Quantitative assessment of past and future tropical forest transition and its dynamic to streamflow of the catchment, Malaysia

Abstract

The consequences of human interventions on land use have been empirically demonstrated to affect substantially hydrological processes and ecosystem services within watershed environments. Since 1970, Malaysia has faced deforestation, driven mainly by logging and agricultural expansion, aligning with its developmental goals by 2020. From 1970 to 2000, deforestation led to a 25.5% decline in forested land, causing a significant 10.2% rise in excess runoff. Moreover from 2001 to 2021, the nation lost 17% of its total land to deforestation. These trends emphasize the need for a thorough investigation of sustainable conservation efforts in Malaysia. This study focuses on the Langat basin in Malaysia, evaluating past and future land use changes and their effects on the basin’s hydrological response. The study employed key informant reports, population growth data, observed land use change, field survey and agricultural land availability considered for developing change scenarios. We emphasized the significance of integrating diverse modeling methods to analyze LULC changes effectively. The use of a semi-distributed hydrological model, SWAT, in combination with Markov chain and Multi-Layer Perceptron (MLP) model and Geographic Information Systems (GIS) techniques proved to be an integrated and suitable tool for comprehensive change analysis and modeling of land use. Markov chain modeling is valuable for predicting land use changes over time, providing input scenarios for SWAT simulations. MLP is a powerful algorithm to capture non-linear relationships and complex patterns in the data, enhancing the modeling accuracy. The simulation results, based on historic land use data (1984-2006-2010) and projected future land use maps (2030-2050-2080), revealed a consistent pattern of urban expansion and deforestation leading to increased streamflow. Projections indicated a substantial rise in streamflow by 20%, 61%, and 71% for the 2030s, 2050s, and 2080s, respectively. To mitigate potential flood and sediment loss, it is crucial to involve local stakeholders such as local communities, government bodies, environmental organizations, and businesses. Such analysis facilitates understanding their perspectives and concerns regarding afforestation and urban expansion control, informing future development programs and land use planning effectively.

Keywords

• Land Use Conversion
• Hydrological Processes
• Land Use Scenarios
• Markov Chain
• SWAT