Combining mechanistic modelling and observations to characterize carbon and water fluxes in mainland Southeast Asia

Abstract

Tropical forests account for approximately one-fourth of the global terrestrial carbon sink, playing an important role in the Earth’s carbon cycle. Importantly, mainland Southeast Asia has the densest vegetation surface but its ecohydrology is historically understudied due to the paucity of field observations and modelling studies. Leveraging on existing flux tower data, remote sensing products, and the mechanistic ecohydrological model T&C, we provide an enhanced understanding of carbon and water exchanges in mainland Southeast Asia. The T&C model is tested to reproduce various ecosystem types of Southeast Asia, including tropical evergreen forests, subtropical deciduous forests, savannas, rubber plantations, and rice fields. The flux tower data including gross primary productivity (GPP) and evapotranspiration (ET) along with remote sensing data of leaf area index and other vegetation indexes, allow us to better refine and constrain model simulations.  With the integration of data and model, we provide a comprehensive picture of spatiotemporal patterns and key drivers of carbon and water fluxes in mainland Southeast Asia. Our findings highlight a strong latitudinal gradient in carbon fluxes and ET associated with seasonality of rainfall as well as an important role of vapour pressure deficit (VPD) and soil moisture content with different responses in wet and dry years. Direct effects of temperature and precipitation are relatively smaller when compared to VPD and soil moisture in driving changes of carbon and water fluxes. These findings, combined with our model framework, pave the road to more accurate predictions of ecohydrological variables in the relatively understudied region of mainland Southeast Asia.