/ ';d"i;;n J ; st«. U!/drol. Geophy~. (lD76), 27. 2. 16;j.

I GH

551. 524 : 551.504.4(083)

Simple formul ae for the estimation of wet bulb
temperature and precipitable water
S. JEEVANANDA REDDY
MeteO'l"oWgical Office, Poona
(Received 5 September 1973)

AnSTl~ACT. A simple formula 1' 1/} = T (0· 45+0. 0013 h vp/1060) has been developed for computing wet bul b
temperatur e (T'l') on electronic computers from. dr y bulb t emperatu re (Tl , relat ive humidity (II) an d pr essure (11).
The results obtained with t his eq uation nrc compare d wit h observ ed va lues IlOU a rc foun d to be in good agreement.
A simple relation of the form T w=c V W has also been derived betwee n wet bulb temperature and precipita ble
wate r (Wj. The resu lts of IVobta ined with t.his equat ion arc compared with t he obser ved valu es and arc foun d to be in
__good agreement .

1. Introduction 24 hourly mean data for t he yea r 1959 for 31

A knowledge of th e moisture cont ent iu the
atmosphere is of great importance in all metcoro-
. logical studies. The parameter genera lly used for
t his st udy is water vapo ur. B ut, for t he estimation
of water vapour content of the ent ire column of
the atmosphere (i .e., precipitable water va pour
conte nt (W) at a particular point requires upper
air data. As t he upp er air data is being recorded
at a limited number of stations over India, it would
be useful to have alternate simple methods for
th e estimation of IV. In biometeorological st udies,
t he wet bulb te mperature is of direct use (Haldane
1905, Leonard H ill 1911-14). These st ud ies requi re
continuous recording (over a day ) of t he wet bulb
temperature. The autographic recordings of t he
element are, however, not ava ilable from 1967
onwards.
Moreover, t here is no simple form ula for th e
estimation of wet bulb temp erature on electro nic
computers from other available meteorological
elements. I n the present st udy, a simple form ula
based on dry bul b t emperature, relative hu midity
an d pressure has t herefore been derived for t he
est imation of wet bulb temperature on elect ronic
computer s. A simple relationship between wet
bulb te mperature and precipitable water has also
been obtained.
2. Derivation 01 simple tormuta for the estimation of T tl.'
~ . l. Correlation st udy between (T-T",) and
different met eorological elements showed high
correlation with h (when compared with e).
Therefore a linear regression equation of the for m
(T- T",) = a+ bIt, was first fitt ed, For this, monthly
163
stat ions well distributed over India were con-
sidered (i .e., 31 dat a points for each month).
The correlation coefficient (c.c.) between (T-Tw)
and 11 is denoted by r, th e regression coefficient
(r.c.) by b aud t he regr ession constant by
a. T, (l an d b in different month s are shown in
Tilble l (a ).
It IS seen from Table 1 (a) t hat a and bare
different for different mont hs, while a is nearly 100
tim es to b in all the respective months. It is a lso
seen t hat t he va lues of r in winter month s are
relat ively small. In order to improve th e c.c,
(i .e., r) in winter months an alterna te form (TwiT)
has been considered in t he place of ( T-T~) .
Correlation study between (TwiT) and h has also
shown high c.c. Therefore> a regression equation
of t he form T wl1'= a' +b'h , has been tried using
all the 12 mont hs data (i .e., 372 data points). The
results of c.c. bet ween (T~/T) a nd 11, th e r.c, b'
a nd the regression constant a' (where, a' and 1/
a re same for all months) obtained from this analysis
are respectively 0 ,95, 0 ·006 a nd 0· 43. H owever,
it is seen from Ra o's (1945) work t ha t l 'lC decrease s
with decreasing pressure (p) for same value of T and
It and also t his decrease is more associated with
h rath er t han with T . Therefore, a pr essur e COrrec-
tio n has been made to h in t he above equation .
Th c above equation has been again solved by mu lti-
plying h with VP (here P varies nearly from
1000 to 750 mb) as, ,.11' increases t hc accuracy of
th e estimates compared when 11 is multiplied with p
(where P is th e annual normal pr essure in mb
which is const ant for a station). Tho c.c. va lue
varies from 0 ·8 3 to 0 · 98 and 0 ·95 to 0·98 res-
pectiv ely when h is multipli ed wit h p and \I f'.