Coastline change and mangrove dynamics on Nias Island, North Sumatra, Indonesia, using Google Earth Engine 2002-2024

ELING TUHONO (Program of Natural Resource and Environmental Management, Graduate School, Universitas Sumatera Utara. Jl. Prof. T. Maas, Medan 20155, North Sumatra, Indonesia)
ZULKIFLI NASUTION (Program of Agrotechnology, Faculty of Agriculture, Universitas Sumatera Utara. Jl. Prof. A. Sofyan No. 3, Medan 20155, North Sumatra, Indonesia)
DELVIAN DELVIAN (Program of Forestry, Faculty of Forestry, Universitas Sumatera Utara. Jl. Lkr. USU, Deli Serdang 20353, North Sumatra, Indonesia)
NISFARIZA BINTI MOHD NOOR (Department of Geography, Faculty of Arts and Social Sciences, Universiti Malaya. Lingkungan Budi, 50603 Kuala Lumpur, Federal Territory of Kuala Lumpur, Malaysia)

Biodiversitas Journal of Biological Diversity
Vol. 27 No. 8 (2026)

Abstract. Tuhono E, Nasution Z, Delvian, Noor NBM. 2026. Coastline change and mangrove dynamics on Nias Island, North Sumatra, Indonesia, using google earth engine 2002-2024. Biodiversitas 27 (8): d270812. https://doi.org/10.13057/biodiv/d270812. Mangrove forests provide important ecological functions, support coastal community livelihoods and contribute to coastal protection. Nias Island is a tectonically active coastal environment where shoreline change may influence mangrove distribution. This study assessed shoreline configuration, abrasion, accretion, and mangrove-cover dynamics using Landsat 7 ETM+ imagery from 2002, 2004, and 2006 and Landsat 9 OLI-2 imagery from 2024, processed in Google Earth Engine (GEE). Shorelines were extracted using the Modified Normalized Difference Water Index, while mangrove cover was classified using a Random Forest model. Extracted coastline length increased from 1,416.7 km in 2002 to 1,506.9 km in 2006, then declined to 1,442.1 km in 2024. Direct comparison of the 2002 and 2024 shorelines identified 10,957.46 ha of accretion and 1,154.92 ha of abrasion, equivalent to annualized mapped-area rates of 498.07 and 52.50 ha/year, respectively. Mangrove cover decreased from 2,278.81 ha in 2002 to 1,736.91 ha in 2006 and then increased to 2,148.62 ha in 2024, producing a net decline of 130.19 ha (5.71%) over the full study period. Overall accuracy ranged from 89.28% to 94.23% for land-water classification and from 87.34% to 93.47% for mangrove mapping. Spatial overlays indicated that some accretion areas coincided with mangrove-cover gain, whereas some abrasion areas coincided with mangrove-cover loss; these patterns represent spatial associations rather than causal relationships. Interpretation is limited by Landsat’s 30 m resolution, uncorrected tidal variation, residual cloud and SLC-off effects, and incomplete historical field validation. The results provide a regional baseline for future tide-normalized, field-validated, and higher-resolution coastal monitoring, while site-specific management decisions require additional field, tidal, hydrodynamic, and sedimentary data.