TY - JOUR
T1 - Assessing the performance and robustness of two conceptual rainfall-runoff models on a worldwide sample of watersheds
AU - Mathevet, Thibault
AU - Gupta, Hoshin
AU - Perrin, Charles
AU - Andréassian, Vazken
AU - Le Moine, Nicolas
N1 - Funding Information:
Authors would like to acknowledge: M?t?o France, SCHAPI-Banque Hydro, EDF, Laurent Coron, Nicolas Le Moine and Audrey Valery for the French data sets, Jai Vaze and Francis Chiew for the Australian data sets (CSIRO), John Schaake and Qingyun Duan for the American (MOPEX) data set, Audrey Valery for the Swiss (M?t?o Suisse and OFEV) and Swedish (SMHI) data sets, Berit Arheimer for the Swedish (SMHI) data sets, and Barry Croke and Ian Littlewood for the English data sets (TDMWG). We thank the two anonymous reviewers and the Associate Editor Roger Moussa for their constructive feedbacks on this work, which helped improving the quality of the article. Hoshin Gupta received partial support from the Australian Research Council through the Centre of Excellence for Climate System Science (grant number CE110001028), and from the EU-funded project ?Sustainable Water Action (SWAN): Building Research Links Between EU and US? (INCO-20011-7.6 grant number 294947). Thibault Mathevet received appreciated support from Annick Gingras-Genois and Federico Garavaglia from EDF-DTG to visit the Department of Hydrology and Water Resources of the University of Arizona. Shervan Gharari, Mari A. Sans Fuentes, Rositsa Yaneva, Natalia Limones, and Exo Roast Coffee baristas are warmly acknowledged for helping to make the time in Tucson both fascinating and amazing.
Funding Information:
Hoshin Gupta received partial support from the Australian Research Council through the Centre of Excellence for Climate System Science (grant number CE110001028), and from the EU-funded project ‘Sustainable Water Action (SWAN): Building Research Links Between EU and US’ (INCO-20011-7.6 grant number 294947).
Funding Information:
Authors would like to acknowledge: Météo France, SCHAPI-Banque Hydro, EDF, Laurent Coron, Nicolas Le Moine and Audrey Valery for the French data sets, Jai Vaze and Francis Chiew for the Australian data sets (CSIRO), John Schaake and Qingyun Duan for the American (MOPEX) data set, Audrey Valery for the Swiss (Météo Suisse and OFEV) and Swedish (SMHI) data sets, Berit Arheimer for the Swedish (SMHI) data sets, and Barry Croke and Ian Littlewood for the English data sets (TDMWG). We thank the two anonymous reviewers and the Associate Editor Roger Moussa for their constructive feedbacks on this work, which helped improving the quality of the article.
PY - 2020/6
Y1 - 2020/6
N2 - To assess the predictive performance, robustness and generality of watershed-scale hydrological models, we conducted a detailed multi-objective evaluation of two conceptual rainfall-runoff models (the GRX model, based on the GR4J model, and the MRX model, based on the MORDOR model), of differing complexity (with respectively, 5 and 11 free parameters in the rainfall-runoff module, and 4 and 11 free parameters in the snow module). These models were compared on a large sample of 2050 watersheds worldwide. Our results, based on the three components of the Kling-Gupta Efficiency metric (KGE), indicate that both models provide (on average) similar levels of performance in evaluation when calibrated with KGE, for water balance (mean bias lower than 2%), time-series variability (mean variability bias lower than 2%) and temporal correlation (mean correlation around 0.83). Further, both models clearly suffer from lack of robustness when simulating water balance, with a significant increase of the proportion of biased simulations over the evaluation periods (absolute bias lower than 2% in calibration and lower than 20% in evaluation for 80% of the watersheds). Simulation performance depend more on the hydro-meteorological conditions of a given period than on the complexity of the model structure. We also show that long-term aggregate statistics (computed on the overall period) can fail to reveal considerable sub-period variability in model performance, thereby providing inaccurate diagnostic assessment of the predictive model performance. Typically the median absolute bias is lower than 8% in evaluation, but the median maximum bias can be as high as 50% within a subperiod, for both models, when calibrated with KGE.
AB - To assess the predictive performance, robustness and generality of watershed-scale hydrological models, we conducted a detailed multi-objective evaluation of two conceptual rainfall-runoff models (the GRX model, based on the GR4J model, and the MRX model, based on the MORDOR model), of differing complexity (with respectively, 5 and 11 free parameters in the rainfall-runoff module, and 4 and 11 free parameters in the snow module). These models were compared on a large sample of 2050 watersheds worldwide. Our results, based on the three components of the Kling-Gupta Efficiency metric (KGE), indicate that both models provide (on average) similar levels of performance in evaluation when calibrated with KGE, for water balance (mean bias lower than 2%), time-series variability (mean variability bias lower than 2%) and temporal correlation (mean correlation around 0.83). Further, both models clearly suffer from lack of robustness when simulating water balance, with a significant increase of the proportion of biased simulations over the evaluation periods (absolute bias lower than 2% in calibration and lower than 20% in evaluation for 80% of the watersheds). Simulation performance depend more on the hydro-meteorological conditions of a given period than on the complexity of the model structure. We also show that long-term aggregate statistics (computed on the overall period) can fail to reveal considerable sub-period variability in model performance, thereby providing inaccurate diagnostic assessment of the predictive model performance. Typically the median absolute bias is lower than 8% in evaluation, but the median maximum bias can be as high as 50% within a subperiod, for both models, when calibrated with KGE.
KW - Calibration
KW - Diagnostics
KW - Evaluation
KW - Hydrological modeling
KW - Kling-Gupta efficiency
KW - Large-sample hydrology
UR - http://www.scopus.com/inward/record.url?scp=85082723950&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85082723950&partnerID=8YFLogxK
U2 - 10.1016/j.jhydrol.2020.124698
DO - 10.1016/j.jhydrol.2020.124698
M3 - Article
AN - SCOPUS:85082723950
VL - 585
JO - Journal of Hydrology
JF - Journal of Hydrology
SN - 0022-1694
M1 - 124698
ER -