基于温湿度的ET_0估算模型应用研究  

Reference Evapotranspiration Estimation Model Based on Temperature and Humidity

作　　者：柳烨[1] 赵文刚[1] 杨珮珮 马孝义[1] 张丽[1] 

机构地区：[1]西北农林科技大学水利与建筑工程学院,陕西杨凌712100

出　　处：《灌溉排水学报》2016年第2期35-39,共5页Journal of Irrigation and Drainage

基　　金：国家科技支撑计划项目(2012BAD08B01);国家自然科学基金项目(51279167);水利部公益性行业专项(201301016)

摘　　要：针对大面积灌区作物蒸发蒸腾量(ET0)分布式监测所需参数较多的问题,开展利用易获取的少量气象参数估算ET0的研究对我国灌区的作物需水量监测和灌区水资源的管理有重要意义。利用人工神经网络技术建立基于温、湿度的ET0月份估算模型,对作物蒸发蒸腾量进行了估算;在此基础上,针对ET0的季节性特征,将估算模型由月份尺度拓展到季节尺度;最后运用陕西省6个基本站点的气象数据对该优化模型进行普适性分析。结果表明,优化后的季节估算模型,在春夏二季的平均相对误差在10%以内,在秋冬二季的平均绝对误差在0.20mm以下,且每对基准站点和邻近站点得到的估算结果具有很好的一致性和稳定性,表明该模型在作物需水估算方向上较强的实用推广价值。In order to solve the problem of large meteorological factors for distributed ET0 monitoring in large irrigation area,how to use a small number of parameters to estimate the ET0 becomes important for water resources management in Chinese irrigation district.This paper set up the reference evapotranspiration estimation model via temperature and humidity based on the standard artificial neural network in the month scale initially.On this basis,this estimation model was extended from the month scale into the season scale based on the seasonality of ET0 values.Finally,this optimized estimation model of ET0 applied and made universal analysis in six representative stations of Shaanxi for example.Results show that the average relative error was less than 10%in spring and summer;the mean absolute error was below 0.20 mm in autumn and winter.More importantly that every estimation result in every pair of stations which included a basic station and an adjacently station could maintain an excellent consistency and stability.In summary,this estimation model confirmed the practical value and had application potential in the direction of forecasting crop water requirements.

关 键 词：参考作物需水量 BP神经网络 估算精度 普适性 灌区信息化 

分 类 号：S274[农业科学—农业水土工程] S311[农业科学—农业工程]