Geospatial Assessment of Urban Expansion and Green Area Decline in Karbala Using Machine Learning: Supporting the Achievement of SDGs

Doaa Taha Yaseen, Department of Civil Engineering, College of Engineering, Wasit University, Wasit, Iraq AND Department of Geomatics Engineering, College of Civil Engineering, University of Technology - Iraq, Baghdad, Iraq Follow
Oday Zakariya Jasim, Department of Geomatics Engineering, College of Civil Engineering, University of Technology - Iraq, Baghdad, Iraq

Keywords

Land cover classification, Machine learning, Sentinel-2, SDGs11.3.1., Urban Sprawl

Document Type

Article

Abstract

Urban sprawl is a major challenge to sustainable development. This research examines geospatial patterns of urban growth and vegetation cover reduction in the Karbala District between 2017 and 2024. Sentinel-2 satellite imagery classification using a Support Vector Machine (SVM) model was evaluated, and accuracy was assessed using confusion matrices and a comparative evaluation against the Esri Sentinel Land Cover dataset. The SVM approach provided better classification precision. The change-detection analysis shows that the conversion of green and cultivated lands into built-up areas is significant. Although there has been an expansion of pivot-irrigated areas in other districts of the Karbala Governorate, the central district has shown little agricultural development. The SDG indicator 11.3.1 was estimated at 4.4971, indicating that urban development has been much faster than population growth. These results emphasise the importance of machine learning and Earth observation for tracking land dynamics.

Recommended Citation

Yaseen, Doaa Taha and Jasim, Oday Zakariya (2026) "Geospatial Assessment of Urban Expansion and Green Area Decline in Karbala Using Machine Learning: Supporting the Achievement of SDGs," Iraqi Journal of Architecture and Planning: Vol. 25: Iss. 2, Article 15.