Predicting Land Use Land Cover Dynamics and Land Surface Temperature Changes Using CA-Markov-Chain Models in Islamabad, Pakistan (1992–2042)

Cellular automata (CA) models are employed for simulating geographical distributions, while Markov-Chain models are utilized for simulating temporal changes. This study aims to forecast the dynamics of land use and land cover (LULC) change and land surface temperature (LST) using a CA and Markov-Chain model for the period 1992–2022 in Islamabad, Pakistan. This research innovatively predicts future LULC and LST and examines their correlations with various vegetation indexes. LULC maps were generated from time series data of Landsat satellite images (Landsat 5, 7, and 8) for the years 1992, 2002, 2012, and 2022, utilizing the random forest algorithm. Five satellite indexes were employed: normalized difference vegetation index (NDVI), bare soil index (BSI), normalized difference water index (NDWI), normalized difference built-up index (NDBI), and soil adjusted vegetation index (SAVI). Through the CA-Markov-Chain analysis, the results showed significant new findings, such as a rise in the built-up area of Islamabad, which increased from 105.63 km2 (11.66%) in 1992 to 447.39 km2 (49.38%) in 2022. The study predicted that the built-up area would further increase to 531.82 km2 (55.38%) by 2042. The study analyzed the enhancement of LST, which was about 2.40 °C from 1992 to 2022, ultimately because of the expansion of uncontrolled urban areas. The correlation of LST with the vegetation indexes NDVI, BSI, NDWI, NDBI, and SAVI was also analyzed through regression analysis. Furthermore, the surface temperature was predicted well by the urban index (UI), a nonvegetation index, demonstrating the positive correlation of R2 = 0.87 with respect to retrieved surface temperature. Using the UI as a predictor of LST, our projections indicate that regions with temperatures ranging from 20 to 24 °C and from 25 to 28 °C will decrease in coverage from 4.63% to 3.85% and from 25.33% to 21.08%, respectively, between 2032 and 2042.