摘要
潜在蒸散发(ET0)是估算作物需水量的基础。根据石羊河流域5个气象站5年的气温、风速、相对湿度等日气象要素资料,采用Penman-Monteith公式计算石羊河流域的ET0,建立六因子、四因子和三因子的支持向量机(SVM)模型与人工神经网络(ANN)模型模拟日ET0,对模拟值与计算值进行比较,以均方根误差(RMSE)、平均绝对误差(MAE)、确定性系数(DC)以及皮尔逊相关系数(R)作为模型的性能评价指标,对模型进行检验以获得模拟精度较高的模型。结果表明:相同因子输入下ANN模型较SVM模型在石羊河流域模拟日ET0有着更高的模拟精度。该研究可为气象要素资料不全的站点提供模拟日ET0的可行方法。
Potential evapotranspiration(ET0)is the basis for estimating crop water requirements.Based on the data of daily meteorological elements such as temperature,wind speed,and relative humidity at 5 meteorological stations in Shiyang River Basin for five years,the Penman-Monteith formula is used to calculate ET0 in Shiyang River Basin.Six-factor,four-factor,and three-factor support vector machine model(SVM)and artificial neural network model(ANN)are established to simulate the daily ET0,and then the simulated values are compared with the calculated values.Root mean square error(RMSE),mean absolute error(MAE),deterministic coefficient(DC),and Pearson correlation coefficient(R)are used as the performance evaluation index of the models,and the models are tested to obtain models with higher simulation accuracy.The results show that under the same factor input,the ANN model has higher simulation accuracy than the SVM model in Shiyang River Basin to simulate daily ET0.The study can provide a feasible method for simulating daily ET0 for sites with incomplete meteorological data.
作者
褚江东
粟晓玲
郭盛明
牛纪苹
CHU Jiang-dong;SU Xiao-ling;GUO Sheng-ming;NIU Ji-ping(College of Water Resources & Architecture Engineering, Northwest A&F University, Yangling 712100, Shaanxi, China;Key Laboratory of Agricultural Soil and Water Engineering in Arid Area of Ministry of Education, Northwest A&F University, Yangling 712100, Shaanxi, China)
出处
《节水灌溉》
北大核心
2020年第8期34-39,共6页
Water Saving Irrigation
基金
国家自然科学基金项目(51879222)。
关键词
日潜在蒸散发
人工智能
支持向量机
人工神经网络
石羊河流域
daily potential evapotranspiration
artificial intelligence
support vector machine
artificial neural network
Shiyang River Basin
