ESTIMATION OF THE TEMPERATURE DISTRIBUTION IN URBAN AREAS BY USING REMOTE SENSING DATA

The objective of this study is to develop practical methods of translating brightness temperature data obtained from satelites such as LANDSAT and NOAA into air temperature. There are two typical cases where the methods sould be applied. In the first case, reference temperature data observed at a sufficient number of points are available, e. g., from AMeDAS (Automated Meteorological Data Acquisition System), and the brightness temperature data obtained from LANDSAT TM and NOAA can be compared with these data. Then spatial distribution pattern of air temperature can be produced by using the correlation formula obtained by regression analysis. This study examines the method for three areas, i. e., the Kanto, Kansai (each area size being 80km×80km) and North-Kyushu (area size being 40km×40km) areas. In the second case, no reference temperature data are available, and another method is developed by using the one-dimensional energy budget model. Here, air temperature is estimated by assuming that the brightness temperature represents the surface temperature. The applicability of this method is tested for the Kanto region.