Temporal and spatial relationships between hydrologic and carbon budgets in an Amazonian watershed: Application of a coupled Subsurface - Land Surface Process Model

Thursday, December 17, 2015 - 08:00
Add to Calendar

Amazonian tropical ecosystems cycle large amounts of carbon and water. The interactions between Amazonian hydrologic and carbon cycles are uncertain and therefore an important research topic. We analyzed the water and carbon budgets in a ~9,000 km2 central Amazon basin near Manaus, Brazil, using an observationally constrained (streamflow) and tested process-based adaptive watershed simulator (PAWS) coupled with the community land model (CLM). The coupled model (PAWS+CLM) includes detailed representations of subsurface and land surface hydrologic and carbon cycle processes. We compared model predictions with site observed carbon states and fluxes (LAI, NPP, and GPP), satellite-based estimates of evapotranspiration and Leaf Area Index (LAI), and water storage anomalies from the Gravity Recovery and Climate Experiment (GRACE) satellite. The simulated results and correlation analysis support the following observations. First, the high-resolution, distributed PAWS+CLM model better matches available hydrologic and carbon cycle observations than the one-dimensional CLM. Second, in this tropical forest headwater basin, most hydrologic components are very sensitive to precipitation forcing, while ET is not. Third, groundwater is the dominant contributor to streamflow, although the streamflow peaks are highly correlated with intense precipitation input. Fourth, carbon fluxes, represented by LAI, NPP and GPP are correlated with hydrologic processes both temporally and spatially, and in particular to ET and groundwater flow. Our simulations provide new estimates of water budgets and storage changes in an Amazonian headwater catchment. The results suggest that prognostic groundwater and lateral flows are important for accurate representation of hydrologic and carbon fluxes at the watershed scale and that accurate prediction of these processes requires accurate climate forcing data.