Remote Sensing Estimation Of Evapotranspiration Of Main Crops In Typical Irrigation Area Based On SEBAL Model

In order to discuss the evapotranspiration of main crops in Wuchang Minyi Township,Longfengshan Irrigation Area,Heilongjiang Province,and Leshan Irrigation Area,Qingyi River Basin,Sichuan Province,Sentinel 2 image data was used as the data source for land use classification.Land use classification effect of decision tree classification method and decision tree classification method based on multi-temporal remote sensing data in Qingyi River Basin.Using Landsat 8 satellite imagery as the data source,the SEBAL model was used to invert the daily evapotranspiration of Minyi Township and Leshan Irrigation District.Based on the classified crops,the daily evapotranspiration inversion of each crop was obtained.In order to verify the inversion results,based on the MODIS evapotranspiration products and meteorological data,the daily evapotranspiration results calculated by the Penman formula shall prevail,and the inversion results are verified and evaluated.The factors affecting daily evapotranspiration and the causes of errors in daily evapotranspiration are analyzed and studied.In the method of expanding from the daily scale to the monthly scale,the method of multiplying the daily evapotranspiration by the number of days in the current month and the crop coefficient method is used to obtain the monthly evapotranspiration of the crop,and the results in the literature are verified to verify these two methods.Conducted evaluation and analysis.The following conclusions are drawn from this:(1)Based on the Sentinel 2 image,for the fifth-generation classification of Leshan Irrigation District in Qingyi River,choosing the decision tree classification method based on multi-temporal remote sensing data can better extract the crop planting information of the irrigation district.(2)Use the SEBAL model to simulate Minyi Township on May 10,2016,June 27,2016,August 14,2016,July 16,2017,September 2,2017,and April 2018 The daily evapotranspiration of rice and corn on the 30th and July 6,2019 were:4.24mm and 3.56mm,5.91mm and 5.34mm,4.56mm and 4.55mm,5.69mm and 4.78mm,3.27mm and 3.18mm,3.40 mm and 2.25mm,5.84mm and 5.05mm.The simulation results and Penman formula calculation results are verified to find that the errors are less than 1mm.It shows that the simulation results are acceptable.The SEBAL model simulation results show that the verification error of crop daily evapotranspiration in Leshan Irrigation is mainly concentrated within 0?1mm,and a few times between 1?2mm.Considering the complicated planting conditions in Leshan Irrigation Area,the simulation results are acceptable.During the verification process,it was found that evapotranspiration products with a too small spatial resolution(such as MOD 16 at 500m)are not recommended to study the daily evapotranspiration of different crops in the same area and the differences in daily evapotranspiration between them(3)In the scale expansion method from daily to monthly evapotranspiration,there are not small errors in the daily evapotranspiration calculated by using the crop daily evapotranspiration multiplied by the number of days in the current month and the cumulative crop coefficient;The determination method is not further improved,the accuracy of the former is better than the latter;the monthly evapotranspiration directly obtained by multiplying the number of days is greater than the result calculated by the crop coefficient method.