基于天气预报信息的参考作物需水量预报研究  

Research on the Reference Crop Evapotranspiration Forecast Based on Weather Forecast Information

作　　者：白依文 鲁梦格 程浩楠 严冬[1] 孙怀卫[1] BAI Yi-wen;LU Meng-ge;CHENG Hao-nan;YAN Dong;SUN Huan-wei(School of Hydropower and Information Engineering,Huazhong University of Science and Technology,Wuhan 430072,China)

机构地区：[1]华中科技大学水电与数字化工程学院,武汉430074

出　　处：《中国农村水利水电》2021年第10期175-179,184,共6页China Rural Water and Hydropower

基　　金：国家重点研发计划项目(2018YFC1508301);国家自然科学基金项目(51879110);湖北省水利重点科研项目(HBSLKY201818,HBSLKY201907,HBSLKY201918)。

摘　　要：获得作物需水信息是农业灌溉管理与水资源优化配置的基础。为探索利用公共天气预报信息进行作物需水预测,研究尝试以PM公式和HS公式为基础,将天气预报信息进行定量转化后预测参考作物需水量,以期解决典型地区参考作物需水量预报难题。通过以武汉市为典型研究区,收集了武汉站2020年3月16日至2020年4月15日的每日对未来七天的天气预报数据,使用天气预报数据与PM公式和HS公式得出的预报值与使用历史监测数据和PM公式采用实测数据得出的标准值进行对比并对本文使用的方法进行了验证。结果表明,不同预报期内,预报值与标准值有一致的变化趋势,两种方法在1~2 d预报期的MAE平均值分别为0.69和0.675 mm/d;RMSE平均值分别为0.91和0.905 mm/d,基本满足需水量精度要求,而随着预报期的增加准确性在降低;两种方法的对比显示,HS法的相关系数总是相对较高且预报后期比PM法稳定,PM主要误差是由实际日照时数导致。当实际日照时数预报精度较差时推荐使用HS法,此方法具有良好的物理基础并且数据容易获取。Obtaining crop water requirement information is the basis of agricultural irrigation management and optimal allocation of water re⁃sources.In order to explore the use of public weather forecast information to predict reference crop evapotranspiration,this paper is an at⁃tempt to use PM model and HS model as the basis for quantitatively transforming weather forecast information to predict reference crop evapo⁃transpiration so as to solve the problem of water demand forecast in typical areas.By using Wuhan as the typical study area,the daily forecast data of Wuhan Station from March 16,2020,to April 15,2020,daily for seven days in the future are collected.The results show that the predicted value and the standard value have the same trend in different forecast period,MAE of the two methods in 1~2 days forecast periods are 0.69 and 0.76 mm/d,RMSE are 0.91 and 0.905 mm/d respectively,which basically meet the precision requirements of water demand,while with the increase of forecast period,the accuracy is decreasing.The comparison between the two methods shows that the correlation co⁃efficient of HS is always higher than that of PM and the prediction during later period is more stable.The main PM error is caused by the actu⁃al sunshine hours.HS method is recommended when the prediction accuracy of actual sunshine hours is poor.This method has good physical basis and data is easy to obtain.

关 键 词：参考作物腾发量 公共天气预报 Penman-Monteith法 Hargreaves-Samani法 

分 类 号：TV93[水利工程—水利水电工程] S274.1[农业科学—农业水土工程]