From Natural Land to Built-Up Areas: Monitoring Residential Expansion Using Sentinel-2 and Support Vector Machine

Abstract. Uncontrolled urban growth, known as urban sprawl, poses a significant challenge to territorial sustainability by affecting ecosystems, infrastructure, and urban quality of life. In this context, the present study aims to analyze residential expansion in Guayaquil between 2016 and 2024, identifying changes in land cover using Sentinel-2 imagery and an SVM classifier, to provide inputs for sustainable urban planning and growth management in peripheral areas. The study area corresponds to Guayaquil, a coastal city in the Guayas province, characterized by flat terrain, estuaries, and high environmental vulnerability. Methodologically, Sentinel-2 satellite images, the Built-up Area Extraction Index (BAEI), and the Support Vector Machine (SVM) supervised classification algorithm were used, along with a multitemporal analysis for the periods 2016–2020 and 2020–2024. The results revealed a considerable increase in residential areas, especially in the city's northern and southern peripheral zones, indicating expansion into natural and high-risk areas. The combination of SVM and BAEI achieved an overall accuracy of over 85% and a Kappa index of 0.84, confirming the effectiveness of the methodological approach. It is concluded that remote sensing is a useful, accurate, and low-cost tool for monitoring urban growth and supporting sustainable territorial planning.