k, 1917. MONTHLY WEATHER REVIEW. . rti9
1918.
Jan. 2 .............................. 5. ............................. 10.. ............................ 10.. ............................ 13. .............................
Feb. 7 ..............................
TABLE 4.--dcaZimum wind velm't , in milcs per hour, fbr ajve-minute
period, midnight to mkhight, at hather Bureau anemometer, and rims of occumence 08 refmtd to the time of shgt from eouthwmt to west.
Date.
blis./hr. 68 w ......... Bo nw ........ 50 sw ........ 50 w ......... On time. 76 w ......... 4 hourslater.
70 w ......... Noevidence.
Mar. 7 ..............................
Apr. 14 ..............................
Oct. a5 ..............................
Nov. 23 .............................. 24. .............................
Dec. 6 .............................. - ____
80 w ......... lghourslater.
56 w ......... 10minuteslater.
80 w ......... 50minuteslater.
52 w ......... 1 hour later. 72 w ......... Noevidence.
88 w ......... Ontime.
_____- - __ .
branches as the stud of radiation or high upper air altitude research (wlere American scientists take a prominent part) we owe the wonderful equipment of modern research to the inventive genius of men like Cleveland Abbe, R. Assniann, C. G. Abbot, C. F. Marvin, etc. As to recording instruments of all kinds the meteorologist of to-day is perhaps better equipped t8han any other student of natural phenomena, and m
tainly much better than the followers of the yo-
sister science of oceanograph In a previous note in thisLvmw (see issue for JW,
1916,44: 3381 I have tried to set out the reasonsfor a close
cooperation be tween meteorologista and oceanographers in an extensive study of the North Atlantic Ocean, of the variable physicd conditions of its surface sheet, and of the influences direct and indirect which these latter undoubtedly exert on the weather and the climate of the surrounding continents. In the following ages I .intend to give a brief description of some nove f instru- ments which highly facilitate tshe collection of oceano- --
gra hical data. 8ver since scientific marine research was for the iimt time taken up on a broad basis by the International Council for the Study of the North Sea, it has become more and more manaest that if results of any practical value for the fishuig industries or for meteorolo cal fore-
investigation must be drawn ti ter, the frequency of
This is special1 the case in coastal regions where our investi ations xave roved physical conditions in the
atmosphere, From our research station Borno in the
Gullmailiord (west coast of Sweden) we possess uni ue series of continuous hydrographical observations $a- liiiity and temperature measurements a t different depths) which have been made daily for over seven years with few breaks. The results prove that the difFerent water layers, differing in density from each other, instead of being a t rest are in a state of almost incessant motion with lar e displacements both vertical and horizontal takuig p f ace in a very short time. Thus the boundary surface between two water layers may, from one day
casts are to be denved from this work then t!E e net of
observations increased, and the fie T d of research expanded.
fluid 2 ement to be a f ost as variable as they are in the
20 meters, which of course
involves a the hydro aphical situa- water Q eing swept out to sea and replaced from below by a simultaneous inva- sion of saltier water of North Sea origin. At the sugges-
tion of the author a series of synoptic parallel observa- tions of these internal movements, previous1 only studied a t Borne, were carried out at five c6fferent localities during November, 1915, b the Swedish Hydro- graphical-biolo ical Commission. h e results prove that
era1 character and that parallel displacements of the boundary occur almost simultaneously at all the difFer- ent points of observation. As re ards the practical importance of these move-
on fish life. 0. Pettersson has repeatedly found par- ticularly large u heavals of the boundary surface to
fisheries. The author has compared e above-mentioned hydrographical series with statistics from the local &&erg and found that the chances of catching mackerel at Borno in summer are from four to six times better when
the bounda surface of 30 per mille salinity is below ita
the more consi % erable internal movements are of a gan-
ments, t B ey must obviously have a profound influence
t!i coincide with ric % catches of herrin during the winter
average leve Y than in the opposite case.
MONTHLY WEATHER REVIEW. h, 1917 160 .
S c l $d ~~ demieter.
In order to follow these internal movements more closely 0. Pettersson in 1909 constructed a recordmg densimeter b means of which the hidden displacements of the boun B ary surface are made to draw their own
€3~. bo. Pettemo~Vs recording d a r n e t a , out-boerd portion.
record. The main features of the arrangement are sketched on figure 1.’ A large cylindrical co per float is free to move vertically from the surface to t e bottom at 30 meters between two guide wires of brass. The float
is com letely filled with sea water and a little araf6n
(B in the figure) so that it barely floats in water of r mille salinity at 10°C. Therefore when the correspon mg bounda surface rises or falls it will carry the float upwa or downward, the movements
oil, an B is balanced against a counterpoise of s o d metal
about 30 $?
3
- -- - -- -- - - - --- --- - ---- -- - -- - -- - _- - -- -- ------- ---- -- -
-T --
--- --- ------ - - --- - -- --- - - - ---- -- --- - - - - --- - - -- - - --- - - - - ---- - - -- - -- --
F~ci. 2.-H. Pettasson’r -tk d-.
being transferred to a wheel on the bridge serving as
support, and from there to a recording pen and drum on the shore. It is not an uncommon experience to find the float one day floating a t the surface m clear salt water and next day out of sight beneath a sheet of brownish water of low salinity, 10 meters or more in depth.
.
1917. MONTHLY WEATHER REVIEW. 161
The record from this instrument, the first by which vertical movements in the sea can be studied, has for
eam been checked by daily titrations on water samples k m different depths and it represents a most valuable material, both for hydrobiological and hydrodynamicel research. Thus boundary waves, both of tidal origin and enerated by seiches, have been observed and studied fheoretically by Nils Zei1on.l A modification of the above instrument intended for work in the open sea, combined with a thermograph, was devised by the author two years a o and ordered abroad, its completion being unfortunatey k still delayed by the war. Hydmsta.Eic &&meter.
In order to get a sim ler and less espensive arrangement
has constructed and tested an entirely novel kind of densimeter based on the hydrostatic principle (see fig. 2).
From a vertical, inverted U-tube of glass, ABC, 10 mm.
for the study of sinii P ar internal movements the author
with the outer water, as well aa to prevent any disturbance in equilibrium b the motion of the surface wave8 and an aspirating e d ects of horizontal currents. h e instrument is inexpensive, reacts instantly to internal movements, also to such of very small am litude, and may therefore be used also for the study o P Helm- holtzian waves. Moreover, the readings are extremely simple, the actual position of the boundary surface being found b a glance a t an instrument on the laboratoq wall wit% as little trouble as that of reading the am temperature. It would obviously be of great advantap if the instrument could be made self-recording, but thls involves considerable technical difEculties. In figure 3 the upper curve shows graphically the
lotted readings taken from a similar instrument at
f30rnii during July 27-31, 1916, com ared with the
agreement is seen to be perfect. Thanks to the kind assistance of Capt. G. Ridder- stad, commanding our research vessel, the S k q d ,
simultaneous curve from the recording a ensimeter. The
12 m. bam. 12 m. 6 pm. I 2 m. 6 am. 12 m b pm. 12 m 6 am. I ? m b pm I2 m b am I? m e pm 12 m. 6 am. 12 m. 6 pm. 12 m.
27.- Juh 2S Jifi 29 Jult YJ Jul; 31. Juli
FIG. &-Readings of the hydroetetio dadmeter (upppa curve) mupared with the aimulteneously m r d e d curve from the recording densimeter.
wide and 100 cm. long, two tubes of rubber or metal lead the instrument has also been tested on board during a
down to depths of say 2 meters and 25 meters, respec- week in July, 1916, when the shi lay anchored at the
tivel The whole system is filled air-free with salt water entrance to the Gullmarfiord. &e results are highly of a b o w n densit , dF, only the upper half of the U-tube interesting as they reveal the esistence of rythmical containing a IiquiJ of lower density dL (toluene colored oscillations in the boundary having a sharply dehed. red with azoamidobenzol). If the average density of the period (15 minutes). water in the fiord between 3 and 35 meters just equals The author hopes that a slightly modified type of the d , then the toluene will be at rest at equal heights in hydrostatic densimeter may be installed on several both branches of the U-tube. But if the salinity in the lightships anchored off the coasts of Denmark and fiord is increased by an upheave1 of the boundary surface, Sweden. then the toluene becomes pressed over from the “deep” to the “shallow” branch of the U-tube. The meniscus between the toluene and the water in the U-tube, which
Water 6ottles for direct nzeasurem.ents of tk W t y .
can be mad to within 1 mm. on a vertical scale, will therefore in general rise or fall in a manner exactly
In Order to as far as Possible the e i q u e of hYdrogra hical soundings where no high accuracy is
arallel to the movements of the boundary surface in the ~~~~~~~l ~$h i ?~~$ \ $~~~p $~e ~~~$’ tion of the salinitsy and the temperature of its contenb. Eord, but on a scale reduced in the pro ortion 1 :r where r = (a, - &)I(& - d J , ds and dz being %e density of the sea water a t 2 meters and a t 25 meters below the surface. for the use Of
special at the lower of the rubber fishermen who may derive information of practical value from similar observations. The water bottle, which can
at any desired depth by means o€ the ordinary closing
This instrument is intended
tub- SBl’Ve to retard the intermkture Of their Contents be made to inclose hemetically a volume of water
SV- €I~&--Bbl@h m m B w t e r , V.
162
.
MONTHLY WEATHER REVIEW. h, 1917
arrangement with a messen er has a cylindrical body of some transparent materiaf gfass or "cellon." Within the latter there are, say a dozen of glass balls of different colors, inset pieces of netting a t the top and a t the bottom of the glass cylinder preventing their escape when the bottle is open. The lass balls have been so balanced
sity 1.027, the next a t 1.026, and so on. On a simple
as to float in waters o ? different density, say one a t den-
thermometer mounted asially within t,he water bottle the temperature can beread to within 0.5 degree (C.), and b counting the number of lass balls
the water bott e when the latter is brought to the surface, one finds the salinity to within say 0.0005. The instrunient is estremely simple and has been used with ood results by two
ing cruises after henin Owing to
tions in stratified waters, the position of a boundary surface may he located with the same accuracy by the aid of such an instrument as from B series of
water yam les taken a t say evnrj- fifth iiinter a n f analyzed by titration to
within 0.00002. The instrument just described would obviously be greatly improved if it could be made to give more exact readings of the salinity. I have con- structed another water bottle answer- P ing to this demand, sketched in fi -
ure 4. The body of this water bot& which in other respects closely resgm- bles the receding, is narrower and longer an8 is graduated by a number of arallel rings, one for every fifth m Ig imeter. Instead of the numerous balls, it contains a single cylindrical float of glass, F, which is held tight b three clamps at the lower end of
open, but becomes released automati-
c d v as the bottle is closed. The float
% a B oat or those 1 ing on the ottom of
common fishermen % uring reconnoiter-
the peculiar hydrograp % ical condi-
i
t g e water bottle when it goes down
the; rises in the c linder to a height which can be rea (9 to within 1 mm. This height is brought into direct pro- portion to the density of the water, to an automatic adjustment of
of a very fine cham of gold or of metal attached to its lower end. latter device was suggested to me by mv friend Dr. Anders hgstronl, of
the "Y uoyancy of the float by means
mo 4 -H Pettemm's U" sda.
sit &itb wste.r &&read& bottle. k i t h a float measuring about 40 cu.
F -&w l m m p m u ~ crn.involumeand achainweighing(sub- merged) 24 mg. per cm., a rise of 1 nun.
b the float corresponds to an increase in the density o 9 the water of 4 in 100,000 or a change in the salinity of about 0.00006, For ordinary investigations in the upper
layers of coastal waters this degree of accuracy is quite sufficient. The temperature is measured on a sensitive thermometer inserted through a hole in the lid. For
exact measurements of the temperature of the water in situ the instrument has to be combined with a reversing
-
thermometer. At depths exceeding a couple of hundred meters neither of the two last instruments can be used, owing to the high pressure on the bulbs or on the float.
Areometer with chain.
The automatic adjustment by means of a h e chain attached to the float may also be used for ordinary are+ metric purposes. In fact, the arrangement represents an almost ideal solution to the old demand for a practical areometer of total immersion. All the errors and corrections caused by surface tension involved in the use of ordinary areometers are avoided with the cham areometer. The only source of error (besides air bubbles and nonhomogeneous temperature in the water) is due to the chain taking up anirregular position, as when it hangs curved instead of depending straight down from the areometer with a shar bend at right an les where
the glass vessel the chain can easily <e brought to a correct position before readings are taken. The sensitiveness of the arrangement, which may be made excessively hi h, depends on the size of the float
difficulty of excluding temperature disturbances. For readings of moderate accuracy, 1 in 10,000 or even
1 in 20,000, the float may be of ordinary glass and the vessel containing it a simple glass tube,. preferably with outlet below (and with gimbal suspension for work on shipboard). An instrument of this kind has just been tested with good results, also in rough weather, by one-of my students, Mag. Sjiistrand, during a short cruise with the Skagemk.
it touches the bottom of t f e vessel. B gen t f y tappmg
and the weight of t f e chain and is limited only by the
[FIG. 5-Rereived too late to be included.]
If B high degree of accuracy is desirable, sa of
100 cu. cm., and should be made of silica in order to
Also the measurements should then be made wit Dewar flask to eliminate as far as possible all temperature disturbances. Figure 5 [not reproduced] shows an in- strunlent of this type used by the t~uthor.~ With the aid of a similar instrument for the deeper samples and for those demanding a higher accuracy, and with a water bottle of the type in figure 4 for the other water samples, the hydrographical so.undings of an expedition may be carried out with a niiiiimum of labor and without the necessity of storing any water Sam 1 s for titration ashore. An important point with regar8 to similar instruments is their "range," i. e., the interval of density over which they may be used. With instruments of moderate
accuracy or with tshe water bottle, the positipn of the float may vary by about 350 scale units whch, if the sensibility is taken to be 1:30,000 corresponds to a
range of 1:60, say, from 6=1.010 40 6=1.026. .This ran e again answers to a maximum Interval in sdnity of a f out 30 (not considering the influence of temperature variations) which meets all t,he requirements of an ordinary hydrographical expedition. With high-precision instruments the total vertical tfisplacenient of the float can not very wed be increased beyond 300 nim. If the accuracy is 2 in 100,000 per mm.,
the range will be about 1 : 160, or, say, from 6 = 1.022 to
6 = 1.038,wliich is againsufficient inmost crises (forbottom
1 : 50,000 or more, the float must have a volume of a i out
avoid corrections for the thermal expansion of e;
-. -. .----
9 The drawing lor Bg. 5 and the YS for Table 1 were sent b the author st a later d8te Lllm the balance of the article here uhlished. hut the mails hled to reach the editor in time for this issue. See %sue for &, 1917.4. A., jr.
Apxur,, 1917. MONTHLY WEATHER REVIEW. 103
layers in coastal regions as well 89 for measurements in the open ocean). Obviously the range of the instrument can in both cases be extended a multiple of times by attachin to or
num or quartz. detaching from the float standardized weights o P plati-
FIO. 6.43. PettePswnC precision chain-eompensated areometer with overhead ch.ln.
A disadvantage of the instrument in figure 5 is the large amount of water required for each measurement, the
capacity of the vessel being some 800 cu. cm. Acting on a
sug estion by 0. Pettersson that the chain ma
to the bottom of the vessel, the author has had con- structed an instrument of the type shown in figure 6.
The spherical float is alwa-ys brought back to its zero position by allowing a suitable len th of the chain to be
float, by unwinding the chain from a cylindrical drum with a spiral groove. With a float of 150 cu. cm. (sensi- bility 1 in 80,000) the net capacity of the Dewar vacuum vessel is only 300 cu. cm., and the range 1:80, the total length of, the chain available for measurements being
100 cm.
be B et down on the float from above as depen i from as well it
accumulated on a small tray attac !.I cd to the top of the
Recording current meter.
It is technically a most difEicult problem to measure currents in the sea at different depths, even with a mod- erate degree of accuracy. A considerable number of
different current meters have been constructed for that purpose. However, also in this case, it has been found deslrable for many purposes to keep up the obsmations for some time, repeating the measurements at regular in- tervals of time, as for exam le, when the tidal element of the currents is to be studiedl With ordinary instruments such observations are extremely laborious and the strain on the human observers very great. The onl succes8ful
vented by 0. Pettersson (modified and improved by the author). Figure 7 conveys an idea of this instrument (vertical section1 .4
A cylindrical envelope of brass, provided with a large rudder which keeps the direction of the current, is kept hermetically closed by a lid and contains the rertezing apparatus, two opaque @ass dish D, and D,, eac with a row of numbers round ita circumference. At every thir- tieth minute those numben, which happen to be before the aperture of a tiny camera are photographed by a flash from an electric lamp L on a film slowly moved by clock- work. One of the dish carries two compass needles whereas the other disk is turned round b the current out- side, acting on a current wheel W b d ow the cylinder. From this wheel the movement is transferred to the in- terior and to the disk by means of two astatic magnets MI and M,, one inside the other outside, the magnetic force acting strai ht throu h the solid brass bottom of the
recording current meter hitherto constructe B is that in-
lealagel. #ach photogram on the iilm two numbers, one representing the average current during the last half hour, the other a t the moment of exposure.
ws
FIB. 7.-The Pettermon recording current meter for d e e m exploratbn.
4Pstkrsron, E. A reeordhg current meter for deep-ses work. Quart. jour.. Royd met’l. soc., London, 1815. Fig. 7 reproduced by permlsslon of the editor of that j o d .
164 MONTHLY WEATHER REVIEW.
The current meter works automatically during a fort- night and may be left for that time unattended sus ended from one of Q. Ekman's submarine buop.4 I!) this mode of sus ension, fkt suggested by the author, one
measurements from an anc ored ship and due to the ro er motion of the latter. The Ekman buoy, by its souhe anchorage, is kept at a depth of say, 5 or 10 me- ters below the surface and there stands as firmly as a rock unaffected by the waves and currents of the surface. A few of these current meters thuq anchored in the Strait of Florida or off the coast of Formosa would serve
to keep under observation and to record the amount of water carried northward; thus it would be possible to determine the ossibility of using such information in making season3 forecasts of the temperatures over the eastern United States and Europe, or over Japan.
Oon.clusio11..
E
avoids all t % e errors usual1 inherent in the ordinary
Notwithstandin the brilliant results gained through
or of Johan Hjort and Sir John Murray, the vast field of research offered by the ocean calls fo? international coo ration on a large scale, if the desmd harvest of
this work appear to the author as particularly worthy of sttention: I. The existing international network of meteoro- logicd observations, sus nded during the war, should be extended also over oceans by means of regular observations from an adequate number of transoceanic liners, reporting by wireless. These telegrams also ought to mclude observations of the temperature and the salinity of the surface water.5 11. A special survey of the most impoftant cold and warm currents and their regions of junctlon or conflict should be systematically maintained by crukm of
hydrographical observations. 111. The internal movements, both horizontal and vertical in the stratsed water near the coasts should be followed b regular observations from a sufEcient num- ber of h e B stations and lightships along the coast line. The results should be compared with those from simul- taneous hydmbiologicd observations (prevalence of fish eggs, larvae, and fkh food or plankton) and the yield of the local fisheries, both as regards quantit and quality, and
of fogs, and, in cold climates, the freezing of fiords and bights. If the oceanographers and meteorologists of the United States, of Canada, and of Japan were to unite their efforts with those of northwestern Euro
soon be on the high road to new and starthg scientific discoveries and also to results of the greatest practical value.
individual efforts t 'ke the cruises of Sir Frithiof Nanseii
useul r results shall be reaped. The following lines for
research steamers fully equlpped for meteorologlca Y and
&o with observations of the local wea ti er, the occurrence
search along these or similar lines we should un R" oubtedly in re-
ON WOBKING UP PRECIPITATION OBSEBVATIONS.
A number of-the younger station officials, enthusiastic in the devel- opment and d u d o n of meteorolwcsl data and particular1 that relating to the rainfall of the count that seem to indicate a lack of famrkrity with the more fully devel
have proposed projects orstud
4 Pdbcrsson. If. A roconling current meter lor d e w work. Quart. jour. Royal Metl. Soc., Landon 1915.
edtlon'by the author has beem ineluded in a p~posal for the reorganization oI't%w% MeteomlogIcsl Servire presented to the Government by the Ewediah delwta~ to the International Council br the Exploration of the Eea.
APBIL, 1917
oped methods of analyeia of observational data and processes for elimi- nating defects or errore due to c h q s in methoda or in obaervera or other thinga that bring about dkcontmuity in a lon series of obeerva- tione. In order to d t such etudenta in +e pm%lem of d' our rainfall observations, we offer the follmng tnanalation of= s in Dr. Hugo Meyer's "Guide to the workiy up of meteoro- r o y observations for the benefit of climatology." Although the
OR 'nal is over 25 yeara old, the methods reaented are still standard anfthe principles stated are still regardec! aa fundamontsl.-Ch:hicf of Bureau.
HOMOGENEITY OF THE OBSERVATIONAL MATERIAL.
In working up or discussing meteorological observa- tions the very first care of the student must be to deter- mine the homogeneity of the series of observations he is using, i. e., to make sure that the changes in values (both periodic and nonperiodic) arise solely from chan es in weather, and that he has excluded all those sucl d en or gradual cha.nges which may arise from a change in expo- sure, or in instruments, or in instrumental constants, or from a change in the observer-changes that at times may be of as great a magnitude as a change in location of the station. Therefore, if one is not perfectly certain that the tabulations he desires to discuss in further detail, actually do present the march of the meteorological elements he should undertake to test the homogeneity of the different factors of the series. * * * Althou h it really seems to be a matter of course that
series of observations before undertaking further discus- sion of them; and althou h Schouw em haaimd the
stance has not been fu y appreciated untll Hann's recent investigations into t8his point.3 Hann has also shown the most convenient way for appl mg tests of homogeneity.
a station is based on the experience that radicalchan es
in weather are rarely confined to a limit,ed region, rat % er they take place with the same sign and with more or less equal intensity over extensive districts. Hence tho. dzfmences [in the case of. pressure or temperature] between si,multaneous observatmns at neigh6orin.g points, are. 9nwch m e c d a n t than the obserwd values themselves. Accordingly the testin of the observations a t a station
observations a t a neighboring standard station whose work is of guaranteed accuracy; or if no such standard station is available then the comparison is to be made with simultaneous observations a t not less than two nei hboring stations.
&e h t method for comparing the observational re- sults on a meteorological element a t different localities which are not too far apart is the phk mthod. The means for all the years (or months T under consideration are plotted on coordinate paper, using the same scale for each station and arranging the corresponding values a t all stations for the same year in the same vertical line; each air of points for the same locality are then con- necte B by a straight line. In this way one secures 6 num- ber of broken lines corresponding to the numbex of sta- tions brought together for comparison. In each of these
lines the rises and falls seem to succeed each othex with- out order. On comparing all the curve8 it must appear, however, that the succession of rises and falls is the same
one shou K d convince oneself of t.he homogeneity of a
The method for testing the o t- servationd matenal from
point as early as 1897,lyet t % e full bearin8 o P this circum-
involves a comparison o K the h t with the simultaneous
I Meper Hwo. Anleitllng zur Bearbeitung meteorologlseher Beobaehtun n fIlr die Klimitologie. Berlin, Jullus Springer, 1891. viii. [4], 187 p., 214 cm. (Sef&tims
arc from pp. 43-45.51,83. and pp. 182-140.)
2 EM in thls connectloll Jdrm Hmn. Untersuchu n Uber die Re nverhlgtltnisse von Oenterrcich-Un rn. I. Thnik Die jBlirliche Perlxder Niedemhz BltsungR- ber.,. Kaiserl. Akafiad. Wlssensch., matli.-naturw. KI., Wlen. 1879, &Is 571-635, particularly p. 573478.