Geographia Technica, Vol 22, Issue 1, 2027, pp. 64-77
HISTORICAL DETECTION OF VEGETATION COVER CHANGES IN THE COASTAL AREA OF BATANG REGENCY (2010 - 2024) USING THE LANDTRENDR ALGORITHM
Muhammad Fauzan RAMADHAN 
, Moh. Rayya Ilham REHARDIYAN , Tjaturahono Budi SANJOTO , Ayu Wulansari PRAMITA
ABSTRACT: Urbanization and large-scale infrastructure development are major drivers of vegetation cover changes in coastal areas, with significant implications for land-use planning and vegetation management. This study aims to detect and analyze historical vegetation cover changes in the coastal area of Batang Regency, Central Java, Indonesia, over the period 2010–2024 using the LandTrendr algorithm. The study area covers approximately 110 km² and is characterized by intensive development pressure from two major national strategic projects: the Batang Coal-Fired Power Plant (PLTU) and the Batang Integrated Industrial Zone (KITB). Vegetation changes were assessed using the Normalized Burn Ratio (NBR) as the spectral index, with the Greatest Change scenario applied to identify the segment of largest magnitude per pixel. Accuracy assessment was conducted using a 3×3 confusion matrix with 225 validation points through stratified random sampling. The results indicate an overall accuracy of 71.56% with a Kappa coefficient of 0.57. Over the study period, total vegetation gain reached 555.57 ha while total loss amounted to 338.13 ha, yielding a net positive change of +217.44 ha. Temporally, gain was dominant in the early observation period, peaking in 2010, while loss increased significantly after 2020 coinciding with active construction phases of PLTU and KITB. Spatially, 80.2% of total loss was concentrated in the PLTU and KITB zone, whereas gain was more dispersed across urban and riparian areas. These findings provide a spatial-temporal baseline for vegetation management and land-use planning in coastal areas facing intensive infrastructure development pressure.
Keywords: Normalized Burn Rasio; Remote Sensing; Time-series analysis; LandTrendr; Land Cover Dynamics.
Full article here