14s APRII. 1959
A NOTE ON VISIBlLITY AT SACRAMENTO, CALIF.
GEORGE C. HOLZWORTH
Weather Bureau Research Statlon, Berkeley, Calli.
[Manuscript received January 13. 19591
ABSTRACT
Daylight visibility obserrations at the Sacranlento Municipal dirport Jvere ns,ed to investigate the possibility
that risibility in tlle area has been declining clue to rising levels of xir pollution. Those risibility observations
made with naturally occurring fog (defined as relatire humidity greater than !10 percent) and/or precipitation were
eliminated from the data. Obserratioms for .July, :I war111, dry month with relatively slight climatic differences
from year to year, and So\-enlber. a cool month that often exhibits it considerable annual variation of elinlate, were
tabulated for the three 4-year Iberiods, 1935-38, 1!14.7-48, nnd 19L-.Xi. The res,ults show that for July the percent
frequency of poor risibilities steadily increased, and of good risibilities steadily decreased over the last 20-odd
years. Such a trend is consistent with the Sac.rmnento 1)opul:ttion increase. For Xuveniber the trend was irregu-
lar. This is believed to be due to ruriations in the occurrence of nleteorological factors, important to atmospheric
dispersion. It is likely that such irregularit,ies are chnr:rc.teristic of other months and areas where there are large
annual climatic variations.
Fro~n concurrent surface \rind observations, the 1)referred wind c.onditions for various risibility ranges were
deterluined. These data indicate that poor risibilities at the nirl,ort, on the southern edge of the city, occurred
most frequently u-ith light lvinds from over the nearby Sacranlento urbun area. Good risibilities were favored
by Inoderate \rinds fronl rural directions. Visibilities \vere rrdncwl when the wind sl)eed 1)ecnnIe strong enough
for loose materials to be 1)iclcecl up from the ground.
1. INTRODUCTION
Residents of Sacramento and other cities in (’alifornia‘s
Central Valley, as well as aircraft pilots in the ricinity,
have stated that, oyer the years tljey have observed a gen-
eral decrease in visibility. These individuals have at-
tributed such a trend to increasing levels of air pollution.
Because air pollution measurements have not) been made
in the Central Valley until recently, studies of trends in
the concentrations of specific pollutants are not possible.
However? the solid ~n t l liquid 1mrticulate mltter sus-
pended in tlle :Ltmosphere, which represents air polltition,
reduces visibility [ 11, and therefore :I measure of visibility
may be interpreted as a measure of air pollution. Hencse,
changes in risibility not attributable to naturally occurring
aerosols may be indicative of mannlade trends in air pol-
lut,ion. This study was undertaken in order to assess t l ~e
possibility that there has been a general deterioration of
1-isibility in the Sacramento area due to hun~nn activity.
2. DATAANDTREATMENT
From tlle hourly weather observations made by the ITS.
Weather Bureau at Sacramento Municipal L4irport, the
J u l y and Kovember data for the three 4-year periotls
l!)X-%? 19X-48, and 19Z-56 were selected and punched
on t*:uds for machine tabnlation. Only d:Lyt.ime observa-
tio~ls, 0730-1730 IW,~ were considered since visibility at
nigllt- depends upon different factors and the observations
are gener:llly less reliable [I ]. All observations, 1935-56,
were made from the same location.
Prior to 1943 visil)ility to be reported by Weather
T3tIrenn observers \\-:IS defined as “the greatest, distance
to\\-artl the llorizorl that, kno\m objects can be identified.”
I his \\-as later revised to “tlle lowest. visibility over the
half of the horizon with the greatest visibility.” Sum-
nlaries for this study do not suggest that this change in
definition had any noticeable effect, on Sacramento
observatiolls.
.J~tly and November visibility trend analyses were con-
sitleretl wit11 respect to the three 4-year periods. The data
n“e1.c: lmnped into 4-year groups in an attempt, to smooth
out, year-to-year variations i n meteorological conditions
swh :IS factors affecting the dispersion of pollutants.
The natural aerosols which limit visibility are fog and
precipitation. Precipitation nlay occur at various relative
r 7
6 6 anti about 41 minutes past the hour for 1935-38. A t Sacramento in
8 ActuaI obserration times are about 28 nlinutrs past the hour for 1945-
Xorrmber the latest wnrise is about 0700 PRT and the earliest sunset is
I I P ~I Y 1700 TST.
APRIL 19.59 MONTHLY WEATHER REVIEW 149
PERCENT
60 R
0 1935-38
48
.,
"
3-6 MILES 7-10
~~~~~ ~
MILES 11-15 MILES 20-30 MILES >30 MILES
F ~c ;r .l c k : 1.-July yisibilities at S.rcranlento Mnilicipal Airl)ort, O'i:(o-
17:10 PS'I'. Percent freqnrncirs when relaltire hun1idity less than
91 1,rrcent and no prec.il)itation occurring.
humidities. but. ~l a t ~u x l f o g is unlikely below about 91
percent. Tllose visibility observations vitll precipitntioll
and/or relative humidity greater tl1nn 90 percent) were
tnbulated separately, and I.emaining risibility observatio~ls
were originally broken down according to various other
relative humidity ranges. Although the visibilities tendetl
tobe lower with higher humidities, tlle trends betweell
4-year periods for various llunlitlity ranges were all
similar. For this reason. and because i t is c1esiml)le t o
deal with large sunlples, all visibilities rerorcied wit11
humidities below 91 percent (and no precipitatioll) were
considered together. 111 96.7 pel*cent of the ,July visibility
cases, humidities were less than 71 percent. I n Koven11)er,
humidit,ies were below 71 percent in 60.2 percent of tllr
cases and below 91 percent in 83.9 percent of the c;~ses.
3. RESULTS
I n the following discussions visibility observxtiolls v-itll
precipitation and/or relative hnrnidity greater tl1an 00
percent, are excluded. This provision is intended to elinli-
nate visibility restrictions that might, be attributed to
nat,urally occurring fog and precipitation particles.
JULY
July was chosen as typical of the warm, :wid sunlnlers
when there are seldonl marked variations in Central Val-
ley weather. The percent frequencies of risibilities in
given ranges for each of the three 4-year periods are shown
in figure 1. This bar diagram indicates that, the frequency
of visibility reports in the ranges 3-6 and 7-10 miles
steadily increased over the years studied. The percentage
frequency of the combined :3-10-mile range increased by
a factor of almost 6 from 1935-38 to 1953-56, apparently
at the expense of the 20-80-mile range whose frequency
PERCENT 1935 -38 1945-48 1953-56
steadily declined f1-onl 62.8 percent in 1935-38 to 44.5 per-
cent, ~I I 195kWj. The frequency of visibilities of11-15
miles c1l:mged only slightly and inco~lsistently with time.
I n the range exceeding 30 miles, there was no regular
trend. There were no July visibilities of 0-2x miles.
('onsidering a11 visibilities, a deteriorating trend is clearly
indicated for J d y .
The most striking risibility trends were in the 3-10-
and 20-3O-mile ranges. By the chi-square test [2 ], dif-
ferences between tlle three &-year periods for both these
visibility ranges are significant) :It the 1-percent level. It
sllould be mentioned that, such tests of significance assume
independence of samples, and hourly weather observations
(lo not fully satisfy this :wumption. However, each of
tlle 4-year groups was considered to more clearly approach
tl~e. property of independence.
NCVEMBER
('lin~ntologic:llly, Sovember is a relatively cool month
and was chosen to represent early winter conditions as
o1)posed to summer. The weatller is often quite variable
within this nlontll and considerable differences may occur
frolu year to year. Some Novembers are rainy and foggy
while others are relatively dry and fog free.
1 he percentage frequencies of visibilities in give11 ranges
for each of tlle three 4-year periods, as shown in figure 2,
indicate that. Kovernber visibilities followed an irregular
trend. As n whole, poor visibilities (0-10 miles) m-ere
least frequent, and good visibilities (20 miles or greater)
were 'most frequent in the middle period, 1945-48. I n the
later period, 1953-56, the picture was reversed with 0-10-
mile risibilities most frequent and visibilities of 20 miles
or grea.ter least frequent. In the early period, 1935-38,
1)oor and good visibilities occurred with frequencies inter-
lnediate between those of the other two periods. The fre-
quency of visibilities of 11-15 miles differed between peri-
ods by only 0.2 percent. Considering all visibilities for
Sovember, they may be described as having been inter-
lnediate in 1933-38, good in 1925-48, and poor in 1953-56.
7 ,
150 MONTHLY APRIL 1959
This irregular va.riation stands in sharp cont.rast to the
steady trend shown for July. As mentioned above, data,
for 4-year periods were lumped together in t,he hope that
important weather variations would be smoothed out.
This result, is more probable for July than for November
when conditions may be quite variable. November aver-
age 4-year precipitation amounts were 1.09, 1.41, and 1.59
inches for the earliest to latest periods, respectively; for
any one November in these 12 years, precipitation ranged
from 0.03 to 3.35 inches. These are cited to indicate to
some extent the degree of wetktller variability in these 12
Novembers. It is thought that such variations in impor-
tant meteorological factors, affecting both dispersion and
emission of visibility-reducing pollutants, have resulted
in t)he irregular November trend shown in figure 2.
VGth respect to pollutant sources, the burning of crop
residues after harvesting should be mentioned. It is
thought that this practice, which has been common in tlle
rich agricultural area of Sacramento and o’tller parts of
the Central Valley, is partly responsibmle for the irregular
trend of November visibmilities. Smoke from what agri-
cultural burning occurs in July is likely to be more rapidly
dispersed by the usual summer conditions of windiness
and instability. However, because the desired facts on
agricultural burning are nonexistent, such remarks must
be regarded as speculation.
4. FURTHER CONSIDERATION
It is generally recognized that as urban areas expand,
pollutant emissions from such sources also increase (in
the absence of emission controls). It is logical then that,
other things being the same, visib’ility should show a gen-
eral decline with increasing urb’anizntion. The Sacra-
mento Municipal Airport is on the southern edge of the
city, about 4y2 miles from the downtown area. Because
Sacramento is the only large city in the county and because
much of the total metropolitan population lies beyond the
official city limits, Sacramento County population figures
have been used. From decennial population data [3] and
others [4], the following populat.ion estimates were made
for Sacramento County for 1 year within each of the
three periods studied :
1937-------------------------------------- 160,000
1947------------------_”------------------ 230,000
1955-------------------------------------- 380,000
Thus the trend of deteriorating July visibilit.ies was
associated with a, rapidly increasing population. I n No-
vember this same trend held true for 1945-48 to 1953-56,
but improving visibilities from 1935-38 t o 1945-48 were
also associated with a population increase. As pointed
out, this latter irregularity is believed to have been due to
variations of important me,teorological factors.
That the Sacramento metropolitan area was indeed a
Source of visibility-reducing pollutants may be shown from
a consideration of surface winds. For each visibility ob-
servation the concurrent wind observation was tabulated
in wind rose form for each visibility range. Wind direc-
tions were divided into two groups, northwest clockwise
through east-soutlleast (NW-ESE) and so’utlleast through
west-northwest ( SE-M7N1V). As slmwn in figure 3, most
of the Sacramento n1etropolit:m population lies in the
sector NM-ESE from the airport. The other sector, SE-
WNW, may be described as relatively rural. The rela-
t,ionsllip of poor visibilities (0-10 miles) to’ winds was
considered first. This was done by summarizing the per-
cent frequency of occurrence of wind in each of the two
direction sectors, NIT-ESE and SE-WNW, by 5-n1.p.h.
speed intervals for 0-10-mile visibilities and for all visi-
bilities. The ratio of the former to the latter was then
determined and these values for July and November are
given in table 1. Each numb’er in this table is the ratio
of percent frequency of given wind conditions for low
visibilit,y to tlle percent frequency of such wind conditions
for all visibilities. Thus, the higher the ratio, the greater
the “preference” of low visibilities for such wind condi-
tions.z For example, table 1 shows that in July the most
preferred wind condition for 0-10-mile visibilit,ies was
from NW-ESE at 1-5 m.p.11. It should be made clear
that these wind condit.ions are not ne.cessarily those with
which t.he visibility was most often 0-10 miles. The pref-
erence is relat.ive to what is expected from a consideration
of all visibmilities.
possible to com8pute the ratios of “percent of time t h e w i n d is from
2Using tables K and F of Local Climatological Data Supplement, it is
given directions for visibilities less than 7 miles with smoke or haze” to
“percent of t i m e t h e w i n d is from corresponding directions for all
visibilities.”
APRIL 1959 MONTHLY WEATHER 151
TABLE I.-Ratio of "percent of winds for visibility 0-10 miles" to "percent of winds f o r all visibilities." Data for Sacramento
no precipitation; 0730-1 730 PST, 1935-38, 1945-48, 1953-56
Municipal Airport when relative humidity less than 91 percent and
Wind speed
m.p.h.
Wind direction
I-
NW-ESE SE-WNW Calm A 11
I I I-
JULY
(No visihility reported in 0-242 miles renae)
1.36
.2 7
.4 7
.1 R
I . 04
Wind speeds exceeding 15 m.p.11. were infrequent, and
so the rat,ios for such speeds had to be conlpult,ed from
small percentages. However, the computed ratios appear
to exhibit reasonable trends as wind conditions vary. A
likely except.ion is the vahe of 2.23 in table 1, July, which
probsably would be lower with more cases of such wi-inds.
This fault might be c0rrecte.d by conlbining some of the
higher wind speeds, but in so doing one of the features
of the table would be lost. It is a reasonable hypothesis
that,, considering nearby communit,y sources of visibilit,y-
reducing aerosols, improving visibility sllould be antici-
pated with increasing wind speed (more favorable dilu-
tion) up to the point, where dust and other loose maiterials
a.re picked up by the wind. This is borne out in tab'le 1
where, considering all wind directions, theare is a. decrease
in the ratios with increasing speeds up to the point where
loose nmteria.1 is picked up. Here the ratios b'egin to
incmase. I n J u l y when precipihtion 'aver:Lged less than
a trace and t,he ground was dry, this reversal began a t
16-20 m.p.11. I n November when precipita,tion averaged
1.36 inches, the ratios did not begin to increase until
speeds exceeded 20m.p.11.
From tab'le 1 it is clear thak wind direct,ions occurring
with low visibilities 1wre most likely to be from the urban
area. This was especially true f o r J u l y when for every
wind speed the ratios were greater for the sector NTV-
ESE tha,n for tE1e SE-TYNW sector. For r\'ovembe,r, con-
sidering all wind speeds, the most preferred direction with
low visibilities was also from the metropolit,arl area.
However, the preference WRS not nearly so strong as in
July. This may have been due in part to agricult.ura1
TARLE 2.-Ratio of "percent of winds for visibility 20-30 miles" to
ipal Airport when relative humidity less than 91 percent and no "percent of winds for all visibilities". Data for Sacramento Munic-
precipitation; 0730-1 730 P S T , 1935-38, 1945-48, 1953-56
Wind speed Wind direction
m.p.h. IGw-EsE I s E -w N w / Calm I All
JULY
burning which provides pollutant sources to the south
as well as to the nolkh.
Table 2 is similar to table 1 except it is for good visi-
bilities, 20-30 miles. This table indicates that for all
directions there was an increase in the ratios, preference
for golod visibilities, with increasing speeds up to the
point where loose materials are picked up, and then the
trend reversed. Considering all wind speeds, the most
preferred direction for goo'd visibility was from the rural
sector, SE-TVKTV.
Hygroscopic nuclei may be active in the humidity
range 71-90 percent. If such relative humidi'ties are re-
lated to wind directions, then the concurrent visibilities
might be lo'mer and t,ab'les 1 and 2 would be inva,lid. July
claytime humidities above 70 percent are very rare. I n
Kovelnber, humidit.ies of71-90 occurred 23.7 percent of
the time. The Novemb'er ra,tios in ltabmles 1 and 2 were
recomputed, using only data for relative humidities be-
low 71 percent (and no precipitation). The recomputed
ratios are almost identical with the original. It is con-
cluded that any relaitionship bsetween humidities of 71-90
percent and wind directions does not alter the ratios of
tables 1 and 2.
5. CONCLUSIONS
The datu for July indicate that Samamento visibilities
lmve been deteriorating over the last 20-odd years. This
is most apparent in the 0-10-mile visibilit,y range, which
showed a steady increa,se in percent frequency, and in
the 20-30-mile range, which showed a steady decrease.
I n November, Visibilities went through an irregular trend
from intermediate to good to poor, from the ea,rliest to
the latest periods, respect,ively. This may have been due
152 MONTHLY WEATHER REVIEV APRIL 1959
largely lto variations i n the occurrence, of meteorological
factors, especially those important to atmospheric clis-
persion of manmade pollutants. It is likely that such
irregularities are cllaracteristic, of other seasons and areas
where t,here are large variations in the weather.
From a conside.ration of wi-ind data it is clear tllxt poor
visibilities were associated with wind direc,tions from over
the urban area and with light, winds. Wind direction and
speed factors were interdependent.. Good visibilities were
most likely with winds from rural directions and at
moderat.e speeds. The data indicate Itll:tt, visibilities were
reduced when winds b’ecarne strong enough to pick 1111
dust, and other loose materials.
ACKNOWLEDGMENTS
This work was carried out in cooperation with the (“ali-
fornilia S a t e Department of Public Health as a part of
the program of the Bureau of Air Sanitation. Various
peoplo in this Hureau made helpful suggest.ions and the’ir
contributions are gratefnlly acknowledged.
The raw data, were kindly made availab’le by the U.S.
Weather Burelan at Sacramento. The encouragement of
the Ofice of Jieteorological Research and the Weather
Hnreau Resetlrcll St’ation alt, the R. ,I. Taft Sanitary
Engineering (‘enter, (lincinnati. Ohio, is’ also appreciated.