Hiroshima Institute of Technology (HIT) in Japan has established a LANDSAT-7 Ground Station in cooperation with NASDA for receiving and processing the ETM+ data on March 15 th, 2000 in Japan. The authors performed a verification study on the surface temperature derived from thermal infrared band image data of LANDSAT 7/Enhanced Thematic Mapper Plus (ETM+) for the estimation of temperatures around Hiroshima city and bay area in the western part of Japan as a test site. As to the thermal infrared band, the approximate functions for converting the spectral radiance into the surface temperature are estimated by considering both typical surface temperatures measured by the simultaneous field survey with the satellite observation and the spectral radiance observed by ETM+ band 6 (10.40–12.50μm), and then the estimation of the surface temperature distribution around the test site was examined.In this study, the authors estimated the surface temperature distribution equivalent to the land cover categories around the test site for establishing a guideline of surface temperature detection by LANDSAT7/ETM+ data. As the result of comparison of the truth data and the estimated surface temperature, the correlation coefficients of the approximate function referred to the truth data are from 0.9821 to 0.9994, and the differences are observed from +0.7 to −1.5°C in summer, from +0.4 to -0.9 *C in autumn, from -1.6 to −3.4°C in winter and from +0.5 to -0.5C in spring season respectively. It is clearly found that the estimation of surface temperature based on the approximate functions for converting the spectral radiance into the surface temperature referred to the truth data is improved over the directly estimated surface temperature obtained from satellite data. Finally, the successive seasonal change of surface temperature distribution pattern of the test site is precisely detected with the temperature legend of 0 to 80'C derived from LANDSAT-7/ETM+ band 6 image data for the thermal environment monitoring. 2003 COSPAR. Published by Elsevier Ltd. All rights reserved.
Read full abstract