Innovative approach to prognostic plant growth modeling in SWAT+ for forest and perennial vegetation in tropical and Sub-Tropical climates

Abstract

The growth of vegetation in ecosystems is influenced by hydro-climatic factors and biogeochemical cycles. Accurately modeling annual vegetation growth dynamics is essential for eco-hydrological modeling to estimate watershed hydrologic balance and nutrient cycling under changing environmental conditions. The Soil and Water Assessment Tool (SWAT) and its upgraded version SWAT+ are process-oriented river basin models widely used. However, the temperature-based approach to plant growth simulation in tropical regions has limitations due to the importance of soil moisture availability as a key driver of plant growth. This study proposes an innovative approach that incorporates a proxy soil moisture availability index based on monthly rainfall and potential evapotranspiration ratio. This approach identifies the start of the growing season within prescribed transition months and controls leaf drop rate throughout the year, a crucial process during leaf senescence. We evaluated the reliability of this approach by comparing SWAT+ simulated Leaf Area Index (LAI), evapotranspiration (ET), and net primary productivity (NPP) with benchmark remote sensing-based datasets for three landcover classes in the Mara River Basin (Kenya/Tanzania). Our results demonstrate that the improved plant growth module in SWAT+ developed in this study can simulate temporal vegetation growth dynamics of evergreen forest, savanna grassland, and shrubland land cover types consistently with good correlations (r > 0.5) and low average bias (<10%). Thus, the SWAT+ model with the enhanced plant growth module can be a robust tool for investigating the coupled carbon, nutrient, and water cycling in tropical and sub-tropical climates.