An Accuracy Analysis Comparison of Supervised Classification Methods for Mapping Land Cover Using Sentinel 2 Images in the Al‑Hawizeh Marsh Area, Southern Iraq

A comparative analysis of supervised classification methods for land cover mapping in the Al-Hawizeh Marsh, part of Iraq's UNESCO-listed Ahwar wetlands, reveals that Maximum Likelihood Classification (MLC) using Sentinel 2B satellite imagery achieves the highest accuracy. With an overall accuracy of 85.32% and a superior kappa coefficient compared to Artificial Neural Networks (ANN) and Support Vector Machine (SVM), MLC demonstrated its effectiveness in distinguishing six critical land cover classes: deep water marsh, shallow water marsh, aquatic vegetation, urban areas, agriculture, and barren soil. This finding is significant for monitoring biodiversity hotspots like the Ahwar, where spectral similarities pose challenges for accurate classification. The study utilized ENVI 5.1 software to analyze multispectral images with a spatial resolution of 10 meters, providing valuable insights for conservation efforts in these ecologically vital regions.

Aziz, N. A., & Alwan, I. A. (2021). An Accuracy Analysis Comparison of Supervised Classification Methods for Mapping Land Cover Using Sentinel 2 Images in the Al‑Hawizeh Marsh Area, Southern Iraq. Geomatics and Environmental Engineering, 15(1), 5–21. https://doi.org/10.7494/geom.2021.15.1.5