Advances in the Study of Near Surface Air Temperature Retrieval from Thermal Infrared Remote Sensing

Abstract

Near surface air temperature is an important environment variable in many earth system models,because it is a key factor in the energy and water exchanges between land surface and atmosphere.Detailed measurements of spatial and temporal variations of near surface air temperature are critical for the effective understanding of climate,hydrology,ecology,agriculture and terrestrial life processes.Traditionally meteorological observation could provide accurate air temperature data at the point scale,but most earth system models need gridded input variables.Satellite remote sensing provides a straightforward and consistent way to observe air temperature at regional and global scales with more spatially detailed information than meteorological data.This paper systematically reviews the air temperature retrieving algorithms for thermal remote sensing data,which include TVX approaches,statistical approaches,neural network approaches and energy balance approaches.The main advantages and limitations of these four methods are also discussed.Finally,the development tendencies of estimating air temperature by remote sensing are pointed out,such as intensive research on thermal radiant transfer model,spatial-temporal scaling of air temperature and improvement of cloud detection.