Accuracy Evaluation of Land Cover Classification Maps Using Remote Sensing and GIS. The Sea of Najaf-Iraq as a Case Study

Abstract

Water scarcity is a growing concern, particularly in arid and semi-arid regions. Sea of Najaf in Iraq is one such region facing water scarcity, with the lake serving as a crucial water source for the local population. Remote sensing systems "RS" and geographic information systems "GIS" are important techniques for monitoring environmental and land cover changes. This paper focuses on Sentinel-2B imagery for surface water mapping in the Sea of Najaf and evaluating the performance of maximum likelihood (MLC) classification method. The study area encompasses Sea of Najaf and its immediate surroundings. Two images for Sentinel-2B one image on December,29, 2015 and December,29,2022, were used for training and validating the classification models. The MLC method was evaluated for surface water classification, with accuracy assessment results presented. The MLC method showed high accuracy in terms of overall accuracy and agreement with reference data over multiple years. Additionally, fluctuations in the surface area of Sea of Najaf, as observed from the classification maps, were analyzed. In 2015, the surface area was 74.36 square kilometers, which increased to 120.30 square kilometers in 2022. The fin, dings highlight the efficacy of the MLC method for surface water classification, indicating its superiority in accurately mapping and monitoring surface water in the Sea of Najaf.