Assessment of urban growth patterns on forest and water resources changes using remote sensing and machine learning techniques

ABSTRACT The landscape of Laos has changed significantly in terms of forest and water resources due to intensive human activities, including agricultural expansion and urbanization. In this study, remote sensing and machine learning models are used to analyze deforestation and surface water reduction to provide a detailed overview of environmental changes in Laos. The results show a complex development: the forest area increased from 90,506 km2 in 1990 to 103,671 km2 in 1995, which can be attributed to targeted reforestation measures. By 2020, however, the forest area had fallen to 76,811 km2, which represents an alarming trend of forest loss. An analysis of the period shows two distinct phases of forest change: rapid deforestation from 1990 to 2005 and a slower decline from 2005 to 2020, probably due to conservation measures. Surface water, which is crucial for the ecological balance, also increases slightly, from 843.22 km2 in 1990 to 852.94 km2 in 2020, indicating a slight increase over the last three decades. The study uses random forest and support vector regression to examine the link between forest degradation and river basin changes, highlighting the need for sustainable resource management in Laos to tackle environmental issues.