Evaluation of integrated CA-Markov model and fuzzy model in predicting land use changes in southwestern region of Isfahan province
Ahmad Amiri
1
(
Master student, Department of Spatial Information Systems and Remote Sensing, Lanjan Branch, Islamic Azad University, Isfahan, Iran
)
saleh abdullahi
2
(
Assistant Professor, Department of Civil Engineering, Isfahan (Khorasgan) Branch, Islamic Azad University, Isfahan, Iran
)
, Ghasem Khosravi
3
(
Assistant Professor, Department of Spatial Information Systems and Remote Sensing, Lanjan Branch, Islamic Azad University, Isfahan
)
Keywords: Land use changes, Cellular Automata, fuzzy model, Remote sensing, GIS,
Abstract :
Uncontrolled and unorganized urban growth have several negative effects on our surrounding natural environment. One of the main problems is the change of valuable natural and agricultural lands to urban, industrial and residential area. This study aimed to extract the past land use changes and predict future changes in the southwestern regions of Isfahan province over a 30-year period using a hybrid model of cellular automata (CA) and fuzzy logic. Landsat 7 ETM+ and OLI Landsat 8 images for 2003-2013-2022 were used together with topographic and geological maps of the region as preliminary data. Satellite images were classified and validated into five land uses classes including agriculture, forest, urban, water and barren lands. The CA model then predicted and produced land use maps for 2023 and 2033. Next, to evaluate the suitability of the projected land use in each location, the selected parameters were mapped based on their nature and role in the study area and the pred, sum and gamma values of 0.7, 0.8, 0.9 were produced using the different operators of the fuzzy model. Combining the results of the two models showed that, 0.98 km2 of urban areas were in very low, 2.63 km2 in low, 20 km2 in moderate, 7.21 km2 in high and 6.12 km2 in the very high suitability class for 2033. Moreover, 35 % of the study area is suitable for urban growth and more than 36 % of the project urban land use are located in these areas. In contrast, 54 % of the project urban land use are located in the class of moderate suitable, and finally about 10 % the project urban land use is located in the class of not suitable, which should be the main concern of the local government. In addition to these statistical results, the process showed that the integration of these two models creates a proper and accurate understanding of future urban changes for urban planners and decision makers.