Author(s): Hashim Ali Hasab

Abstract: This study investigates land use and land cover (LULC) changes in the western region of Iraq and neighboring countries within Sections 24 and 37, utilizing satellite remote sensing data and geographic information system (GIS) techniques. Landsat-9 imagery for the year 2025 was processed through geometric correction, image fusion, and digital classification methods to generate thematic maps of the study area. Two classification approaches were employed: Decision Tree (DT) and Support Vector Machine (SVM). Comparative evaluation of the two methods demonstrated that the SVM approach consistently produced results closer to the real field data than the DT method. Specifically, the integration of mathematical linear equations with GIS analysis showed that the SVM classification achieved a correlation coefficient of R² = 0.97, reflecting superior accuracy and lower standard error in estimating LULC elements. By contrast, the DT method recorded a slightly lower performance with R² = 0.96, highlighting its limitations in capturing spatial variability in arid and semi-arid environments. The findings recommend the adoption of SVM classification for environmental monitoring, as it offers higher reliability and precision.

DOI: 10.22271/27078302.2025.v6.i2b.73

Pages: 87-93 | Views: 118 | Downloads: 29

Download Full Article: Click Here