Biogeochemistry and Earth System Modelling

River Carbon: Modeling inland water greenhouse gas fluxes - ORCHI-LEAK


I. Context: The conventional wisdom is that inland waters simply transport terrigenous organic carbon to the oceans. This view is perpetuated by current assessments of the global carbon cycle that largely ignore inland waters as represented in, e.g., the Intergovernmental Panel for Climate Change (IPCC). In the six years since the publication of the last IPCC report, it has become apparent than the global flux of greenhouse gases (GHG) from inland aquatic sources to the atmosphere is much larger than previously suspected (Regnier et al., Nature Geo., 2013; Raymond et al., Nature, 2013). Despite the potential importance of these GHG emissions, their inclusion in current Earth System Models is still missing.


II. Description of work: The proposed research is a collaboration between ULB-BGEOSYS and the Institut Pierre Simon Laplace (IPSL-LSCE). Its aim is to to incorporate the aquatic C cycle (land to ocean) into the land surface scheme ORCHIDEE of the IPSL Earth System model, to validate it against recently established global maps of river pCO2 and fluvial C exports and to quantify the presently unknown atmospheric feedbacks between inland aquatic carbon evasion fluxes and drivers such as climate change and anthropogenic eco-hydrological disturbances. This project naturally follows from a series of high-profile papers (Regnier et al., Nature Geo. 2013; Raymond et al., Nature 2013; Bauer et al., Nature, 2013), which highlight the significance of the inland water flux of greenhouse gases (GHG) to the atmosphere for the global carbon carbon budget.


The project will require the technical implementation and calibration of fluvial transport, trapping and transformations of particulate organic carbon (PC) and dissolved organic carbon and CO2 (DC) into ORCHIDEE (ORCHI-LEAK, Task 1), and its validation against observed data (Task 2)


Task 1 Technical implementation: The existing surface-runoff routing scheme of ORCHIDEE will be enabled for the lateral transport of DC and PC with the water flow through rivers, over floodplains, and through lakes and reservoirs. This will allow simulating the whole lateral translocation of DC + PC from the upland soil down to the coasts, including all sinks from the colluvia at the foot slope, small upland impoundments, downstream lakes and reservoirs and in floodplains. The proposed plan will require integration of fluvial fluxes of dissolved organic C and CO2 (DC) into ORCHIDEE and allow for modelling of the lateral export of all soil derived C (TC = PC+DC) and its transformation between particulate, dissolved and gaseous phases. Processes will comprise adsorption and flocculation (DC->PC), production of DC from the decomposition of PC, and oxidation of organic C to CO2 and the exchange of CO2 through the air-water interface. While periodical flooding of floodplains is already simulated in ORCHIDEE, lake (GLWD) and reservoir databases (GRanD) will be integrated into the river routing scheme of ORCHIDEE to simulate Pc burial. The damming effect on C cycling will be simulated in the context of the EU project C-CASCADES (www.c-cascades.ulb.be).


Task 2 Validation of fluvial C transfers and transformation for present day conditions : The model will be validated for present day conditions against observed fluvial DC + PC flux data and reported Pc burial rates as well as observed Dc and Pc concentrations in lakes and reservoirs. For this purpose, the model will be run offline at variable resolutions (0.5° and 1°) using forcing files for climate (CRU-NCEP) and land cover. A sensitivity analysis will be performed and uncertainty of model predictions will be quantified. The regional focus will be on the two contrasted watersheds : the Amazon and Rhine catchments.


III. Results:

This figure shows the CO2 evasion from the rivers and floodplains of the Amazon basin simulated by ORCHI-LEAK.



For more details on the ORCHI-LEAK model configuration and results, click here.