Watershed

Oral

394396 - Propagation of Structural Uncertainty in Watersheds

Monday, June 4
2:00 PM - 3:30 PM
Location: Greenway IJ

A distributed hydrologic model is typically a non-linear function of several parameters that need to be estimated through calibration. Apart from measurement uncertainty, parameter uncertainty exists because of limited number of observations, while structural-uncertainty results from models that are simplified versions of reality. Often, observations are available at the outlet or select locations within a watershed, and model parameters are calibrated to replicate observed streamflow values. Subsequently, one can quantify uncertainty in streamflow estimation at the measured locations. We are interested in quantifying the structural uncertainty at unmeasured locations of the river network. If we assume that the observations are error-free, we need to propagate parameter and structural uncertainty. Propagation of parameter uncertainty is a relatively simple task, and we focus on the problem of structural uncertainty propagation in a river network. We propose to treat the hydrologic model as a stochastic one instead of a deterministic function to accommodate the errors due to the structural uncertainty. The challenge is to find a stochastic representation of the function which would enable us to propagate the structural uncertainty to other sites after calibration using data from measured locations. We will present results of a perturbation method for the structural uncertainty propagation, and discuss its strengths and limitations.

Abhinav Gupta

Graduate Student
Purdue University

Presentation(s):

Send Email for Abhinav Gupta

Rao S. Govindaraju

Professor
Purdue University

Presentation(s):

Send Email for Rao Govindaraju


Assets

394396 - Propagation of Structural Uncertainty in Watersheds



Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Propagation of Structural Uncertainty in Watersheds