Is the physical hydrologic model HYPE well suited for the simulation of water quantity in North-American watersheds? - A modelling experiment with the newly developed RDRS meteorological reanalysis data
Abstract
In the frame of the Great Lakes Runoff Inter-comparison Project funded by the Integrated Modelling Program (IMPC) for Canada as part of the Global Water Futures (GWF) project, the HYPE model was selected to simulate water availability in transboundary North-American watersheds. HYPE - Hydrologic Predictions for the Environment - is a semi-distributed hydrologic model that is capable of simulating hydrological processes in ungauged catchments because of its unique combination of elevation, soil type and land use. For this project, HYPE is initially setup for the Lake Erie watershed, but the modelling domain is being extended to the whole Great Lakes Basin and the Nelson Churchill river basin (NRCB) in an effort to develop strategies to handle cross-border issues of data availability. This presentation will report on the modelling effort in the Lake Erie watershed.
For model calibration, 28 non-regulated gauge stations located upstream and 31 regulated and non-regulated gauging stations located downstream of the Lake tributaries were selected. A set of five-year meteorological forcing data is used and it is derived from the Regional Deterministic Reforecast System in development at Environment and Climate Change Canada. A single objective function combining NSE and PBIAS computed on daily discharge is used to optimize the model parameters which are general or coupled to soil type and land use. Preliminary results after calibration are generally in good agreement with observations. The median values of NSE and PBIAS are 0.52 and -3.8 for the set of upstream gauge stations. For the downstream gauge stations, the median value for NSE is 0.48 while that of PBIAS is 5.9. The lowest performances are obtained for the gauge stations whose upstream areas are dominated by croplands and mosaic vegetation. The simulated streamflow are being spatially and temporally validated using flow signatures in addition to conventional metrics. Key hydrological processes such as evapotranspiration and soil moisture are also considered for validation against available remote sensing data products.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.H33M2162A
- Keywords:
-
- 1807 Climate impacts;
- HYDROLOGY;
- 1834 Human impacts;
- HYDROLOGY;
- 1847 Modeling;
- HYDROLOGY;
- 1873 Uncertainty assessment;
- HYDROLOGY