The Study On Methodology For Spatialization Of Instantaneous Air Temperature Under Clear Sky Based On RS And GIS

Near surface air temperature is a key factor in energy and water exchanges between land surface and atmosphere, and it is an important input parameter in many models. It is of great significance to obtain spatial distributions of the near surface instantaneous air temperature when there are rare or no traditionally meteorological observations.Using Jiangsu province as the study area,the paper compared precisions of the ordinary kriging interpolation,multivariate linear regression analysis,temperatu re-vegetation index method.The surface energy balance model used for retrieving near surface air temperature.and then analyzed the characteristics and variations of the error distribution from the three methods.Furthermore,the applicabilities of different simulation methods were discussed.Using the modeling time data,Results show that the error of the ordinary kriging interpolation method is minimum with its mean absolute error (MAE) and root mean square error(RMSE) are0.47℃and0.62℃respectively. The retrieval precision of the surface energy balance model is relatively lower with its MAE and RMSE are2.14℃and2.82℃respectively.The error of the temperature-vegetation index method is the lowest with its MAE and RMSE are2.49℃and3.32℃respectively.Using the validation data, Results show that the precision of the surface energy balance model is minimum with its MAE and RMSE are increased by0.19℃and0.20℃.On this basis,this paper discusses that these three methods have different applicabilities in various data sources and underlying land surface conditions. Ordinary kriging interpolation method require data from meteorological stations,which have great dependence on the number and distribution of meteorological stations.Especially when the multiple linear regression analysis method is applied to other study period, the inapplicability of the model is enhanced, and the precision of the result changes greatly.TVX method can use in a small amount of meteorological stations in the study area.If the meteorological stations is little or not and vegetation coverage is low in the study area, we can consider the surface energy balance model for the temperature inversion, the physical meaning of the model is clear,and the model has good portability and universality.