江苏地区不同参考作物蒸发蒸腾量估算模型

为了研究不同参考作物蒸发蒸腾量ET₀估算方法在江苏地区的适用性,收集了江苏省徐州市、高邮市和昆山市1957年1月至2019年12月的气象数据,采用12种不同模型估算了各站点的ET₀,其中模型Priestly-Taylor,Hansen,Jensen-Haise,Makkink是基于辐射数据的模型;MC-Cloud,1985 Hargreaves,Thornthwaite是基于温度数据的;Copais,Valiantzas 1和Valiantzas 2是综合法模型;XGBoost和SVM是机器学习模型.12种ET₀的估算模型计算值分别与Penman-Monteith模型(PM)计算值进行比较,结果表明:各站点的综合评价指数GPI最高的为机器学习模型中的SVM模型;在输入参数相同的情况下,机器学习模型模拟精度优于综合法和温度法以及辐射法中的Pristley-Taylor和Makkink模型;机器学习模型随着输入参数减少,模拟精度依次降低.研究结果可以为江苏地区气象数据不完善时估算ET₀提供科学依据.

Estimation model of evapotranspiration(ET0) of different reference crops in Jiangsu area

To study the applicability of ET₀ estimation methods for different reference crops in Jiangsu area, this study collected meteorological data from January 1957 to December 2019 in Xuzhou site, Gaoyou site, and Kunshan site, Jiangsu Province, and used 12 different models to estimate the refe-rence crop evapotranspiration(ET₀)at each site. Among the estimation models, Priestly-Taylor, Hansen, Jensen-Haise and Makkink are models based on radiation, MC-Cloud, 1985 Hargreaves and Thornthwaite are based on temperature, Copais, Valiantzas 1 and Valiantzas 2 are integrated me-thods, SVM and XGBoost are machine learning models. The calculated values of 12 models for estimating ET₀ were compared with the Penman-Monteith model(PM). The results showed that: The SVM model has the highest GPI(comprehensive evaluation index)value of the three sites. With the same input parameters, the simulation accuracy of the machine learning model is better than that of Priestley-Taylor and Makkink models in the synthesis method, the temperature method, and the ra-diation method. As the input parameters of machine learning model decrease, the simulation accuracy of the machine learning model decreases in turn. The above research results can provide a scientific basis for estimating ET₀ when the meteorological data in Jiangsu area are imperfect.