Generating Remotely Sensed Land Surface Temperature With High Temporal Resolution Based On Diurnal Temperature Cycle Model

Land surface temperature(LST)has crucial research significance in climate,hydrology,ecology,biogeochemistry and lots of other research fields.In addition,LST is also one of the significant parameters at the interface between the earth's surface and the atmosphere.Remote Sensing is a unique way to measure the diurnal cycle of LST over extended regions.This thesis aims to develop methods for generating land surface temperature with high temporal resolution from remote sensing based on the diurnal temperature cycle(DTC)models.(1)Comparison and analysis of the temporal normalization methods of Remote Sensing land surface temperature.Temporal normalization method based on the diurnal temperature cycle(DTC)model and the relative normalization methods(RN Methods),i.e.,histogram match(HM),linear regression(LR),and polynomial regression(PR)methods,were compared using airborne LSTs of Heihe Watershed Allied Telemetry Experimental Research(HiWATER).The limitations of these two methods were analyzed.The relative radiance temporal normalization results show that the best accuracies are obtained for the HM and LR methods,whereas the worst accuracies are achieved for the PR method.While compare with the results of the DTC based method,the DTC based method is better than the RN methods.(2)Improvement and development of DTC models.This paper summarizes and deduces the equation of improved diurnal temperature cycle(IDTC)model based on the DTC model and combined with the surface energy balance equation and the soil heat conduction equation.In addition,we then established the definite IDTC model that analyzes the influence of meteorological factor(i.e.,wind,air temperature,air humidity,atmospheric pressure,and wind direction)and fluctuations that cannot be measured by the DTC model.(3)The validation and evaluation of IDTC model.Based on the ground measured LST and ASTER LST,we verify and evaluate the temporal normalization of TASI LST.Results show that: first,the accuracy of IDTC model is better than the DTC model;second,the TASI LST is more consistent with ground measured LST compared with the results normalized by DTC model;third,the TASI LST normalized by IDTC model is more consistent with the ASTER LST visually,especially in Gobi desert and desert stepple.