Hamrin Lake monitoring and sustainability in Iraq using GIS-based machine learning and remote sensing

Abstract

The present research is focused on the study of the changes that have occurred in Hamrin Lake and to test the ability of Geographic Information System (GIS)- based Support Vector Machine (SVM) to provide reliable water monitoring. The main objective of this research is to explore the evolution of the Lake by using GIS, Sentinel-2 satellite images, and a Machine Learning ML approach, SVM classification. Based on three multi-temporal images of the years 2017, 2022, and 2026, we assessed the changes in lake extent and its surroundings. The classification results indicated that in October 2017, the Hamrin Lake area was 169.09 km², but showed a significant drop in April 2022 to 30.08 km², and a rise to almost 185.77 km² in January 2026. Such variations are driven by the drying conditions experienced within the catchment, hydrological variation, evaporation, and water management within the study area. The interpretation and consistency of the classified water boundaries with the classification in the generated maps showed that the SVM classification had an estimated overall accuracy of about 94%. The observed fluctuations highlight the need for monitoring support for sustainable water resource management and putting in place effective water allocation and conservation strategies in Iraq.