Comparison of infiltration models on disturbed soils using parameter optimization

Date Issued
March 1, 1980
Author(s)
Innes, Gordon
Advisor(s)
Donald E. Overton
Additional Advisor(s)
Richard L. Church
Dennis W. Weeter
Permanent URI
Abstract
The infiltration models of Horton, Green and Ampt, Philip and the SCs-CN method were reviewed and their parameters were optimized for a data set of 103 rainfall simulations. The parameters, after being fitted to simulate steady state storms, were related to soil characteristics, the most significant of which was the initial soil moisture. Results indicate that the goodness of fit is dependent upon the initial soil moisture. Some models fit well if the soil was wet; others fit well if the soil was dry. None fit all conditions well. The parameters of Philip's model and Green and Ampt's model were linearly related to the change in soil moisture. Good correlations occurred when the sample size was adequate.

The accuracy of the infiltration models was concluded to be dependent upon accuracy of the estimate of the time of ponding. Since Horton, Green and Ampt, and Philip do not give methods for estimating this time, however, another model with new parameters and assumptions would have to be used in conjunction with them. For this reason it is recommended that either the SCS-CN method or Mein and Larson's model be used because they are both capable of simulating the time prior to runoff and subsequent infiltration rate.

The results are applicable to disturbed denuded soils such as would be found on strip mined spoil banks, construction sites, and plowed fields. Infiltration information about soils in this condition is important for simulating the stormwater response and soil erosion of disturbed areas.

Degree
Master of Science
Major
Environmental Engineering
File(s)
Thumbnail Image
Name

Thesis80I554.pdf

Size

2.93 MB

Format

Unknown

Checksum (MD5)

ed5fb59b2fd13d6d9fcd7d67ee9c3851