干旱区生态环境知识资源中心

The first step in the determination of irrigation water requirements for use in project planning, design, and operation commonly involves the prediction of the reference crop evapotranspiration. This paper presents a simple mathematical model for forecasting reference crop evapotranspiration at three stations in California that represent humid, semi-arid, and arid weather conditions, covering three extremes in the California climate. A time domain-time series model is identified for the reference crop evapotranspiration data at each of the sites and compared to other simple methods of forecasting using monthly average and yearly difference approaches. Based on the results of model comparisons it is concluded that a generalized time series model can be used to forecast reference crop evapotranspiration at each of the stations, and that the time series model can provide a reasonably accurate, yet simple method of forecasting reference crop evapotranspiration in many agricultural areas of California.