融合Kriging算法的河套灌区ET估算方法评价

Evaluation of ET estimation method in Hetao irrigation district based on Kriging algorithm

为解决数据缺乏时对大区域ET估算的限制,评价了融合Kriging空间插值算法的河套灌区6种ET估算方法。基于Kriging算法和气象数据,得到7种ET估算方法的河套灌区区域ET值,将试验站实测ET值与6种方法ET估算结果进行精度验证,并以PM计算结果为标准,评价融合Kriging空间插值算法的不同方法在河套灌区的估算精度与适用性,分析不同方法与气象因子响应关系的差异性。结果表明,综合法(KP、PVB)与PM的结果最接近,计算结果最大偏差0.045-0.490 mm·d^-1。推荐综合法为融合Kriging空间插值算法的河套灌区ET最佳估算方法,适用于灌区所有土地类型ET估算。平均气温、温差、相对湿度较小的地区可用HS法。日照时数、平均气温、相对湿度较大的地区与风速变化较小的地区可用PT法。相对湿度较小的地区可用IA法和PVB法。

In order to solve the limitation of large-area ET estimation when data are lacking,this paper evaluated six ET estimation methods in Hetao irrigation district with Kriging spatial interpolation algorithm.Based on the Kriging algorithm and meteorological data,the ET values of the Hetao irrigation district were obtained from seven ET estimation methods.The measured ET values of the test station and the ET estimation results of the six methods were verified by the accuracy.The PM calculation results were used as the standard to evaluate the fusion Kriging spatial interpolation.The accuracy and applicability of different methods of the algorithm in the Hetao irrigation district were analyzed,and the differences in the relationship between different methods and meteorological factors were analyzed.The results showed that the comprehensive method(KP,PVB)and PM had the closest results,and the maximum deviation of the calculation results was 0.045-0.490 mm·d^-1.The proposed comprehensive method was the best estimation method for the estuary irrigation area ET based on Kriging spatial interpolation algorithm,which was applicable to the ET estimation of all land types in the irrigation area.The HS method could be used in areas where the average temperature,temperature difference,and relative humidity were small.The PT method could be used in areas with large hours of sunshine,average temperature,and relative humidity,and areas where wind speed changes were small.The IA method and the PVB method could be used in areas with relatively low relative humidity.