Potential evapo-transpiration (PET) data are required for a number of
climatic analyses. As measured PET values are seldom available, partic
ularly in historical climate data sets. it is necessary to estimate PE
T from available climatic data, which are often limited to rainfall an
d temperature in ordinary, rural meteorological observatories. Simple,
empirical PET estimation procedures, even though generally location o
r region specific, are often the only alternative. In this study we co
mpared open pan evaporation and estimates of daily PET from two empiri
cal formulae viz., Linacre and Campbell (a modified Priestly-Taylor fo
rmula). with PET estimates by the Penman formula, usually considered t
he most accurate PET formula. for five diverse arid and semi-arid loca
tions in India. Estimates of PET from the Linacre equation were (1) ge
nerally more linearly correlated with Penman estimates of PET, and (2)
the standard deviations of the difference between Linacre and Penman
estimates were lower, relative to similar comparison between either op
en pan evaporation or Campbell PET estimates and Penman PET. Estimates
of PET by the Linacre formula were consistently higher than the Penma
n equation estimates, but with an appropriate calibration factor. deri
ved from a few years of complete weather data, the Linacre formula can
be used to estimate PET for semi-arid locations where input data for
the Penman equation arc not available.