2nd UF Water Institute Symposium Abstract

Submitter's Name Anna Cathey
Session Name Poster Session: Hydrologic, Biogeochemical and Ecological Processes 1
Category Hydrologic, biogeochemical and ecological processes
Poster Number 205
Author(s) Anna Cathey,  University of Florida, Ag. & Bio. Engineering
  Rafael Munoz-Carpena,  University of Florida
  Gregory Kiker, University of Florida
  Global Sensitivity Analysis of a Rainfall-Runoff Model in a Largely Ungauged Watershed, Okavango Basin in southern Africa
  The Okavango Basin and Delta comprises a unique hydrologic system in southern Africa that delivers an important flood pulse for wildlife and humans from humid Angola to arid Namibia and Botswana. The Angolan portion of the Basin contributes that vast majority of the rainfall and occupies 60 percent of the 192,500 km2 system. Hydrologic monitoring in Angola was suspended because of a civil war that took place between 1975 and 2002. As a result rainfall and flow data is unavailable for much of the contributing watershed. A hydrologic model of the Okavango Basin is especially important as management decisions concerning development in post-war Angola arise. The Pitman model, which is embedded in the GIS user interface SPATSIM, has been used to model the river flow in the Okavango River. This is a rainfall runoff model that was calibrated on 12 years of data and tested on an additional six years of data. Global uncertainty and sensitivity analyses were conducted through this research in order to quantify model reliability along with parameter importance. Each of the 41 input parameters in the Pitman model is examined for first and higher order sensitivities and the total model uncertainty is determined. The Morris Method is a qualitative sensitivity analysis technique that is used to screen the input parameters in order to detect those that are the most sensitive. The variance based extended FAST quantitative uncertainty and sensitivity analysis method is then used on the subset of most sensitive parameters. This global uncertainty and sensitivity analysis of the Pitman model in the Okavango Basin will aid managers, scientists, and modelers in making decisions regarding the risk of development scenarios and allocating monitoring resources as well as improve our overall understanding of the system.