ios of these nuni- Iers. The follow~ng values were thus ob!ained:
these two points. Very few publishe B works on rainfall satisfy
are incom P ete, we have proceeded in the following uianner:
interpolated in t I e manner esplained further on.
toh P amount of min collected each month during the coniinon
Jnniiary
-
1.209 0.659
0.997
93.3 It;?. 1 1W. 7
109.2
106. H IUi 4
1.253 O.&W
1.017
April.
1.218 0.m
1.015
M l l .
I , 96Y
1.724 2.
1.747
1.1.12 0. w25
1. Ei
75.8
95.7
N i .4
As the ratios for a particular moiith ma be quite largely influenced by heavy ixinfalls unequally B istributed at the
f
respective stations, these ratios have been smoothed hy takin for any one month the mean of the ratio proper to this mont and the half suni of the ratios corresponding to the months >receding and following. Thus, for the ratio San Sehastiani h~yonne in January, we take, in place of the crude number
1,YUti dirwtly obtahecl for .January, the number 8 [1.206 -+ 8 (1.130 + l.SHS) ] = 1.209 resnlting froni the combination of the number for January with those for December and Fehru-
aryl and similarly for the other numhers. By multiplying the mean depths of rainfall received at the three stations under coniparison during the interval 1861-1890 by the correspond- ing smoothed ratios, we obtain three independent values for the probable dept,h of rainfall at San Sebastian during this
sctnie period, of which values we take the mean. The follow- ing numbers illustrate this process:
95.7
P2ti.Y 165.1
January.
C!omputd. 1.Wl-lSW.. 107 t:lbwrvcd, ls7s-lsW.. ................. 110 ................
Mrrn.. ............................ ........I 107.5
February. March. April. Annual.
1% 11.1' 1.377 110 w 1-52 1.454 9G
Ilbtal rcri~all1sIS-1.pyo.
6anLbastian ................................... Bsyonne ........................................ Aragorri ........................................ Bilbao ..........................................
Ratio8 qf Sar Sebastian to meh Ptation.
Bsyonne ........................................ Aragorri ........................................ Bilbao ..........................................
Xm. Mwt. Mni . 1.433 1.435 1.140 1.1% 1,110 %%
P.lx'2 2:s 1,536 1.447 1,:Bl 1;Si
l.:W 1.293 1.218 0.W O.G.12 tl.657
0.990 1.07s U .9 2
9G.0 I 119.7
April.
1.24
0. 794
1.116
9Fi. 1
19s. R
1w. 6
116.4 1W.Y
118.8
Keeping only the round nuinhers of millinieters, we finally obtain for the coin mtrd probable values of the rainfall at San Selmtian during t i!l e period lStj1-18MO the following monthly
and annual nunihers, with which we give the niean of the thirteen years of actual observation:
The rallies thus coiiiputed for the years 1M1-1890 differ very niucli i n certain months froni t,he crude mean of t.he thir- teen years of observations. The departures are notably con- aiclcral~le for March and April. and it is easy to assure one's self that. the computed numbers are niore appropriate t,hm t,he olmrvecl niiiii tiers for t,he construction of general charts and the discuw~ion of bha rainfall regime. The arerage of
thirt,een 4'eaL1.s-lS5r3-1sHc.,--Kivrs for San Sehastian for March
n c uant.ity of rain scarcely niore than half that of April; on the ot t er hand, the niiml~er~ deducecl for the period 1Stil-1890 give
:t total for March somewhat Inrger than that for April, as is
found to be the case at all stat.ions in the northern part of Spain. Similarly, the total for August now becomes very nearly the sanie as that for July, whereas it was much larger during the period of thirteen years for which we have actual observations The sanie met,hod of computation has been applied to every inconiplete series: the niiniber of stations for comparison has never been less than t,hree and has been four when the series was very short and the comparison stations near at hand. The numbers have not been subjected to any correction other than this reduction to t8he mean of thirty years. The rainfall .ctations of Spain are not yet sufficiently numerous to allow of determining in a precise ninnner the rCgiuie of certain regions.
APEUL, 1902. MONTHLY WEATHER REVIEW. 239
~~ ~~~~ ~ ~~
* * * It is especially regrettable that observations mere suspended at Lirgos, since now there exista no station on the southwest coast between the mouth of the Tagus and that of the Guahlquivir. The records of the rainfall during each month do not allow us to easily compare the pluviometrical r6 ime of the difler-
tributed throughout the course of the year. For this kind of study it is more convenient to reduce all the anniial siinis to the same value-for esaiiiple, l.O0;)-:tnd to idcullite for e d i month the Itrritjiii.:~tr.ic w;ficieiif-that is to say, the ftzwtion
to thiu month. * * * This mode of presentation of the annual variation of rain- fall by the fraetioiis of the totd annua,l which correspond to the different tiionthfi, although on tlie one hand very advanta- geous, nevertheless presents, on the other littiid, a disadvantage on account of the unequal lengths of t,he months. The rain- fall for February. for es~iiiple, may be less t.han that for January or March. :tnd the relative rtiinfall i n February niight exceed that of either of the other two nioiiths. To remedy this inconvenience, 1 have propased two iiiethods which I will mention aptin here. If thc rain were equally distributed diking the whole yew there would fall 0.0S5 of the total rainfall durin each of the
thirty days, and 0.077 in Februatry. respctively tho proper one of these niini-
given station, we obtain nnmhers which represent, in thou- sandths of the total rainfall, the fraction by which the rain collected each month differs froin that correspmding to R
uniform distribution. These dilferences, which I propose to a l l rehtitv . /wioiia&ic dq,at.trrPv*. represent. at once, inde-
station and the line ilal lengths of tlie nionths, the relat.ive
We thus discover dry months and wet months. or uiinus de artiires and plus departures. fnsbud of subtracting from t,he pluviometric coefficient for each moiith that which corresponds to a iiiiiforni distribution, we niight take the quotient of these two nunibera; we would thus obtain the r-rlofr PP p7wit)iiwf cvwtJi&irfx--that is to say,
the ratio of the quantity of rain that actually fell in any one
month to that which would 1)e colleat,ed if the rainfall were equally distributed during the whole year. Thus at Sm-
tiago there falls in Jtmiary 0.119 of the total rainhll; the proportion:tl pnrt for this month is only tJ.OS5; the relative
pluviometxic coe,ffic.ient for Jt~ti~iar.~ at t.liis atation mill. therefore, be 119i85 = 1.40: that for .Jdy in a similar iiian- ner would he 0.31!0.S5 = 0.36. ' h i s the dry months would be charticterized by a relative plnviometric coetlicient less than unity; the wet months by a coeficient greater th:tn nnity. This latter mode of presenttition of the an timil varia- tion of rainfall. although aniounting siibst,antinlly to the mnie thing as the preceding. will perhaps I -J ~ preferahlo to it,. becaase it leads to certain nuiiibers whose iti~~iiedii~te signiti- cation is plainer. * * * But aa thtn conipntitt.ion is i-ather longer than for the preccding, we shall consider generally the relative pluviometric departures themsc!ves.
. The variability of the iainfall from gear to year is cledured
.from the longer recorda by t.he method of Icnwt squtwes, it varies between 8 and 3S nini. for 17 of the Sptinish st.- .1 t' 1011s.
I n his general work On the irtinfall of western Eiirope, An-
got (Annales Bur. Cent. 1895) says:
In the study of the distribution of rainfall it is not possihle
to content ourselves wit.h annual means. Paris, hlarseilles, and Berlin, not to cite other esamples. receive annaall~ very
nearly the same average rainfall; but it is evident t.hat the
etttstationa-that is to say, the manner in w fl ich the rain is dis-
in thousan x tbs of the total annual mitifall which corresponds
months having t,Iiirt,y-onc days. 0.0W during ~L C ? 7 i of thoso of
bers from the ' p uvioiiietric.coeffilioietit for each tiiottth for a
pendently o f the ahsoliite quantity of t,he rain colleuted at the
distribution of rainfa 7 1 during the whole year. .
By subtract'
pluviometric rk 'mes are absolutely different at these three
it is necessary to consider. periods much shorter than a
year. * * Next after temperature the rainfall is the most important element of climatology. Notwithstandin the interest that it
undertaken i n a systematic manner. The cause is probably to he found in three special difficulties that this subject presents.
(a) The average rainfall received by two stations, even close together. niay be very different, for of all the elements the rainfall is that which iti most affected t)y to)ographic c-mndi-'
of stations. (b) As the quantity of lain received at any point often wries from year to year within very lar e limib, some- times as much as tenfold for the corresponcfing months of two consecutive years, the mean values have no significance, unless for each station they are the means of a large niiniber of yeam ((e) Finally, it is also necessary that the observa- tions should relate to the sanie series of years, since without t,his we iiiight he led to compare a relatively dry period observed at one stdon with a ditferent relstively moist! period observed at nei hboring stations, which would entirely falsify
our results. 'hi# last cuuse of error, t.0 which sufficient :itt,ention has not always been given, niay cause considerable error, even if we consider'quite R long series, such as ten years, for example. Thus. over a reat part of Austria the
I N 0 was scareel? one-half of that corresponding to the ten years lSdl-lS70. The only countries for which to my knowl- edge any one has as yet esecuted works on kinfall satisfyin the three conditions above mentioned are the British Isles an the Iberian Peninsula. For the fornier, Mr. G. J. Svmons in 18S3 published the details of the tohl monthly rainfalls at 367 nbtions during the fift,een years 1866-1880; but this extensive work contained only the observations with no discussion of, results and no charts. Mr. A. Buchan in 1895 published a
r&iiniA, but for Scotland only, of observations made at 344 stations. and reduced them all t,o the same period. viz, the twenty-five years of 1866-1890. For the Iberian Peninsiila I myself published in 1695 the details of all the observations made in that country, lSGl-lrS90, with a discussion of these observations, the general means reduced to the same period of thirty years aiid 16 charts, which allowed one to easily spprehend the peculiarit,ies in the pliivionietric r6ginie of this
I n the present work, which extends this study to the whole
of Europe for the sanie thirty years, lS61-1S!10, I hare made use of about, 375 stations having complete rerords for this period. The addit.iona1 stations, to the nuniber of more than
TH OW, which have been employed in the preparation of the ch:trh, present more or less important gaps. I have never used any record which did not contain at le& ten years' ohservat,ions. Ordinarily the gaps have been filled up by i titerpolation, following esactly the method that I have indi- rated i n my studies relative to the Iberian Peninsula. When
t,he gaps were filled the means were taken by periods of ten years. and I have always been careful to verify the niean of ten y e d y rainfalls hy the sums of the twelve corresponding monthly iiieans. This method of interpolation is quite long, Ibut we can siiiiplify it in a' special case in which we possess mean values for the same series of years. For a certain nnmber of stations having complete observations during thirty ycsrs we may csldate separatelv the monthly means of the r.otii dete period and that of the short period corresponding to
cient by which to niulti ~l y the tiieans of the partial series, in
I n order that this method iiiay give good results, it is evi- dently necessary that the coefficients corresponding to stations
Ytations. I n or r er to obtain IL clear idea of the phenomenon
presents, the study of this phenomenon fl as never yet been
tions; it is therefore necessary to niake we o \ a large nuniber
nit'an rainftdl received in Jatiii~~iy d uring the ten years 1881-
%
country. * * *
mot b er stat,ion. The ratio of these means gives the coeffi-
order to ohtain the pro t Jatde nimn for thirty years.
240 ' MONTHLY WEATHER REVIEW. AP~LIL, 1908
in the same region should differ very little from each other. The small differences t,hat these coefficients present in the
same month in the groups of British stations thus reduced to a c.entid one, and especially the regular manner in which they vary with the geographic positions of the stations, prove
the exact.ness of this iiirthod of reduction. At the same time
we see that the reductions can not be neglected. * * * I give a second example. which in much less favorable bemuse it is drawn from L very iiiounhinous region, where the rainfdl
regime is less regular than in Engltind, and because the period
of observations embraces only ten years. * * * I n this series for the Austrian Alps tlie coefficients of reduc-
tion for any. given month form n regular variation depending upon the position of the stations. so that t.he reduction would
seem to have considerable precision; hit even i n t,hese condi- tions we can, by meaiis of ten years of ohnorvation, calculat~e
the probable avertige for t,hirty years with nn error of 3 or 4
per cent. B.y following the methods thus esplained me nee that it is
pFPible to compute with sufficient :wcur:tcy t,he prol~tihle
monthly mean rainfall for t.he :Lvenige of a long p~riod. surh
as twent years. l)y nietiiid of st,irtions whose ol-wrvations have r i d 9 y estended over :I ritut*h shorter pt*t*iod. Of course
we must not. push this prinviplr hjo fair. A series of t.hrec or
four years of ol>serv:ttiiJns wouIiI pcwc.rally Ibe quitme insuifi-
cient. and 1 1i:tve never ut,ilized st:it.ions hiving less t,li:in ben
years of record. Tlie total tiutnlwr of shit,iotis thus used Is-
keeds 3,000, and I niiLg ndd t h t the oIwrvationn of c:wh station hare always heen cnrefully cwnip:trecl with thost. of its
nearest neighbor, and me have rejected till those t.liat8 seen1
to present t,oo large divergences frcmi the others.
[From AiTgot's hhle of 2.71 iniporhuit statmioils we copy the
following stat,ions that hare mi alt,itude of l?WI) niet.c.rs or
more, and also those having the largest minfxlls, namely,
those above 9,000 millimeters:]
'
aihser- tude east, Lntitiidr. ~ennof' Lringi- I
rutinn. of Paris. !
.~
Stations.
' M~Ier8. MIII.
........................ 157 Mirnt Pemt ........................ 9233 Bentenberg ........................ 260 Bcvrri ............................. '263 Engelherg ......................... 269 St. Bcmnrrl ........................ 270 Bils Mnria .......................... 4ti 58 1.w10 932
76 Alt-Ausser ......................... :?O I 11 L! I 4; 39 !U; ?,OS*
101 Rnihl ................... : .......... '16 I 11 I-L i 41j '3(i 9Sl '1.1s1
-. .. .. - .~
~
All observations have been unifornily reiluced to thc sanie
thirt,y years. lN3l-lS!W, :is Iwfore espltiintxl. The niem
monthly rainfalls have heen ch:i.rt.ed on t i large s d r :itid
isohyetds have heen drawn. These 1:irge partial charts have been followed in preparing tlie published redurecl rhr-irt on t.he
scale of 1 I 1~1.000.000. For a phenonienon whoso geogrtiphiad distrihot,ion is as rariahle :ts that, of the r~itifiill. it is indis-
pensable t,litit, t,he isohyetdx I)e dmwi with grerit, exavt,nrss;
therefore, I have myself dixwn t,lie ciirws in detni I . so t,hnt nothing was left for t,he engraver to do hut. to (:o 'y them.
by the charts in those regions t,hat do not offer too. nituiy
topographical irregulnritics ; but. this is no longer t,he cast' in
nio~int~ainoos countries. The drawing of isohyr,tals in very
mountainous countries is alninxt, irii >ossil)le. for one ot't,t*n finds very difierent tiuioiuit,s of i-,linfnl t nt neighlming stat,ions.
acrordinpas they are on the side of a liioi1titiLitl or in the I~otboiii
of the valley, or are 110th situ:rted at the l)c.)t,toiii of t.he s:mp
valley but h:tvin different orientations. Tli~ts, for esnriiple, in the valley of t e Inn, the three neighboring st:tt,ions. Inns-
bruck. St. Martin in the Cfnadenwald and Hall, having the
I n B general way the clistrihut,ion of rain is fait, I ifidly shown
alt,itudes 573, 837, 1,188 meters respectively, record as mean rainfall for the years 1881-1890, th'e values 7ci!3,1,008, and 1,226
millimeters. respectively. Under these conditions exact
inohyehl cwves can not be drawn exce t upon chartsof a
very large sale and when one has at his t&q~osal a very large
number of stations. The char& that IUT.O~ any this present work therefore represent t.he clistrihution orrainfall for very
nionnt,ainous regions, not in detail, but only in the most
general features. Refore studying the distrihntion of rain nionth by nionth,
we tiinst consider the general lams that govern this phenome
non. la order that rain may he produced. it is necessary that t,he aqueous vapor cbnhined in the air shall be condensed very
nipidly; n slow condensation gives only cloiids or fog. The mixture of two masses of air, both of them saturated, but at
different temperatures, is. as we know, powerless t~ furnish
ntiy apprecirtble quantity of rain. The foriiiation of rain is
therefore depenclent on the sudclen cooling of the tiir. This
cooling can he produced directly, tis when a mass of warni
tiioist trir passes over a colder region; hut, the rains produced
by c1ircc.t cooling are relatively of little importance. In fact, rc-mling at€ t?cth$ only t,he layers of atriios >here new the ground.
r:Lin it,self, :ire ehcient causes of the mrtrniinp up of the soil.
By far t,he niore iniportmt cause of the prochic-tion of rain is the coolitig by cspansion which accompanies every ascend- .
itig niovenieiit of the. air. Tlie laws of this cooling are well known. The. mcending nioveinents of the air can originate i n three different ways, corresponding to which are three dif-
ferent. classes of min: (A) Certain ascending iiiovenienb take
place in connection with the general circulation of t.he atnios-
yliere; for example, thc asctending currents of the equatorial
rryions: the c~onvectional rains correspond to the.: se move- nients. (B) Other ascending niovenients accompany bnromet-
ric depreseions and thunderstorms over a grt!ater or less art
atortiis. (C) Finally, whenever aerial currents encounter
mountains or ereti slight elevations of the soil, they produce
niechanically on the Ilttnks of these moiintains ascending inorenletits from which there result the so-called orographic:
rains. We should add that the initial temper:tt.ure of the air exerts
:t preponderating influence on the intensit? of the rainfall; the wwnier the air the niore moisture it contains for a given per-
centage of relative humidity and the more liquid water it will
give up for the same lowering of temperature. The influence t,Ixit is iiiost prominently nianifest,e.d on the rainfall chart? is
that of the orography. It has often been said that u raintall
i h r t is only a rough copy of the hypsometric chart of the
s~tiie region, bat in fact t,he rainfttll chiirt is far more compli-
rated. On the side of a mountain exposed to the wind, t,he
rain increases with sltitucle, at first rapidly, but, we can easily we that thin increase is not indetinite. If the mountain 18
diciently high, we ohserve B zone of maximum rainfall at a
rertniii :~ltitode, above which the dirninution is very clear.
An nn:rlogons phenomenon is produced if the mountainous niaSs
is very brortd; the rainfall is especially heavy on the border
of the niass, whereas the central part receives far less rain. I'he charts that we piildish oder B st,riking example of this in
t,he region of the Tyrolian and Austrian Alp. 1.11 every ciise without exception the center of this titc~~~nt~aitio~is region
a iuininimii of r:iin relative to its northern atid s o i i f ~~~
Hanks. The influence of the orogirtphy is very cleiw; every
c1i:iin of mountains shows a ninxiniiini of rainfall. The snialler
elevations sonictinies suflice to develop vcqy appreciable max-
ima. Another causc of complesity in the irtinfall niap arises ft-ont tlir fact that thc sir, aft,er having rei*ipiLited iipoii the
t:tined and after having snrmountecl the obstacle, is now much
poorer in vapor and less capable of producing rain than if this
h1oreovt.r. the rapid ititlow of tiir. an cf even the forniation of
,
of their estent; the82 give birth to cyclonic rains or thiin tl er-.
H:i.nk of the niountrtin a large part of t R e water that it con-
APRIL, 1903. 'MONTHLY WEATHER REVIEW. 841
obstacle had not existed; therefore to the leeward of every maximum of rainfall there should be t~ niinitiium. This
occurrence of a minimum to the leeward of a maximum with
reference to the direction of the prevailing wind is also very clearly shown on the ma s of European rainfall. This double
maxima and minima of rainfall in tieighhoring regions, is
the principal cause of the complexity of the rainfall charts. In order to study the manner in which the iainfall is distrib-
uted month by niont.h, it does not suffice to directly conipare the monthly means of various stations ainong themselves, since
the absolute quantities of rain vayy very iiinch from one sta-
tion to another. This conipnrison is facilitat,ed by dividing the monthly means by the annual totnl; the quotients, or niont.hly
plwiometric ratios, re.present the, fractional part:s of the totd
rainfall independent of the absolute quantity of rain. The numbers thus obtained differ but, little for any given niont,h,
for all the stations within a large area; t,heg nre therefore f:w
more convenient than the absolute heights for st.udying bhe
pluviometric r6gime. The monthly fract,iotis present one small inconvenieiic?e rcsulting froiii the ineqiutlities of the
months. In order to represent the pluviosity i n a siniple ant1
exact manner, I h p w proposed to calculat,e the pluviontet~ric~ coefficicnt for each nionth I J ~ t:tking the ratio of the qit:uit,it.y
of actual rainfall to thRt which moitld have heen observed if the min had heen distrihited uniformly t.hroughoatj the year.
If the pluvioniet,ric coeljicient for any month is lens than unitry it shows that less water httx fallen t,lian corresponds t,o n uniform
distribution. The months will be dry or wet.nccordinp as their coefficientu arc less or greater t,hm one. * * * By
these exaniples and especially 1:y coniparin the cha.rts of
bhe actual depth of rainfall we see how niuch iiiore simple is the
consideration of reliitive pluvionity than of absolute quantity.
influence of the orograp \ y, which produces at the same time
plnviometric rCgii1ie or relative pluviosity wit a those t.hat give
(32) PROF. A. J. HENRY.
I n Bulletin J), The Rainfall of the United S t a h , by Alfred
J. Henry (lY97), t,he a d i o r has collected all accessible records,
including such long ones as that, for eighty-three years at, New
Bedford and sisty years at St. Louis. We quote a.s follows:
With regard to th6 elevation of the Weat,her Bnreaa gauger, high above ground on the roofs of Iar e flat, buildings: In gen-
eirtl, it does not seem possihle to svoif the conclusion that the
observed anionnt of precipitation falls short of the t,rue nmoont.
by qnantities varying from 5 to 10 per cent of tlre annual
rainfnll. Uniformity of the years of observntion: Heretofore it has
been suflicient to accept, the available registers of t,he varioiio districts. whether of ten or twent:y years' clnirttion as re re-
shown, however. that. gears of fat and lean ra.infal1 do not
a1ternat.e i n orderly seqitence and that a nrunher of consecu-
tive years of hesvy rains can not be safely ac.rept,cd as indi-
cating ti pernianent, increase of rninfall. Length of record required for a normal: A t,riie nornial
may be defined as one which will not be materinlly itltercd, however iiiiwh longer the obwrvntions may be c*mtinur.cl.
* * * The writer does not know of a single rninfall re is-
conditions of environment and observational accuracy. In
order to obtain the estreine variation and the powible err(~r of a libpear period. t.he average of the tirst 10 ytws of ewh
register was computed, then dropping'tlie tirst year the aver- age of a second period of 10 years was computed. Proceed-
ing in like manner, 74 separate cotlibinations of 10-year periods
were obtained for the New Bedford record. The average for
senting the true average precipiht,ion. *' * * It wil 7 be
ter that was estrthlished and has been perpetuated under i f ea1
83 years io 43.5 inches; that for the 1p years 188G-1893 is
50.4; for the 10 years lS37-1846 it is 38.S. As coiii red
per cent and 11 per cent, respect.ively. I n like manner the estreme variations for decades at Cincinnati were 20 and 17 per cent; at St. Louis, 17 and 13; at Fort Leavenworth, 16
and 18, and at Sail Francisco. 9 and 10. I n a similar way
the extreme varitLtion of the iiiesn of N 25-year record was 10 per cent. The conclusioti is reached t,hat at least 35 or
40 years' continuoils ohservat,ionn are required to obtain a result that will not, dppart, iuore than 5 per cent above or
helow the nortiinl. The addition of the 5-year period 1893-
1896 t,o the monthly and ttnniial averages up to the end of 1891,
:LS published in Bulletin C!, does not materially change the
averages he.retufore determined except over the west Gulf
coast,. L)uring the great,er portion of the period 1SPi-1896 drought prevailed in intiny purts of the L~nited States, and
there does not. seein t,o he any law of conipensation hy which
a deficit i n one district, is baltuiced by a surplus in another.
The locd distribution of ininfall is exceedingly errnt,ic; thus the catch of two gtoges I!aving prwtically the same exposure
and but n few miles apart tilay differ 8s much as 10 or 15
itdies in the totd of t,he gear. For (wnvenience of wniparison bhe niontahly averages were
rediit*ed to )erc.cntit~es of t.he nnnital fall. Arranging these
coiiipaintivelg large area,* is prrtctic?ally uniform, and h a t the
protile of ti single stat.ion iiiny represent the entire district. We may t,herefore view the'rainfall of the United States not
as :I single c*onc:rete system, hut rather as being composed of fourteen separat.e and distinct types, which are described in
detail. Periods of heavy and light rainfall continuing for two or
three yearn are not infrequent,. Professor Henry gives the following centers of groups of dry growing seasons, viz,
1860, 1863. 18Ttk71, 1881, 1887, and 189G95. Similar peri-
ods for the whole year, but. no regnlar periodicity, are easily
debected. A very general deficiencg occurred during the ten consecutive years 1XS7-1886, thus showing how long a series
of years is nee,ded in orde,r to o h i n normal values. The
oscillations from dry to wet, or vice versa. are oft,en very remarkable. Thus on Mount, Hamilton, California, 911 inches
fell in 1884, but only 1 Y in 1885, and this contrast, prevailed over a great, part. of t,he Pacific Coast States and the plateau
region. During December, lS89, the whole system of atnios- pheric circulatioti and storm niovenients seem to have been
shifted for the t h e being 5 or 10 degrees to the southward.
with the mean for 83 years, these decades are in error r y 16
I)y geograp I it* clist,rtc.t,s, it is seen t,hnt t,he distrihut.ion over
-
(35) C. A. SCHOTT.
The Smit~hsoniwi Tahles ancl Results of t,he precipitation in
miti and snow in the United States were compiled by the late
Mr. Chides A. Schott. the tirst edition published in March,
1879, and t.he second edition in May. 18M1. Charts of mean
a n n d precipitntion (mi ti and iiieltecl snow) acconipatig thee
volutnes. Ditferent, cditions of tbis chart of anniial iltinfall
nre d:tted August, 1868, March, 1S70, titid 1877. I n geneid,
the cIiart,s accotiipiuiying the second edition of the t,ext include
data up to the end of 1874 and in some cases the end of 1876.
For each station all wceasible rainfall records are published,
heginning in one case, Charleston, Y. C.. with the year 1738.
There is no evidence that, the nunierow shorter seiies were
reduced by any niethod of interpolntion to homogeneity with
one fundauiental period. Only in t.he case where a few months
were missing were these interpolated, in order to give com-
'
2453 MONTHLY WEATHER REVIEW. APRIL, 1908
plete years. . It was very often neceasary to accept,, instead oi
nielted snow, the observer's rule of laking one-tenth of t h e
depth of the measured snowfall. Iii the second edition, May,
1881, the charts themselves how by shadings and isohyetal
lines every ti inches of annual precipitation from 8 inches up
to tis, thereby attempting to show more detail than would
ordinarily be considered advisable in view of the small 1111111-
her of stations and inequalities ah to time. Records are given
for about 1,500 stations, hut only 1,300 were plott,ed for t.he
compilation of the chart of 111ean aiiniinl rainfall. On these
matters Mr. Schott says:
An asterisk atlisecl to any niiiiiber indicated that it is derived
from less than twelve months of observation. I n no case, how- ever. are annual amounts given for which niore than three months had to be found by interpolation, which was effec?trd either by using the observed anioriiitr at the nearest station dur-
ing the t h e required, or by using the nieans from ncljacwit yea13 for the same month or months. The tirst mode of inter- polation is quite reliable :tiid \vas always preferred to the sec-
ond. The anniia.l iiienns at t.he bott,oni .of each column i n
Tahle B are t&en from the precwling 'l'able A whencver the series was not continuous. The inntusl rel:ttions and nigniticancy of the tabular resulk
crtn best be brought. ont by a gralihicd presentation. In this forin the amount and genom1 clist,rihution of the rainfall over the country cnn'at. onve lie seen, and admit+, at the same
time. of a close study of its spe+d features. The increased nin-
terial at disposal since the publication of the first edition of
the tables and the importance to the agriculturist of :t knowl-
ed e of the distrihution of, the rainfall in the several seasons
struct two new ones, thus presenting five. viz, one for t,he year and one for each of the four seaisom. For the delineation of the geographical distribution of the aqueous precipitation over the area of the United States, the same base ma has been made use of which served for
ture, published by the Sniithsonian Institution in 1 S W U It has,
however, been improved hy the iotroduction of the mountain systems, and it is believed that the study of the relationshi i
of the distribution of temperature and of rainfall will he faci f - itated by this uniforiiiity of projection rind scale. For the
generalization of the resulB of the greater part of the western and elevated portion of t.he United States, the scale of the
nia appears inconveniently lurge. our material-being too lim-
adapted for the exhibition of the general and the detail fea- tures presented i n the numerical results. To explain the con-
struction of these charts it sufioes t.0 show it for the one
exhibiting the annual distribution. All sthons for which the observations extend over four or inore yearr~ were plotted by their coordinates, latitude and longitude, and against the dot was written the ainoiint, of precipitation in inches and t~ the nearest. tenth of an inch; for it11 ot.her stations with series of
observations shorter than four years the position was inarked
as hefore. hut only the nearest whole iiwh was written against the dot; the relative vitlue of the resulh WHH thus, in a nieas-
ure. indicated in the construct.ion of ciirves of equal rainfall. These curves were drawn wibli a free hand aiiiong the clots by
graphical inter >alation. and with due regard to the iniportance
hyetal lines, and construeted i n 'the intinner of contour lines generally. are graduated for certain equal increments of rain. The difference l?et.ween :uljnc?ent curves resulted from the con-
u Smithsonian Contributicms to Knowledge. No. 377, "Tables, distribn- tion, and variationsof the at.niruJpheric temperature in the United States "
By Charles A. Schott.
in 8 wed t.he Institution to enlarge the charts, as well as to con-
the exhibition of t 1 e distribution of the atmospheric t,eniperit-
ite f to sdequat,ely cover so large an wen. otherwise it is well
of long and s i!l ort series. These ciirves. clesignxt,ed as iso-
-- ____-~_____
Washington, 1876.
sideration of the probable uncertainty of the results. If di-awn too close-that is, if too many curves were shown- they would exhibit temporary or accidental inflections, which would only tend to complexity and confusion; on the other hand, if the curves were too wide a art there would be danger of losing portions of permanent P eatures in the distribution. The distinction between long and short series in the graph- ical process is of importance in a phenomenon of such great variations froiii year to year and froin the ~aiiie season in dif- ferent years, and the numbers of the second class given in whole Inches were used to ooinplet,e, modify, and generally to iniprove t,he curves resting iipon the niore reliahle data. Special consideration was necessary to select for each chart those particular c:urves and their gi-aduation which would best bring out its leltdin features, and further, to facilitate the
charts, cdor shading was introduced. The curves. whether principal or iatermeditlte, are indexed. and ('an thus be easily followed by the eye. and each chart, is su plied with a suffi-
cient. explanstion t,o be understood, twen w K cn detached from
the text,. I t was neither necessary nor prti.t?t4ic!itl)lc. fro~ii want of spac?e, to indirate on the charts the iiidiviclual stat.ions nnd their ainoruit. of rainfall, though they are caroaded in o n the manuscript charts. Thus the iiiiiiiher of stations plotted and utilized for the chart of the nie:~n annual distrilmtion is alJout
1.300, and the nunibers are largor for each of the season
chrts. Comparing the new w i b h the old rain charts, the siipcriorit,y of the former will tie :Lpparent; and while perhaps too nioc~h detail was given on the ch:trts of the tirst edition, which, through the increase of observat.ions. is now known not to forni part of permanent feat.ures, but arose from insaffi- cieiit data at that t.ime, yet the nppttre~lt distortion of the curves, when the two sets are conipared, may be produced by small changes in the aiiiount of rainfall; and, while the en-
era1 features are preserved, the present charts will bring t f em out only uiore prominently and broadly.
understanding and t % e ready interpretation aad use of the
(34) ALEXANDER BUCHAN, ESQ.
In the magnificent Atlas of Meteorology, published 8.8 Vol-
ume 111 of Bartholoniew's Physical Atlas, London, 1899, there
is given a list of. iiieteorological services and stations, froin
which it would appear that there are about 30,000 stations for
observing rainfall scattered over the continentn and islands of
the world. The regions that have one or inore to every 40
square miles are rJamaica, Barbados, 8t. Kitts, Great Britain
and Ireland, Denmark, Saxony, the Straits Settlement.q, Vic-
toria, and Maurit.iiis. For the greater part of the world there
are only scattered stations; for Rurope and the United States,
India and Australia, the genei-al avei-age is about one for every
5,000 square miles. Notwithstamding this apparently large
number of stations, yet, when we 1?0111e to'inake up R rainfttll
nisp, we find so m+uy short or broken series t.hat the actual
nuiiiber available is reduced to :t third of what we appear to
have. The Bnrtholoiiiew atlas has collected together nearly
R l l that is known about'rainbll, and the following remarks by
the editorrr are appropriate to the present occasion. On pages
1 to 4 Alexander Buchan sap:
The use of the r:ain gauge dates as far hack as the time of Leonarclo de Vinci. but as rain gauges continued long to be
placed on houses and other ohjeotionatble situations. the ob-
servations in only a few cases are comparable wit.h those now made. Not until the middle of the present century-can the
A m , 1902. MONTHLY WEATHER REVIEW. 243
quality of the observation8 and the numher of trainfall station$
be regarded a i sufficient to represent this all-im ortant facto1 oPclimate with a first approach to accuracy. - E * * The
maps of this atlas show in the most conclusive manner that the rainfall of any particular region is determined by the pre- vailing winds of that region, considered in the first place in relation $0 the regions from. which they have come, and, i n the second, the physicltl configamtion and the teniperature of
that part of the earth’s surface over which they now blow. I t is then seen, for esani ;e, that the maximum rainfall is precipitated by winds whic!, having tmversed a large breadth of ocean, come up against and blow over a niountaino~~s ridge lving across their ath; and the amount deposited is still fur-
latitudes or through regions the temperature of which is con-
stantly hecomin ctolder. * * * The rainfall is ~t i ~i s ~a l l ~
small or even ni f when the prevailing winds have not prevr-
, ously tmversed a considernhle extent of ocean or have crossed
a mountain ridge and .advance at the same time into lowe!
latitudes, that is, into regions the temperature of which con-
tinues to become higher, * * * and this peculiarit\. ir? presented in the most pronounced form when the winds arriv- ing from the ocean blow out inimedietely from a well-marked anticyclone which presses c.lose toward the shore, that is. out of
a region characterized by great atmos heric dryness. * * *
our econornic needs and froni its intimate and vitd connectioii with changes of weather, stations for its. observation require to be mort-. thickly planted than has as yet been done. Lever- rier was not far from the inark when he urged the parochial ohservntion of rainfall. I n mountainous regions more rain-
fall stations are niuch needed, especially in connection with the water supply of our great centers of popiilntion and prob-
ably the gra3t extension of the use of water power in industry. Until this be done, the engineers’ deiiiand can not be met for
a statementt of the rates of increase of rainfall with increased height under different clinithc conditions.
ther increased if t K e winds pass at the saine tinie into higher
From the ininiense iniportance of t. f e rainfall in relation ta
(Sa) A. J. HERBERTSON, ESQ.
On page 17 of the test of Bartholomew’s Atlas of Meteor-
ology, Mr. A. .J. Herherkon says:
“ In rainfall maps the actual mean rainfall values are entered
without any correction for :iltitnde. atid then the isohyets are drawn. In preparing i.ainfsl1 imps for the months. awouiit should he taken of their differentr lengths. All the monthly
rainfall tilaps in this atlas, escept those for the United States of America, show i:iinfall values reduced to one-twelfth of a
year. The isohyets on these monthly niaps have therefore two meanings: (1) The tigiires attached to them show the nieaii monthly i-~iinfttll espressed i n terms of one-twelfth of n yc:~r;
and c2) the linen also wpresent an actual irec4pitation during
and the exact aniount of which will be found on consulting
the month. differing slightly from the ~a r lies marked on then1
the Table XU.” [According to this table I 0 0 nun. of rainfall in one-twelfth of a year corresponds to 101.5 in 31 days, or 98.6
in 30 days, or 92.5 in SS+ day.--E~.]
(S 6 ) PROF. VICTOR ICEEXSER.
In the Meteorologische ZeiBchrift for July, 1500, Vol. XVII,
pp. 489-317 and 337-355, Yrof. Dr. Victor Kremser, in the
course of his review of the climatic conditions of the Memel,
Yregel, and Weic.hse1, discusses the distribution of rainfall from
several points of view, but we will summarize only that which
relates to the variation with altitude. The distance from the
sea, in and of itself, seeins not to he so important as the other
factors, siichas the distribution of storms, the direction of the
wind, etc. With regard to altitude the following bables show
it variation in the winfnll gradient, depending on the peculiar-.
itks of location.
~~~ ~
Pregel and roast regions.
In the valley of the Weichsel, on the northwest slope of
the Prussian territory, 14 stations, having an nvemge alti-
tude of 139 meters, gave an average annnul precipitation of
592 mm., whereas on the southeast slope, 14 stations with an
average altitude of 146 meters gave ail average annual rainfall
of 55.7 mni. I n the valley of the lower Weichsel the preaipi-
tation increases toward the south and east, and the effect of
elevation is shown in the preceding table.
In the upper Weichsel region we have the following figures:
€3$ comhining both altit.udes and horizontal distances, Krem-
ser shows that with one esception there is always a diminution
of precipitation as we go from the west toward the east.
regards the influeiwe of the gpneml trend of the coast of the
Baltic Sea, he shows that along n line stretching eastward 4
stntioiis gave a niean rainfall of 4% nini. ; stretching north-
ward, Y stations gave 543; stretching westward, 6 stationsgave
658.
AH F
C
6
C
S
8