基于天气预报的参照作物腾发量中短期预报模型研究

Middle and Short Term Forecasting Models for Reference Evapotranspiration Based on Daily Weather Forecast

以新乡市1970—2011年逐日实测气象资料代入FAO 56 Penman-Monteith（PM）方法算得的ET0作为基准值,对HG、P-T、M-K、M-C模型进行参数修正,将新乡市2012—2014年冬小麦生育期间预见期为1、3、5、7、10d的天气预报数据代入修正后的模型进行ET01～10 d的中短期预报,并以2012—2014年冬小麦生育期间逐日实测气象资料由PM公式算得的ET0为基准值,对天气预报的精度及ET0的预报精度进行评价。结果表明：经过参数修正后HG、P-T、M-K、M-C模型的精度均有提高;最高气温、最低气温、风速、日照时数的预报精度均随预见期的增加呈逐渐下降趋势,最低气温预报的精度稍高于最高气温;不同预见期的ET0预报模型中,P-T模型预报的ET0平均准确率在众模型中较高（95.06%）,其次为HG-M模型（94.66%）、PMT1模型（94.34%）、M-K模型（93.89%）,且P-T、HGM两种模型计算程序较简单,因此优选P-T、HG-M模型进行ET0的中短期预报。

The Hargreaves （HG） equation, McClound （ M - C） equation, Makkink （ M - K） equation and Priestley-Taylor （P-T） equation were amended against the FAO 56 Penman- Monteith（PM） by using the observed meteorological data from 1970 to 2011 at Xinxiang City. Then, the calibrated equation and weather forecast data were used to predict the 10 d ETo. The precision of weather forecast and forecast ETo was evaluated, and an analytical method was developed to translate daily weather forecast message into the variables needed by estimation ETo. The results showed that the mean accuracy of calibration HG （HG- M） equation, P- T equation, M -C equation and M -K equation were relatively higher. However, the maximum and minimum air temperature, wind speed, hours of sunshine tended to decline with the increasing forecast horizon; the accuracy of minimum air temperature forecast was higher than that of the maximum air temperature. The mean accuracy of P -T model was the highest （95.06%） among all models, meanwhile, the HG - M equation, M - K model, PMT1 model were in turn. The prediction accuracy of selected four models were higher than other models, but the P - T and HG - M models could satisfy the requirement of higher accuracy and simplify the calculating procedure. So for ETo middle and short term forecast, the P - T and HG - M models could make a great contribution to improve the accuracy of irrigation forecast.