Monitoring Spatio-Temporal of Vegetation Indices Using NDVI and EVI Algorithms on Google Earth Engine (GEE) in Tuntang Watershed Area, Indonesia

Abstract Climate change affects increase rate in flood fluctuations, drought, and deforestation in watershed areas. The condition of Tuntang watershed has decreased in carrying capacity due to several factors. Google Earth Engine (GEE) is one of the significant technological breakthroughs in remote sensing and GIS. The aims to analyze the Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) in the Tuntang watershed using GEE and analyze spatio-temporal variations in vegetation density conditions. The method is processing NDVI and EVI algorithms for five years (2018-2022). This research to the full set of images acquired with the Sentinel-2A sensor between August 2018 and September 2022. The findings indicated that the highest forest area for five years with the NDVI method was in 2018, namely dense to very dense density of 716.92 Ha (54.78 %), while in the EVI method, the highest in 2020 was moderate to dense density of 504.58 Ha (38.55%). Based on the accuracy test, NDVI and EVI have the same accuracy overall of 83.33%, with a kappa index of 0.78 NDVI and 0.79 for EVI. Based on the correlation test, both methods have a correlation value of 0.799, indicating a strong and unidirectional relationship.