Study of soil salinity in the Ardakan area, Iran, based upon field observations and remote sensing

The main purpose of this study is to evaluate the capability of TM imagery for mapping land cover types and to study soil salinity conditions in the Ardakan area located in the Central Iranian Deserts. A Landsat TM image recorded on September 11th 1990 was used. In this study some soil salinity problems were enhanced. Consequently the field work and sampling strategy were designed and the training areas were selected. In a first attempt to classify the image, the TM band 6 was neglected. A maximum likelihood classification was done using the TM bands 3, 4, 5, and 7. The results obtained showed that a great confusion between some classes prevented a high classification accuracy. Therefore it was decided to include also the thermal TR band in the band combination. The result of the image classification based on the combination of TM bands 3, 4, 5 and 6 showed a large improvement of the classification accuracy. The result obtained revealed that the TM thermal band plays an important role in separating: a) saline soil from gypsiferous soil, b) saline soil from non-saline soil, and c) urban area from dark to brownish salt crust. The behaviour of the TM thermal and TM reflective bands on the land cover types was compared through the relationship between: a) TM thermal and TM reflective bands, and b) soil salinity and digital numbers. Based on the results we may conclude that the behaviour of the TM thermal and TM reflective bands is highly dependent on the type of land cover