Modeling the Effects of Land Use and Land Cover Changes on Desertification Intensity in Mokhtaran Plain Using CA-Markov Method

Abstract

Land use and land cover (LULC) is one of the most important factors that affect the desertification risk in this study, the desertification sensitivity of the Mokhtaran basin in South Khorasan province was estimated by integrating LU/LC scenarios with the MEDALUS environmentally sensitive areas (ESAs) model to predict the desertification risk. The four main MEDALUS criteria including soil, climate, vegetation, and management were examined to assess the sensitive areas to the desertification. Land use maps were categorized using the Landsat satellite imageries of TM, ETM+, and OLI sensors for 1987, 1998, 2003 as a past scenario, and 2015 as a current scenario. Land use maps for 2025 and 2035 were produced as the future scenario based on the simulation of CA-Markov models. The validation results confirmed the model accuracy by calculating the kappa coefficient of 0.95. The land use map was predicted for the years 2025 and 2035 based on the transition rules and a transition area matrix. The results showed that the rainfed area was reduced by 68.29 km2 and the agricultural land was increased by 25.35 km2 during theperiod. In the protection area of playa-bare lands, the changes showed this area was increased by 26.86 km2. The rangeland has also experienced positive changes with an increase of 18.83 km2. Compared to the current scenario, the desertification trend in the future scenario was positively predicted by increasing the area of critical areas from 30.9% to 48.7% over 20 years. The most susceptible lands to desertification were known as playa-bare lands.

Keywords

• land use
• simulation
• medalus
• remote sensing
• markov