ABSTRACT

Introduction: Presently, unplanned changes of land use have become a major problem. Most land use changes occur without a clear and logical planning with little attention to their environmental impact. Since those changes in land use take place in large and extensive areas, so, remote sensing technology is a necessary and valuable tool for land use change detection.

Objectives: The purpose of this research is to investigate land use changes in the past and evaluate the efficiency of SVM model in modeling land use changes.

Methodology: In this study, images of TM (1988), ETM+ (2001) and OLI (2023) were processed using Support Vector Machine (SVM) in Cham Gardalan watershed with 476/751 square Kilometer and occurred changes were investigated in two time periods, from 1988 to 2001 and 2001 to 2007. In order to accuracy assessment of this method, after collecting ground truth data which were carried out through field visiting, Google earth images and aerial photographs, overall accuracy and Kappa Coefficient was used.

Geographical Context: The geographical area of this research is the Cham Gardlan watershed of Illam province.

Results: According to obtained results, overall accuracy of resulted maps through classification using Support vector machine method for TM, ETM+ and OLI images was 93%, 95% and 94%, respectively, that states high accuracy of this algorithm in classification of satellites images.

Conclusion: Based on results, during 1988 to 2001, lake was created due to dam Construction and decreasing rangeland and in second period from 2001 to 2007, change in cultivation method from irrigated cultivation to dry farming and increasing rangeland and forest land was the most obvious change which occurred in this watershed.

Conclusion: Based on results, during 1988 to 2001, lake was created due to dam Construction and decreasing rangeland and in second period from 2001 to 2007, change in cultivation method from irrigated cultivation to dry farming and increasing rangeland and forest land was the most obvious change which occurred in this watershed.