Assessing the river forecasting capabilities of the Flux-Penn State Integrated Hydrologic Model and the Antecedent Precipitation Index-Continuous Model for the Little Juniata River Basin
Open Access
Author:
Kramer, Ryan Joseph
Area of Honors:
Meteorology
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
Kenneth James Davis, Thesis Supervisor Chris Eliot Forest, Thesis Honors Advisor
Keywords:
hydrology model comparison river forecast lumped distributed
Abstract:
Researchers have maintained discussion regarding the effectiveness of distributed, physical hydrologic models as river forecasting tools compared to the lumped, conceptual models currently used prevalently in the National Weather Service River Forecasting Centers. We assess the river forecasting capabilities of the distributed, physically-based Penn State Integrated Hydrologic Model, coupled with the Noah land surface model (Flux-PIHM), in comparison to the lumped, conceptual Antecedent Precipitation Index (API)-Continuous model. We produce and analyze reanalysis model discharge output from Flux-PIHM and the API-Continuous model for the year 2010 at the Spruce Creek stream flow gauge in the Little Juniata River Basin in Central Pennsylvania. Twelve precipitation events were selected from the year 2010 for further analysis. We evaluate, in relation to USGS stream flow observations, each model’s ability to accurately simulate peak discharge magnitude during a storm event, the elapsed time between the start of the event and the occurrence of the peak discharge value and the total runoff. We also compare multiple precipitation datasets to determine the effects that alternative forcing data may have on model output. Results indicate that, among other trends, Flux-PIHM overestimates base flow and peak discharge during the winter months while more accurately simulating peak discharge in the summer months compared to API-Continuous. Furthermore, both models simulate shorter time to peak discharge compared to observations for a majority of the events. Many of the possible causes for the identified trends in this study point to a need for various improvements to Flux-PIHM and API-Continuous calibration and parameterization.
