A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Hydrological water quality modelling of nested meso scale catchments
In this study, the HYPE model was tested for simulation ofdischarge and stream water inorganic nitrogen (IN) concentration in two different mesoscale catchments of the German lower mountain range, the Selke (463 km2) and Weida catchments(99 km2). Results showed that IN concentration and daily IN load had a proportional relationship with discharge, indicating that IN leaching ismainly controlled by runoff in managed catchments. The HYPE model was proved to be capable of capturing dynamics and balances of water and IN load with a Nash-Sutcliffe coefficient above 0.83. PEST (Model-Independent Parameter Estimation & Uncertainty Analysis) and DREAM(ZS) (DiffeRential Evolution Adaptive Metropolis algorithm)were combined with the HYPE model to implement parameter calibration and uncertainty analysis. Results showed that multi-site calibration improved model performances at internal sites and decreased parameter posterior uncertainty ranges and prediction uncertainty, indicating the importance of observations from internal sites for spatially distributed prediction. Compared with the parameter calibration against biweekly nitrate-N concentration measurements, nitrogen-process parameters calibrated using daily averages of nitrate-N concentration observations produced better and more robust model performance on simulations of IN concentration and IN load, narrower posterior parameter uncertainty ranges and IN concentration prediction uncertainty. This is attributed to the fact that different hydrological conditions are covered under a temporal high resolution monitoring program. Both PEST and DREAM(ZS) are found to be efficient for hydrological water quality parameter calibration. However, DREAM(ZS)is more sound and appropriate than PEST because of its capability to evolve parameter posterior probability density functions and estimate prediction uncertainty objectively based on Bayesian inference.
Hydrological water quality modelling of nested meso scale catchments
In this study, the HYPE model was tested for simulation ofdischarge and stream water inorganic nitrogen (IN) concentration in two different mesoscale catchments of the German lower mountain range, the Selke (463 km2) and Weida catchments(99 km2). Results showed that IN concentration and daily IN load had a proportional relationship with discharge, indicating that IN leaching ismainly controlled by runoff in managed catchments. The HYPE model was proved to be capable of capturing dynamics and balances of water and IN load with a Nash-Sutcliffe coefficient above 0.83. PEST (Model-Independent Parameter Estimation & Uncertainty Analysis) and DREAM(ZS) (DiffeRential Evolution Adaptive Metropolis algorithm)were combined with the HYPE model to implement parameter calibration and uncertainty analysis. Results showed that multi-site calibration improved model performances at internal sites and decreased parameter posterior uncertainty ranges and prediction uncertainty, indicating the importance of observations from internal sites for spatially distributed prediction. Compared with the parameter calibration against biweekly nitrate-N concentration measurements, nitrogen-process parameters calibrated using daily averages of nitrate-N concentration observations produced better and more robust model performance on simulations of IN concentration and IN load, narrower posterior parameter uncertainty ranges and IN concentration prediction uncertainty. This is attributed to the fact that different hydrological conditions are covered under a temporal high resolution monitoring program. Both PEST and DREAM(ZS) are found to be efficient for hydrological water quality parameter calibration. However, DREAM(ZS)is more sound and appropriate than PEST because of its capability to evolve parameter posterior probability density functions and estimate prediction uncertainty objectively based on Bayesian inference.
Hydrological water quality modelling of nested meso scale catchments
Hydrologische Gewässergütemodellierung in genesteten mesoskaligen Einzugsgebieten
Jiang, Sanyuan (author) / Universitätsbibliothek Braunschweig (host institution) / Meon, Günter (tutor)
2014
Miscellaneous
Electronic Resource
English
DDC:
627
| TIBKAT | 2020
| HENRY – Federal Waterways Engineering and Research Institute (BAW) | 2020