The first step in the determination of irrigation water requirements f
or use in project planning, design, and operation commonly involves th
e prediction of the reference crop evapotranspiration. This paper pres
ents a simple mathematical model for forecasting reference crop evapot
ranspiration at three stations in California that represent humid, sem
i-arid, and arid weather conditions, covering three extremes in the Ca
lifornia climate. A time domain-time series model is identified for th
e reference crop evapotranspiration data at each of the sites and comp
ared to other simple methods of forecasting using monthly average and
yearly difference approaches. Based on the results of model comparison
s it is concluded that a generalized time series model can be used to
forecast reference crop evapotranspiration at each of the stations, an
d that the time series model can provide a reasonably accurate, yet si
mple method of forecasting reference crop evapotranspiration in many a
gricultural areas of California.