Dynamics and Determinants of Forest Changes Across Mainland Vietnam in the Recent Three Decades

Abstract

In the recent few decades, Vietnam has experienced a considerable change in land use/land cover (LULC), especially forest land. However, there is not a comprehensive analysis of the dynamics and drivers at the nationwide spatial scale for a long-term period. In this research, we estimate the socioeconomic and biophysical drivers of forest changes at the commune scale. Utilizing our results from the Vietnam-wide annual LULC database available in the Japan Aerospace Exploration Agency (JAXA), we first computed the dynamic changes in forest land from 1990 to 2020. To decide the major drivers of the changes, we conduct a synthesis of case studies working on the analysis of the forest changes in Vietnam at various spatial levels. Subsequently, a machine learning technique was adopted to measure the drivers of the forest changes. Our results indicate that although the forest area has increased from 2005 to 2010, it has undergone a decrease over the full study period. There is a dramatic conversion between forest and agricultural land, especially in the North-West and Central Highlands. This conversion is mainly driven by agricultural expansion/shifting, topographic position index, accessibility/infrastructure, population growth/migration, and distance to systems such as irrigation, drainage, and mining/industry. The identification of the drivers in this study is likely to help enhance the accuracy of the land use/land cover change prediction. These findings provide coherent evidence-based information about the dynamics and drivers of forest changes at the nationwide spatial and decadal temporal scales and thus can support informing land policies in Vietnam.