You are here:
MODELING UNCERTAINTY OF RUNOFF AND SEDIMENT YIELD IN TWO EXPERIMENTAL WATERSHEDS
Citation:
Hantush*, M M. AND L. Kalin**. MODELING UNCERTAINTY OF RUNOFF AND SEDIMENT YIELD IN TWO EXPERIMENTAL WATERSHEDS. Presented at First Interagency Conference on Research in the Watersheds, Benson, AZ, October 27 - 30, 2003.
Impact/Purpose:
To assess the impact of uncertainty in soil parameters on predicted hydrographs and sedimentographs over two small USDA experimental watersheds, using the distributed KINEROS model.
Description:
Sediment loading from agriculture is adversely impacting surface water quality and ecological conditions. In this regard, the use of distributed hydrologic models has gained acceptance in management of soil erosion and sediment yield from agricultural watersheds. Soil infiltration and antecedent moisture conditions can have a significant impact on soil erosion and sediment transport. The objective of this paper is to assess the impact of uncertainty in soil parameters on predicted hydrographs and sedimentographs over two small USDA experimental watersheds, using the distributed KINEROS model. We therefore attempt to discern whether variability of observed flow rates and sediment discharge can be attributed to uncertainty in soil and model parameters or to an imperfect modeling system. Probabilities for runoff and sediment discharge are generated from probability distributions of net capillary drive, saturated hydraulic conductivity, and other soil and channel parameters, using Monte Carlo (MC) simulations. The MC simulations are compared with measured values of runoff and sediment yield from the watersheds, and the results show that the saturated hydraulic conductivity, net capillary drive, initial relative saturation, and soil porosity have considerable impact on the probability distributions of peak flow, total flow, peak sediment discharge, and peak sediment discharge. The impact of uncertainty in the soil parameters on time to peak flow and sediment discharge was relatively moderate. The results have implications on the use of the hydrologic model for risk management and TMDL development in watersheds.