2. Research and development of a Web application based on Google Earth Engine cloud computing platform to support forest resource management in Ha Tinh province This paper presents the research results on developing a web application based

Abstract

This paper presents the research results on developing a web application based on the Google Earth Engine (GEE) platform to support forest resource management in Ha Tinh province. The study utilized Landsat (5, 7, 8, 9) and Sentinel-2A satellite imagery along with remote sensing indices including NDVI, EVI, and NBR to assess forest area changes across the province during the period 2009 - 2023, and to evaluate damage caused by a forest fire in Nghi Xuan district in June 2019. Key methods applied in the study include remote sensing image processing techniques and comparative analysis for validating the results against ground-truth data. Optimal threshold values for NDVI, EVI, and NBR were determined to be 0.750, 0.457, and 0.598, respectively. Validation results indicated that NDVI was the most stable index, yielding the lowest error (PBIAS -0.565 %). Additionally, the test implementation of the web application on the GEE platform achieved an average error of approximately 6.26 % when estimating forest area, compared to official statistics. For the 2019 forest fire in Nghi Xuan, the application estimated a burned area of 125.55 hectares - around 30 % higher than the reported figure. Despite this discrepancy, the results showed a strong alignment with the report from the Provincial Forest Protection Department. This highlights the great potential of the application as a rapid assessment tool for forest managers in Ha Tinh province.