JWSS - Journal of Water and Soil Science

In recent decades, land use changes have been one of the most important environmental issues worldwide. This study was carried out to investigate and analyse land use changes in dust sources of south and south-east Ahwaz and surrounding lands using remote sensing technique. Firstly, based on the standard precipitation index (SPI), the years 1986, 2002 and 2016 were selected as years with near normal situation. In the next step, land use maps of the study area were extracted using a Landsat satellite images and supervised classification methods. MNDWI index was used to increase the accuracy of image classification. In order to evaluate the efficiency of each method, the overall accuracy and kappa coefficient were used. Finally, to investigate land use change a post-classification comparison method was employed. According to research findings, in the first period (1986-2002), the area of vegetation, wet land with vegetation and water bodies have increased by 419%, 219% and 40.7%, respectively, in contrast, the area of barren and poor range has reduced by 36.6%. One of the major causes of these changes is human factors such as population growth and climate variables such as proper amounts of precipitation. In the second period (2002-2016), the area of wet land with vegetation, water bodies and vegetation has been decreased by 94%, 49.3% and 46.7%, respectively; in contrast, barren and poor range land has been increased by 45.4%. Recent droughts increase the temperature and incorrect and non-principled management of water resources has been effective in reducing green cover and water bodies and increasing in bare land. The results show that, during the three decades, built-up land has been increased by 157%, while River bed has decreased by 28.8%. These results indicate a total degradation in the region due to climate change and human activities.

The southern regions of Kerman Province have repeatedly encountered dust storms. Therefore, the objective of this study was to identify dust sources using effective parameters such as vegetation cover, land surface temperature, soil moisture, soil texture, and slope as well as to detect dust storms originating from these regions based on 31 MODIS images in 2016 and SRTM data. After normalizing parameters, the dust source map was prepared by fuzzy logic and assessed with an error matrix and available dust source map. Results showed that 30.5% of the study area was classified as a low source of dust, 39.55% as moderate, and 29.85% as severe-very severe. The overall accuracy of the produced map was about 70% and the producer and user accuracy of the severe-very severe class was more than 87%. The detection of dust storms originated from the identified dust sources also confirmed a crisis situation in the region. Due to the repeatability and continuity of obtained dust source map at pixel scale, it can be used to update available dust source maps and manage dust crisis in the region, properly.