ABSTRACT:
Sensitivity analyses generally aim to quantify model result changes with respect to input changes. This study instead focuses on the site- and condition-specific nature of sensitivity, to demonstrate why users benefit by conducting their own sensitivity analyses as an essential step in model applications. This is demonstrated by employing the RUSLE model, which predicts average annual soil loss within a field. Sensitivity of computed soil loss to changes in selected parameters and variables is quantified under a range of sites and conditions.
Input parameters representative of several common forms are varied about base values for the variety of scenarios. The scenarios include a corn/soybean rotation near Chicago, Illinois, the same scenario moved to the Atlanta, Georgia, area, and a corn/soybean/winter wheat rotation near Topeka, Kansas. Both conventional and no-till management schemes are tested at the three locations. Results show that model sensitivity may vary both with site and with management conditions being simulated. These results serve to caution users of complex computer models not to rely upon sensitivity results that were conducted under conditions other than those being simulated A serendipitous result is the demonstration of how the form of presenting results afects user perspective of the importance of those results.
Footnotes
V. A. Ferreira is a mathematician with the USDA-Agricultural Research Service, TERRA Laboratory, Ft. Collins, CO; G. A. Weesies is a conservation agronomist with the USDA-Natural Resources Conservation Service, National Soil Erosion Research Laboratory, W. Lafayette, IN; D. C. Yoder is an assistant professor, University of Tennessee, Knoxville; G.R. Foster is laboratory director, USDA-ARS, National Sedimentation Laboratory, Oxford, MS; and K. G. Renard is a retired research hydraulic engineer, USDA-ARS, Southwest Watershed Research Center, Tucson, AZ.
- Copyright 1995 by the Soil and Water Conservation Society
This article requires a subscription to view the full text. If you have a subscription you may use the login form below to view the article. Access to this article can also be purchased.