USING ATMOSPHERICALLY RESISTANT VEGETATION INDEX TO DETECT FOREST COVER CHANGE IN LAC DUONG DISTRICT,  LAM DONG PROVINCE

Dinh Cao Phi; Nguyen Hai Hoa

Keywords:

ARVI, Lac Duong, land cover, land use, NDVI

Abstract

This study tested the potential of using Sentinel-2-derived ARVI for land covers mapping compared to NDVI in Lac Duong district, Lam Dong province. Overall accuracy assessments of ARVI-based land covers mapping are calculated from 88.9% (in 2016) to 92.0% (in 2022) with Kappa coefficients of 0.79 and 0.86, respectively. The comparison between Sentinel-2-derived NDVI and ARVI for land covers mapping indicated that ARVI has performed well in land covers mapping in Lac Duong district. The Sentinel-2 time series were therefore used to generate spectral-temporal changes in land covers based on ARVI thresholds. Our estimation of land covers and forest cover change shows that the forest cover in Lac Duong increased by 1718.4 ha from 2016 to 2022, while non-forest covers reduced by 1662.1 ha at the same period. Conversely, during the segmented period of 2018-2020, the extent of forest cover reduced by 1156.1 ha, whereas non-forest covers increased by 1032.8 ha. The study also identified the main drivers of LULC change, including forest cover change, these changes are due to the expansion of coffee production, effectiveness of PES policy implementation, and a lack of agricultural production and sustained likelihoods. Our study suggests that ARVI can be used to monitor and detect changes in land covers, especially in deforestation and forest degradation detection in the tropical and mountainous regions of Vietnam. More focus of comprehensive understanding of LULC change and its management should be expanded beyond the poverty-environmental relationships in Lac Duong district.