-

Abstract

Landsat 5 and Landsat 7 were launched on March 1, 1984 and April 15, 1999 respectively. Since bands 1-5 and 7 have a matching spatial resolution (30? 30) and the only difference is the resolution of the thermal band (band 6), so their efficiency and information content can be compared in a particular location. Damghan playa was selected as the first case where land cover change during the study was observed and Kashan playa was selected as the second area where no change observed. Tabriz was selected as the third study area. In Damghan playa ETM+ image acquired on July 20, 2000 and Landsat 5 TM image dated May 30, 1998 were used. In Kashan area ETM+ image dated August 9, 2002 and Landsat 5 TM data acquired on September 5, 1988 were used. ETM+ and TM images acquired August 2, 2001 and August 18, 1998 were used in the Tabriz area. Initially acquisition time of Landsat 5 and 7 were studied. Then for accurate comparison, these images were geometrically corrected and geo-referenced. Using statistical and mathematical techniques the effect of land cover changes were removed and the correlation between bands were examined in both processed and unprocessed images. In Kashan playa, due to land cover changes, only correlation and statistical relationship between bands were studied. In Tabriz, samples were taken from unchanged areas located inside and outside of urban area. The two state ETM+ thermal band (Low and High gain) images were also investigated. The results have shown that in all areas the correlation between thermal band and reflective bands in ETM+ is lower than TM and could be interpreted as richer information contained in the image. The reason for the observed dissimilarity between TM and ETM+ sensors could be attributed to the differences in acquisition time, sun elevation, azimuth and different calibration coefficients. Further studies are required to compare TM and ETM+ information content in a completely identical acquisition time.
Keywords: TM, ETM+, Thermal Band, Information Content, Remote Sensing