In the present paper, land use/land cover (LULC) change was predicted in the Greater Isfahan area (GIA), central Iran. The GIA has been growing rapidly in recent years, and attempts to simulate its spatial expansion would be essential to make appropriate decisions in LULC management plans and achieve sustainable development. Several modeling tools were employed to outline sustainable scenarios for future dynamics of LULCs in the region. Specifically, we explored past LULC changes in the study area from 1996 to 2018 and predicted its future changes for 2030 and 2050. For this purpose, we performed object-oriented and decision tree techniques on Landsat and Sentinel-2 satellite images. The CA-Markov hybrid model was utilized to analyze past trends and predict future LULC changes. LULC changes were quantitatively measured using landscape metrics. According to the results, the majority of changes were related to increasing residential areas and decreasing irrigated lands. The results indicated that residential lands would grow from 27,886.87 ha to 67,093.62 ha over1996–2050 while irrigated lands decrease from 99,799.4 ha to 50,082.16 ha during the same period of time. The confusion matrix of the 2018 LULC map was built using a total of 525 ground truth points and yielded a Kappa coefficient and overall accuracy of 78% and 82%, respectively. Moreover, the confusion matrix constructed base on the Sentinel-2 map, as a reference, to judge the predicted 2018 LULC map with a Kappa coefficient of 88%. The results of this study provide useful insights for sustainable land management. The results of this research also proved the promising capability of remote sensing algorithms, CA-Markov model and landscape metrics future LULC planning in the study area.

在本文中，预测了伊朗中部大伊斯法罕地区（GIA）的土地利用/土地覆被（LULC）变化。近年来，GIA一直在快速增长，尝试模拟其空间扩展对于在LULC管理计划中做出适当决策并实现可持续发展至关重要。使用了几种建模工具来概述该地区土地利用和土地利用的未来动态的可持续方案。具体来说，我们探索了研究区域1996年至2018年过去LULC的变化，并预测了2030年和2050年的LULC未来变化。为此，我们对Landsat和Sentinel-2卫星图像执行了面向对象和决策树技术。CA-Markov混合模型用于分析过去的趋势并预测未来的LULC变化。LULC变化使用景观指标进行定量测量。根据结果​​，大多数变化与增加居民区和减少灌溉土地有关。结果表明，在同一时期，居民土地面积将从1996年至2050年的27,886.87公顷增加到67,093.62公顷，而灌溉土地的面积将从99,799.4公顷减少到50,082.16公顷。2018年LULC地图的混淆矩阵是使用总共525个地面真点建立的，得出的Kappa系数和总体准确度分别为78％和82％。此外，基于Sentinel-2映射构建的混淆矩阵作为参考，以判断Kappa系数为88％的预测2018 LULC映射。这项研究的结果为可持续土地管理提供了有用的见识。研究结果也证明了遥感算法的潜力，