Modelling hydrological and dissolved carbon fluxes from forests under current and future rainfall regimes

The project will consist in modelling future alterations in the forest carbon balance, by implementing the results in coupled climate, hydrological and forest ecosystem biogeochemical models. More specifically, it will assess the dissolved organic carbon (DOC) export from deciduous and coniferous forest, and hereby close ecosystem carbon balances in two Belgian ICOS-sites. Future climate conditions involve increased temporal variability, with both increases in precipitation intensity and changes in the number of rain storms (more rainfall events in the wet winter periods and less rain in the dry summer periods). These changes lead to an intensification of the hydrological cycle, with higher runoff flows in the wet periods and stronger dry-out of the soils during the longer drought periods. Such alterations in precipitation patterns will cause large changes in the forest carbon balance. Hydrological conditions are the main driver of DOC leaching, both on intra-annual and inter-annual timescale. Longer periods of drought, and occasional strong precipitation events, could cause a more event-driven carbon export from soils through dissolved losses. It also will be assessed whether a shift will occur in gaseous vs. dissolved carbon with changing rain regimes. Site-specific empirical findings, provided by other researchers, will be extrapolated to the larger spatial scales by means of hydrological modelling. A novel hydrological modelling approach will be implemented and applied for that purpose. Particular focus will be given to the spatial resolution of the model. The approach will allow a direct coupling to climate change scenario analysis, hydrological and phenological characterization and the biogeochemical assessment of dissolved carbon fluxes. On the one hand, there is a high spatial resolution required for studying the surface hydrological processes of the local study plots around the ICOS sites. On the other hand, to model the role of the groundwater a larger scale model covering the entire watershed is required. The fine resolution model at the local scale will be nested in the coarse resolution model at the larger scale. In addition, depending on the results of the empirical observations, conceptual process implementations will be tested and evaluated. This project thus applies an innovative ‘‘method of multiple working hypotheses’’ where the model structure is adjusted or inferred from data and field evidence. The coupled hydrological/biogeochemical/phenological assessment will finally be implemented in detailed climate scenarios for Belgium. Future climate conditions will be assessed through downscaling of climate model projections. The downscaling will be a dynamic-statistical one. The uncertainty in the climate change projections for Belgium will be quantified by an ensemble modelling approach, considering the full set of available climate model runs in the databases of CMIP5 and EURO-CORDEX, complemented by high-resolution climate models. The latter models allow to dynamically downscale to large-scale changes in atmospheric conditions to local changes. They also allow the link observed fluxes to local climate patterns. By means of statistical downscaling, the uncertainties in the large-scale (CMIP5 and EURO-CORDEX based) future projections will be superimposed on the local climate model impact results.