Hendry Frananda, Paus Iskarni, Fitri M. Sari, Muhammad Ismail, Eva Purnamasari, Asra I. Khairi
This study assesses multi-temporal changes in mangrove cover along the coast of Padang Pariaman by integrating Sentinel-2A imagery from 2016 and 2025 with a Random Forest classification framework on Google Earth engine. Pre-processed composites and derived spectral features were used to generate land cover maps, yielding a high classification performance with an overall accuracy of 89.2% and a Kappa coefficient of 0.89. The results reveal a marked decline in mangrove extent, decreasing from 746 hectares in 2016 to 679 hectares in 2025, indicating a net loss of 67 hectares over nine years. Spatial analysis demonstrates that this degradation is predominantly associated with the expansion of shrimp ponds, which intensified after 2017 and led to progressive fragmentation of mangrove stands. These findings highlight the ecological vulnerability of the region and underscore the importance of systematic remote sensing–based monitoring for informing coastal management, regulating aquaculture development, and prioritizing mangrove conservation and restoration efforts. © 2026, BIOFLUX SRL. All rights reserved.
Department of Geography, Universitas Negeri Padang, Padang, 25171, Indonesia; Department of Statistics, Universitas Negeri Padang, Padang, 25171, Indonesia; Remote Sensing and Geographic Information System, Vocational School, Universitas Negeri Padang, Padang, 25171, Indonesia; Department of Visual Art, Universitas Negeri Padang, Padang, 25171, Indonesia
