KARIMZADEH MOTLAGH Z., LOTFI A., POURMANAFI S., SAFIANIAN A., SAFIANIAN A.,  ENVIRONMENTAL MONITORING AND ASSESSMENT, Vol. 192, No. 695, PP. 1_19, 2020.

Abstract:

In the present paper, land use/land cover(LULC) change was predicted in the Greater Isfahanarea (GIA), central Iran. The GIA has been growingrapidly in recent years, and attempts to simulate itsspatial expansion would be essential to make appropri-ate decisions in LULC management plans and achievesustainable development. Several modeling tools wereemployed to outline sustainable scenarios for futuredynamics of LULCs in the region. Specifically, weexplored past LULC changes in the study area from1996 to 2018 and predicted its future changes for 2030and 2050. For this purpose, we performed object-oriented and decision tree techniques on Landsat andSentinel-2 satellite images. The CA-Markov hybridmodel was utilized to analyze past trends and predictfuture LULC changes. LULC changes were quantita-tively measured using landscape metrics. According tothe results, the majority of changes were related toincreasing residential areas and decreasing irrigated lands. The results indicated that residential lands wouldgrow from 27,886.87 ha to 67,093.62 ha over1996–2050 while irrigated lands decrease from 99,799.4 hato 50,082.16 ha during the same period of time. Theconfusion matrix of the 2018 LULC map was builtusing a total of 525 ground truth points and yielded aKappa coefficient and overall accuracy of 78% and82%, respectively. Moreover, the confusion matrix con-structed base on the Sentinel-2 map, as a reference, tojudge the predicted 2018 LULC map with a Kappacoefficient of 88%. The results of this study provideuseful insights for sustainable land management. Theresults of this research also proved the promising capa-bility of remote sensing algorithms, CA-Markov modeland landscape metrics future LULC planning in thestudy area.