The Response Of Evapotranspiration To Changes In Land Use/Land Cover And Climate And The Model Heterogeneity Error Correction

Land surface evapotranspiration(ET)is a central component of the Earth's global energy balance and water cycle and a key process that links the atmosphere,hydrosphere,and biosphere.Timely and accurate understanding ET is important in quantifying the impacts of human influences on the hydrological cycle and thus helps improving water use efficiency and strengthening water use planning and watershed management.The most important factors influencing ET changes in recent years are land use and land cover change(LUCC)due to human activities and climate change.However,their contributions to ET are still with large uncertainties.China has experienced tremendous LUCC as a result of urbanization and ecological restoration under a broad background of climate changes.There is an urgent need to fully understand environmental effects on water resources for science-based management and rational allocation of water resources.But due to environmental changes caused by the surface heterogeneity effect is aggravated in some regionals,this problem makes the remote sensing ET simulation produce a large error,and can not be ignored.How to solve the problem of heterogeneous error have also become the difficulties of remote sensing evapotranspiration estimates in the current study.Based on these,the above issues were studied.Specific research work and conclusions of this paper are as follows:(1)This study used MODIS data products to analyze the temporal and spatial variations of ET in China in the period 2001-2013.Results showed that the average annual ET in China decreased at a rate of-0.6 mm/yr from 2001 to 2013.Spatially,average annual ET increased gradually from the northwest inland areas to the southeast coastal areas.The significant decrease in ET was mainly confined to the northwest China,the central of southwest China,and most regions of south central and east China.(2)We explored the separate contribution to the estimated ET changes by combining LUCC and climate data and identified the leading factors of each factor.The trends of four climatic factors including air temperature,wind speed,sunshine duration,and relative humidity were determined,while the contributions of these four factors to ET were quantified by combining the ET and climate datasets.Our study shows that TA in China has contributed positively to ET changes since the beginning of the twenty-first century,but that the negative contributions of WD and SD outweighed the positive influence of TA.Thus,ET still showed a decreasing trend even in the context of obvious global warming.Among the four climatic factors,sunshine duration and wind speed had the greatest influence on ET.Sunshine duration was the dominant influence on changes in ET in China.LUCC data from 2001 to 2013 showed that forests,croplands and grasslands show a trend of mutual transformation in China,the overall area of forests and croplands increased,and the area of grasslands decreased.The ET effects from changes from forests to other land cover types were much greater than those from other land cover types to forests.Our study suggests that,in a certain area,the effects of deforestation on ET cannot be offset by the effects of afforestation or reforestation.(3)Through quantitative separation of the distinct effects of climate change and LUCC on ET,we conclude that the changes in ET were mainly controlled by climate change in different regions of China during the period 2001-2013.The effects of climate change on ET were more significant than the effects of LUCC on ET,although LUCC was more influential in some years than in other years.Climate change had a dominant effect on change in ET in China as a whole during the period 2001-2013.Effective water resource management and vegetation-based ecological restoration efforts in China must consider the effects of climate change on ET and water availability.(4)This study applied the widely used the Surface Energy Balance Algorithm for Land(SEBAL)and the Surface Energy Balance System(SEBS)models to estimate surface ET in a small agricultural watershed in Jiangxi Province for the surface heterogeneity research sample.The differences between model parameters and energy components were analyzed and compared.The two models were validated by using the Bowen Ratio Energy Balance System on site scale and the Large Aperture Scintillometer on regional scale.On the regional scale,the relative precision of SEBAL and SEBS were 86.75%and 79.32%,respectively.On the site scale,by using a map of land-cover types,it was found that the ET trend for different land-cover types obeyed the following order:paddy fields>orange groves>dry farmland.Howerer,the result showed that both models overestimated pixels in paddy fields and orange groves and underestimated them in dry farmland.(5)We infer that the model estimation error was mainly due to surface heterogeneity caused by mixed pixels.The linear spectral mixture model was introduced in this research to unmix the image.A set of overdetermined equations was defined to improve the estimation accuracy.The revised results show that the problem of overestimating and underestimating various pixels in the two models has been better solved.The statistical distribution of regional average evapotranspiration is more concentrated and close to the measured value.The relative precision of improved SEBAL and SEBS were 94.62%and 89.38%,respectively.The findings from this research are expected to provide new ideas for future development of accurate remote-sensing ET model estimates.