中文摘要：

Penman-Monteith方法需要的气象资料较多，在气象资料不全时无法使用，并且该方法不能直接预测参考作物蒸发蒸腾量（ETo）。参考天气预报的常规预报项目，将日最高气温、日最低气温、反映天气状况的日照时数、反映风力等级的风速以及Penman-Monteith公式中可以直接计算的相关因子作为自变量，以Penman-Monteith公式计算的ET0为因变量，引入偏最小二乘回归算法（PLS）建立了逐日ETo预测模型。实例计算结果表明，PLS模型的模拟和预测精度较高，基本能够满足农业生产实践的需要，为有效预测逐日E7＂o提供了一种新方法。

英文摘要：

Calculating the reference crop evapotranspiration （ETo） requires much meteorological data, in the course of ETo calculated by Penman-Monteith method, there will be many serious correlations among the calculational variables. Based on routine weather data, the article introduced partial least-squares regression algorithm to establish model to forecast daily ET0 by the top and low air temperature. The application results show that the preferable simulation and forecast precision of the model in practice, and presents a better new method of forecasting of daily reference crop evapotranspiration.