Study On The Urban Surface Temperature By Fusing NPP-VIIRS And Landsat-8 Thermal Infrared Band Data

Land surface temperature(LST)is a key parameter of great interest in research and application of surface energy balance and heat exchange.It is widely used in many fields such as agriculture,geology,hydrology and meteorology.The thermal infrared remote sensing data reflects the thermal radiation on the surface and is widely used for predicting and monitoring of surface temperature.However,to our knowledge,no single satellite system can produce thermal infrared(TIR)images at both high spatial and temporal resolution to retrieve LST till now.Therefore,various algorithms have been developed to enhance the spatial or temporal resolution of TIR data with high temporal resolution but low spatial resolution(e.g.MODIS)and low temporal resolution but high spatial resolution(e.g.Aster).VIIRS is one of five instruments carried by the Suomi National Polar-orbiting Partnership(S-NPP)satellite which was launched in 2011.This National Oceanic and Atmospheric Administration/National Aeronautics and Space Administration(NOAA/NASA)program was tasked to meet the requirements for the next generation of environmental research.The onboard 22-band VIIRS sensor builds on MODIS and AVHRR heritage,and high radiometric accuracy and spatial resolution multispectral imagery can be provided.To our knowledge,Visible Infrared Imaging Radiometer(VIIRS)TIR data have not yet been used in thermal downscaling with Landsat-8 TIR data.In order to verify the feasibility of VIIRS data for spatial thermal infrared data fusion and its application prospects,this paper proposes the fusion of VIIRS and Landsat-8 TIR data,the fusion results were then converted to LST.Landsat-8 TIR data have a spatial resolution of 30 m with a revisit frequency of 16 days while VIIRS provides daily TIR data at a spatial resolution of 375 m.This study aims to generate daily LST images at Landsat-8 resolution(30 m)by fusing VIIRS and Landsat-8 TIR data for the first time.Among the algorithms to enhance the spatial or temporal resolution of TIR data,the Spatio-temporal Adaptive Data Fusion Algorithm for Temperature mapping(SADFAT)data fusion model is considered to be more efficient,which is also employed in this study.Case studies were conducted to predict LSTs in Beijing,Urumqi and Yinchuan,China.The predicted LSTs were compared against the observed LSTs obtained from the Landsat-8 TIR data using the single channel algorithm.The results of study area(a)indicate that the prediction accuracy for the study area ranged from 1.1 K to 1.4 K,and the correlation coefficients between the observed and predicted LSTs varied from 0.71 to 0.89.This study suggests that ?RS data can be used as a good alternative to MODIS data for generating daily LST images by fusing Landsat TIR data.In order to verify the application of SADFAT model in VIIRS and Landsat-8 thermal infrared data fusion,this paper analyzes the sensitivity of the model parameters,we change the conversion coefficient in the model from the pixel-to-pixel ratio difference to the mean conversion coefficient of each land cover type in the moving window,to get smoother and more accurate results.At the same time,through experimental verification,we found that when the size of the moving window is 5×5,we can get the fusion result with better precision and higher efficiency.