|
Activity Number:
|
453
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #301000 |
|
Title:
|
Bayesian Data Assimilation for Parameter Estimation in Hydrological Systems
|
|
Author(s):
|
Darl D. Flake, II*+ and Mevin B. Hooten and Luis A. Bastidas
|
|
Companies:
|
Utah State University and Utah State University and Utah State University
|
|
Address:
|
Department of Mathematics and Statistics, Logan, UT, 84322,
|
|
Keywords:
|
data assimilation ; Bayesian statistics ; hydrology
|
|
Abstract:
|
Conventional approaches to parameter estimation for hydrological systems often involve mechanistic model calibration. Because of the potential complexity of these models, Bayesian methods provide an appropriate framework to assimilate data into the calibration process and thereby allow us to formally account for uncertainty while estimating parameter distributions. We compare and contrast the effectiveness of two such runoff models for appropriately characterizing the physical system under study (i.e., HyMod and the Sacramento Model). Specifically, we present a model formulation that explicitly accommodates the non-negative support of the data and focus the application on historical data from the Leaf River basin in Mississippi. We formally evaluate the utility of the statistical model as well as the mechanistic models involved and discuss additional aspects of the project.
|
