Optimizing Hydrologic Model Selection for Low-Flows

Undergraduate Honors Thesis

Citations

Citeable URL: https://scholar.colorado.edu/concern/undergraduate_honors_theses/v979v364x

Abstract

Conceptual rainfall runoff models (CRRM) are used to predict the flow characteristics of waterways by correctly classifying the hydrologic exchanges occurring in the natural environment. Their practical applications range from predicting catchment yields to filling in gaps in stream flow data. Many of these models were devolved for predicting and quantifying mean and peak flows and have shown difficulty in replicating the low-flow hydrology of a waterway. We collected stream flow, rainfall, and evapotranspiration data from 16 gauges across Australia over a 30 year period. The gauges displayed four different types flow characteristics. We then tested the ability of three commonly used CRRMs: SimHyd, Sacramento, and Australian Water Balance Model (AWBM) to predict the low-flow hydrology of the 16 different catchments. We found that the predictive performance of AWMB was consistently better than the two other models. While SimHyd fit well to many flow duration curves, it struggled to model low-flows correctly. The information gathered from this study suggests that AWMB should be used when trying to predict low-flow hydrology. While AWBM was the best model, many improvements can still be made in consistently replicating the low-flow hydrology of intermittent systems.

Creator