Published: Dec. 1, 2023
BENSO, M. R
Language: Английский
Computers & Chemical Engineering, Journal Year: 2024, Volume and Issue: 182, P. 108559 - 108559
Published: Jan. 3, 2024
Language: Английский
Citations
7
Hydrology and earth system sciences, Journal Year: 2025, Volume and Issue: 29(2), P. 567 - 596
Published: Jan. 30, 2025
Abstract. While global hydrological models (GHMs) are affected by large uncertainties regarding model structure, forcing and calibration data, parameters, observations of output variables rarely used to calibrate the model. Pareto-dominance-based multi-objective calibration, often referred as Pareto-optimal (POC), may serve estimate parameter sets analyse trade-offs among different objectives during calibration. Within a POC framework, we determined optimal for WaterGAP (WGHM) in two largest basins Indian subcontinent – Ganges Brahmaputra, collectively supporting nearly 580 million inhabitants. The selected through multi-variable, multi-signature sensitivity analysis, were estimated using up four types observations: situ streamflow (Q), GRACE Follow-On terrestrial water storage anomaly (TWSA), LandFlux evapotranspiration (ET), surface (SWSA) derived from multi-satellite observations. our analysis ensured that parameters most influential identified transparent comprehensive way, rather number 10 16 had negative impact on identifiability process. Calibration against observed Q was crucial reasonable simulations, while additional TWSA basin helpful Brahmaputra obtain simulation both TWSA. Additionally calibrating ET SWSA enhanced overall performance slightly. We several objectives, with nature these closely tied physiographic hydrologic characteristics study basins. particularly pronounced basin, particular between SWSA, well ET. When considering observational uncertainty decreases cases. This indicates an overfitting singular observation time series algorithm. therefore propose algorithm identify high-performing Pareto solutions under consideration data.
Language: Английский
Citations
0
Journal of Cleaner Production, Journal Year: 2025, Volume and Issue: unknown, P. 144915 - 144915
Published: Jan. 1, 2025
Language: Английский
Citations
0
Journal of Hydrology, Journal Year: 2024, Volume and Issue: 637, P. 131417 - 131417
Published: May 26, 2024
Language: Английский
Citations
1
自然资源学报, Journal Year: 2024, Volume and Issue: 39(9), P. 2140 - 2140
Published: Jan. 1, 2024
Language: Английский
Citations
1
Published: Feb. 9, 2024
Abstract. While global hydrological models (GHMs) are affected by large uncertainties regarding model structure, forcing and calibration data, parameters, observations of output variables rarely used to calibrate the model. Pareto dominance-based multi-objective calibration, often referred as Pareto-Optimal Calibration (POC), may serve estimate parameter sets analyse trade-offs among different objectives during calibration. Within a POC framework, we determined optimal for WaterGAP Global Hydrology Model (WGHM) in two largest basins Indian subcontinent—the Ganges Brahmaputra, collectively supporting nearly 580 million inhabitants. The selected through multi-variable multi-signature sensitivity analysis, were estimated using up four types observations: in-situ streamflow (Q), GRACE Follow-On total water storage anomalies (TWSA), LandFlux evapotranspiration (ET), surface (SWSA) derived from multi-satellite observations. our analysis assured that parameters most influential identified transparent comprehensive way, rather number 10 16 had negative impact on identifiability process. against observed Q resulted be crucial reasonable simulations, while additional TWSA was basin helpful Brahmaputra obtain simulation both T. Calibrating also other observation enhanced overall performance enabled more accurate representation balance. We several objectives, with nature these closely tied physiographic hydrologic characteristics study basins. The particularly pronounced basin, particular between SWSA, well ET. When considering observational uncertainty decreases cases. This indicates an overfitting singular time series algorithm. We therefore propose algorithm identify high-performing solutions under consideration data. Recognizing uncertainties, anticipate actual lower roughly 90 % cases.
Language: Английский
Citations
0
Published: Feb. 27, 2024
Abstract. Understanding how physical climate-related hazards affect food production requires transforming climate data into relevant information for regional risk assessment. Data-driven methods can bridge this gap; however, more development must be done to create interpretable models, emphasizing regions lacking availability. The main objective of article was evaluate the impact risks on security. We adopted climatic impact-driver (CID) approach proposed by Working Group I (WGI) in Sixth Assessment Report Intergovernmental Panel Climate Change (IPCC). In work, we used CID framework select most indices that drive crop yield losses and identify important thresholds indices. When these are exceeded, probability increases. then examine two types (heat cold, wet dry) represented extremes considering different datasets, focusing maize soybeans central agro-producing municipalities Brazil. random forest model a bootstrapping experiment Then, applied Shapley Additive Explanations (SHAP) with XGBoost explanatory analysis caused impacts. found mean precipitation is highly CID. However, there window which crops vulnerable deficit. For soybeans, many Brazil, below 80 mm/month December, January, February represents an increasing losses. This end growing season those regions. case maize, similar pattern 100 April May. Indices represent variability. Nevertheless, including means remains recommended studying agriculture. Our findings contribute body knowledge critical informed decision-making, policy development, adaptive strategies response change its
Language: Английский
Citations
0
Published: April 15, 2024
Abstract. Understanding how physical climate-related hazards affect food production requires transforming climate data into relevant information for regional risk assessment. Data-driven methods can bridge this gap; however, more development must be done to create interpretable models, emphasizing regions lacking availability. The main objective of article was evaluate the impact risks on security. We adopted climatic impact-driver (CID) approach proposed by Working Group I (WGI) in Sixth Assessment Report Intergovernmental Panel Climate Change (IPCC). In work, we used CID framework select most indices that drive crop yield losses and identify important thresholds indices. When these are exceeded, probability increases. then examine two types (heat cold, wet dry) represented extremes considering different datasets, focusing maize soybeans central agro-producing municipalities Brazil. random forest model a bootstrapping experiment Then, applied Shapley Additive Explanations (SHAP) with XGBoost explanatory analysis caused impacts. found mean precipitation is highly CID. However, there window which crops vulnerable deficit. For soybeans, many Brazil, below 80 mm/month December, January, February represents an increasing losses. This end growing season those regions. case maize, similar pattern 100 April May. Indices represent variability. Nevertheless, including means remains recommended studying agriculture. Our findings contribute body knowledge critical informed decision-making, policy development, adaptive strategies response change its
Language: Английский
Citations
0
Published: May 15, 2024
Abstract. While global hydrological models (GHMs) are affected by large uncertainties regarding model structure, forcing and calibration data, parameters, observations of output variables rarely used to calibrate the model. Pareto dominance-based multi-objective calibration, often referred as Pareto-Optimal Calibration (POC), may serve estimate parameter sets analyse trade-offs among different objectives during calibration. Within a POC framework, we determined optimal for WaterGAP Global Hydrology Model (WGHM) in two largest basins Indian subcontinent—the Ganges Brahmaputra, collectively supporting nearly 580 million inhabitants. The selected through multi-variable multi-signature sensitivity analysis, were estimated using up four types observations: in-situ streamflow (Q), GRACE Follow-On total water storage anomalies (TWSA), LandFlux evapotranspiration (ET), surface (SWSA) derived from multi-satellite observations. our analysis assured that parameters most influential identified transparent comprehensive way, rather number 10 16 had negative impact on identifiability process. against observed Q resulted be crucial reasonable simulations, while additional TWSA was basin helpful Brahmaputra obtain simulation both T. Calibrating also other observation enhanced overall performance enabled more accurate representation balance. We several objectives, with nature these closely tied physiographic hydrologic characteristics study basins. The particularly pronounced basin, particular between SWSA, well ET. When considering observational uncertainty decreases cases. This indicates an overfitting singular time series algorithm. We therefore propose algorithm identify high-performing solutions under consideration data. Recognizing uncertainties, anticipate actual lower roughly 90 % cases.
Language: Английский
Citations
0
Published: Feb. 14, 2024
Language: Английский
Citations
0