论文部分内容阅读
为实现气象资料缺乏情况下参考作物蒸散量(ET0)的精确模拟,利用川中丘陵区3个气象站点1999-2013年的逐日气象资料作为输入量,以FAO-56 Penman-Monteith模型计算的ET0作为标准值,建立基于遗传算法优化神经网络的ET0模拟模型(GA-BPNN),并将其模拟结果同Hargreaves、Mc Cloud、Priestley-Taylor和Makkink等4种常用ET0计算模型的计算结果进行对比。结果表明:GA-BPNN模型能够很好地反映ET0同气象因素之间的非线性关系,模拟精度较高;当基于温度资料模拟ET0时,GA-BPNN模型模拟精度高于Hargreaves和Mc Cloud模型;当基于温度和辐射资料时,GA-BPNN模型模拟精度明显高于Priestley-Taylor和Makkink模型。因此GA-BPNN模型可以作为气象资料缺乏情况下川中丘陵区ET0模拟的推荐模型。
In order to accurately simulate the reference ET0 in the absence of meteorological data, daily meteorological data of three meteorological stations in the central Hilly area of Sichuan Basin in 1999-2013 were used as input and ET0 calculated by the FAO-56 Penman-Monteith model as (GA-BPNN) based on the genetic algorithm to optimize the neural network, and the simulation results are compared with the calculation results of four commonly used ET0 calculation models such as Hargreaves, Mc Cloud, Priestley-Taylor and Makkink. The results show that the GA-BPNN model can well reflect the nonlinear relationship between ET0 and meteorological factors, and the simulation accuracy is high. When simulating ET0 based on temperature data, GA-BPNN model has higher accuracy than Hargreaves and Mc Cloud models. GA-BPNN model simulation accuracy is significantly higher than Priestley-Taylor and Makkink models when based on temperature and radiation data. Therefore, the GA-BPNN model can be used as a recommended model for simulating ET0 in the hilly area of Central Sichuan under the lack of meteorological data.