Traditional Numerical Weather Prediction Systems focus on the modelling of the atmosphere, ocean and land processes, ignoring some more complex hydrological processes including the simulation of river discharge. In this talk, we describe how integrating simple one-way coupling of hydrological processes within a land surface and atmosphere modelling system can help understanding key biases and limitations in the simulations, using the ECMWF ECLand platform as an example.
Hydrological processes are generally defined within Earth System Modelling by vertical infiltration and evaporation fluxes without long-term land storage other that snowpack. In the real world, however, horizontal water movements and water land storage mean that river discharge measured in a catchment outlet is a natural integrator of water balance processes both in time and space. Over the last three years, ECMWF conducted multiple experiments to generate river discharge timeseries with different configurations of ECLand, the earth system modular platform of its Integrated Forecast System. The experiments used CaMAFlood, a global river hydrodynamic model, to transform daily runoff generated by HTESSEL, the Land Surface component of the ECLand, into river discharge multi-annual reanalysis time series. Simulations were verified against observations from the Global Runoff Data Centre worldwide for catchments selected for the limited impact of anthropogenic influence. Results showed the power of river discharge to amplify any discrepancy between modelled and observed timeseries and identified some weakness in the existing operational configuration of ECMWF’s IFS, including, for example, the impact of Data Assimilation on water balance closure and long-term trends in reanalysis datasets such as ERA5, or how multi-layer snow modelling could affect water processes in permafrost areas.
The work demonstrates the power of a holistic Earth System Modelling approach integrating river discharge as a diagnostic tool for land surface processes verification, paving the way for a future more complex coupling of atmosphere and land components.
How to cite: Prudhomme, C., Arduini, G., Balsamo, G., Brown, A., Boussetta, S., Cloke, H., de Rosnay, P., Harrigan, S., Mazzetti, C., Pappenberger, F., Sandu, I., and zsoter, E.: EC-Land Hydro: the benefits of integrating hydrology in earth system modelling, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-692, https://doi.org/10.5194/ems2022-692, 2022.
