| Evapotranspiration is an important segment in the hydrological cycle between soil,vegetation and atmosphere,and evapotranspiration plays a crucial role in maintaining surface water balance and surface energy balance.Quantitative and accurate estimation of regional surface evapotranspiration not only can provide a powerful theoretical basis for analyzing regional climate change,hydrological and ecological changes,but also agricultural irrigation and water resources allocation and management.China is a traditional agricultural country,and the rational allocation of agricultural water resources is essential.It is difficult to monitor the evapotranspiration of large areas by traditional methods,because of complex terrain conditions.Therefore,with the help of remote sensing technology to overcome the limitations of traditional methods,evapotranspiration monitoring can be expanded from point scale to surface scale.It provides a reliable guarantee for the real-time and periodic regional acquisition of regional surface evapotranspiration data.This article took Liaocheng City as the research area,and the article was based on the SEBS model which developed by the theory of surface energy balance.The regional meteorological data,DEM elevation data,MODIS remote sensing data and Landsat8 remote sensing data were took as data sources.And Arc GIS,ILWIS and Envi multi-platforms were used to estimate the evapotranspiration data of typical days from March to September.Surface albedo,surface emissivity,surface temperature,normalized vegetation index(NDVI),surface net radiation,sensible heat flux,latent heat flux,the soil heat flux and evapotranspiration of wheat and corn in Liaocheng City were analyzed.The data of Xinxian station was combined with the FAO Penman-Monteith formula to verify the accuracy of SEBS model.The evapotranspiration estimated by the model was extended to the daily and monthly scales,and the long-term regional evapotranspiration distribution was analyzed.At the same time,the correlation between evapotranspiration and key parameters such as meteorological parameters and surface parameters was further discussed.Through the analysis and discussion,the main conclusions of this paper are as follows.(1)The evapotranspiration in agriculture fields of Liaocheng changed obviously with the change of seasons.In 2019,the daily evapotranspiration in the farmland area of Liaocheng City was more consistent with the spatial distribution of NDVI while most of the lowest values were found at the boundary of non-farmland areas.The daily evapotranspiration fluctuated more widely in wheat season,the temporal and spatial distribution of evapotranspiration was significantly affected by the change of climatic conditions,land use types and crop types.According to the time-scale expansion analysis of evapotranspiration,the change of evapotranspiration in the study area from 2015 to 2019 had a significant periodicity,and the annual fluctuation curve presented a bimodal change pattern.In January and February,the evapotranspiration remained at a low level,less than 1mm/d.And in March,evapotranspiration increased rapidly with the rapid growth of wheat,and reached the peak value in May,close to 5mm/d.The evapotranspiration decreased in June,but it fluctuated in July and August and rose to a new peak while its value was lower than the peak in May.When maize ripened,the evapotranspiration gradually decreased and remained at a low level in November and December.(2)Temporal and spatial variations of surface parameters such as albedo,emissivity,temperature,NDVI,liquid radiation,sensitive heat flow and latent heat flow have been analysed;Heat flow from soil and evapotranspiration of wheat and maize in Liaocheng City were analyzed,while the spatial distribution of soil heat flux and sensible heat flux was contrary to the NDVI distribution.The fluctuation range of surface albedo was small,and it would not fluctuate greatly with the change of seasons and crop types.The trend of surface net radiation over time was exactly opposite to the trend of surface albedo over time,and the distribution of net surface radiation values under wheat cover was more uniform than that under maize.In the wheat season and the corn season,there was a strong correlation between the surface temperature and air temperature,and the R~2of linear fitting were 0.8812 and0.9050,respectively.(3)FAO Penman-Monteith formula and crop coefficient were used to calculate actual crop evapotranspiration to verify the accuracy of model inversion values.The correlation coefficient R,R~2,root mean square error(RMSE),mean square error(MSE)and mean absolute error(MAE)between the value of calculation and model were 0.825,0.6807,0.601mm/d,0.361mm/d and 0.493mm/d respectively.The significance value(Sig)was 0.004,indicating that the fitting effect is good.Since the correlation coefficient R of linear fitting for wheat season and maize season was 0.837 and 0.851,respectively,the fitting effect of remote sensing estimation of evapotranspiration in maize season was better than that of wheat.(4)Correlation between evapotranspiration and meteorological parameters and surface parameters after standard evaporation treatment was analyzed by using SPSS.As a result,temperature,wind speed,air pressure,sunshine time,ground surface temperature,NDVI and surface net radiation were significantly correlated with evapotranspiration,and surface albedo significantly correlated with evapotranspiration.Among all parameters,only air pressure and surface albedo showed negative correlation with evapotranspiration. |
PDF Full Text Request