Comparing the impact for hydrology of the new ERA5 reanalyses dataset over ERA-Interim for 8 hydrological models in 6 catchments using the eWaterCycle community modelling environment.
- 1Delft University of Technology, The Netherlands (r.w.hut@tudelft.nl)
- 2Netherlands eScience Center, The Netherlands
- 3Deltares, The Netherlands
- 4Utrecht University, Utrecht, The Netherlands
- 5Wageningen University, The Netherlands
- 6University of Washington, USA
- 7European Center for Medium range Weather Forecast (ECMWF), UK
- 8CRP Gabriel Lippmann
- 9Swedish Meteorological and Hydrological Institute (SMHI), Sweden
- 10Institut Català de Ciències del Clima, Spain
- 11University of Bristol, UK
- 12University of Saskatchewan, Canada
- 13Arizona State University, USA
- 14Community Surface Dynamics Modeling System (CSDMS), USA
- *A full list of authors appears at the end of the abstract
The release of the European Centre for Medium-Range Weather Forecasts (ECMWF)’s Re-Analysis 5 (ERA-5) global climate forcing dataset is expected to greatly improve the quality of hydrological modeling. Following this release there is great interest in assessing the improvements of ERA-5 relative to its predecessor ERA-Interim for hydrological modeling and predictions.
In this study we compare streamflow predictions when using ERA-interim vs ERA-5 as forcing data for a suite of hydrological models from different research groups that capture the variation in modelling strategies within the hydrological modelling community. We check whether physically based models, defined as those that do not require additional parameter calibration, would lead to different conclusions in comparison to conceptual models, defined as those that require calibration. Based on the hydrological model structure we expect that conceptual models that need calibration show less difference in predicting discharge (skill) between ERA-5 and ERA-Interim, where-as the physical based (non-calibrated) models most likely will benefit from the improved accuracy of the ERA-5 input. This assessment will provide the HEPEX community with answers on how the ERA-5 dataset will improve hydrological predictions based on different hydrological modelling concepts.
An additional key objective while conducting this study is compliance to the FAIR principles of data science. To achieve this we held a workshop in Leiden, the Netherlands, where multiple hydrological models were integrated into the eWatercycle II system. eWatercycle II is a hydrological model platform containing a growing number of hydrological models. The platform facilitates research and cohesivity within the hydrological community by providing an Open-Source platform built specifically to advance the state of FAIR and Open Science in Hydrological Modeling. We also use this study to demonstrate the feasibility of eWatercycle II as a platform for FAIR hydrological models.
Preliminairy results from this comparison study were presented at the AGU Fall Meeting 2019. Here we will present the full results of the comparison study.
Ben van Werkhoven Ronald van Haren Yifat Dzigan Jaro Camphuijsen Fakhereh Alidoost Inti Pelupessy Berend Weel Gijs van den Oord Stefan Verhoeven Bouwe Andela Peter Kalverla
How to cite: Hut, R., Drost, N., Aerts, J., Bouaziz, L., van Verseveld, W., Jagers, B., Baart, F., Sutanudjaja, E., Melsen, L., Bennett, A., Arnal, L., Fenicia, F., Santos, L., Gelati, E., Dal Molin, M., Knoben, W., Gharari, S., Hall, C., Hutton, E., and van de Giesen, N. and the the eWaterCycle comparison team from the Netherlands eSciencecenter.: Comparing the impact for hydrology of the new ERA5 reanalyses dataset over ERA-Interim for 8 hydrological models in 6 catchments using the eWaterCycle community modelling environment., EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-10219, https://doi.org/10.5194/egusphere-egu2020-10219, 2020.
