CGRG Bibliography of Canadian Geomorphology
Search Results
Author : Bingeman, A.; Kouwen, N.; and Soulis, E.D.
Date : 2001.
Title : Using flood frequency curves to validate a hydrological model.
Publication : Annual Scientific Meeting of the Canadian Geophysical Union. May 14-17, 2001, University of Ottawa, Ottawa.
Issue :
Page(s) :
Abstract
There are several types of hydrological models available; WATFLOOD/SPL is a distributed physically-based model. This model subdivides a watershed into hydrologically significant subgroups and uses a physically-based description of the hydrologic processes. The large number of parameters complicates model calibration and validation. The calibration requires that the internal hydrological processes be examined to ensure they are operating correctly, and that appropriate streamflow volume is calculated. There are numerous validation techniques available. The first validation checks the calculation of streamflow. Another validation technique is to check internal parameters such as: soil moisture, evaporation, snow melt and accumulation, and groundwater flow. A final validation is to examine the model’s ability to re-produceobserved flood frequency curves. A flood frequency analysis is based on the annual maximum in each year. Therefore, good reproduction of the frequency curves indicates that the model reproduced the peak flows, in addition to streamflow volumes. The WATFLOOD/SPL model was calibrated and validated for the Columbia River Basin in B.C., using meteorological forcing data calculated by the High Resolution Boundary Layer model (Atmospheric Dynamics Corp.). There were 95 years of simulated streamflow available foranalysis. The observed streamflow database, from Water Survey Canada, contains 32 streamflow stations with an average of 34 years of data (and a range from 5 years to 91 years). The flood frequency curve to the right shows the comparison between the observed and simulated frequency curves for theoldest Columbia River Basin streamflow station (91 years). The simulated frequency curve was “shortened” to match the observed frequency curve, so that the comparison could be made. These data show that the WATFLOOD/SPL model was able to simulate the peak flows at this station.The WATFLOOD/SPL was able to reproduce the observed flood frequency curves at most of the streamflow locations. Hence, flood frequency curves canbe calculated from the WATFLOOD/SPL simulated streamflow. In addition, when a longer time series of simulated streamflow is available, it may be used to replace or augment the observed streamflow database to improve the flood frequency curves.
Bibliography of Canadian Geomorphology