Parameterization of the GPFARM-Range model for simulating rangeland productivity

Authors: ADIKU, S.G. - Colorado State University; Ahuja, Lajpat; Dunn, Gale; Derner, Justin; ANDALES, A - Colorado State University; GARCIA, L - Colorado State University; Bartling, Patricia

Submitted to: Book Chapter
Publication Type: Book / Chapter
Publication Acceptance Date: 11/15/2011
Publication Date: 12/15/2011
Citation: Adiku, S.K., Ahuja, L.R., Dunn, G.H., Derner, J.D., Andales, A.A., Garcia, L., Bartling, P.N. 2011. Parameterization of the GPFARM-Range model for simulating rangeland productivity. In Ahuja, L.R., Ma, L. editors. Methods of Introducing System Models into Agricultural Research. Madison, WI: American Society of Agronomy, Crop Science Society of America, Soil Science Society of America. p.209-222.

Interpretive Summary: One of the major limitations to rangeland model usage is the lack of parameter values appropriate for reliable simulations at different locations and times. In this chapter we seek to show how the GPFARM-Range, a rangeland model, which has been previously parameterized, tested and validated for the central locations of the Great Plains, could be re-parameterized to extend its domain of application to other locations of the Plains. Two main parameter determination methods are proposed: (i) manual adjustment of default parameter values and (ii) direct empirical parameter determination from some experimental data. It was recommended that in view of the level of information and expertise required for the second method, at this point the users should follow the first method; comparing the simulated output with observed data in a statistical sense such that the sum of squared deviations between the simulated and observed is minimized. It is noteworthy, however, that the published literature is a major source of data that could be used for parameterizing models. A number of texts are suggested for consultation to assist in the empirical determination of parameter values. It is shown that by varying the default values of only a few key parameters, the GPFARM-Range model could simulate forage growth under varying weather and grazing conditions at Miles City, located in northern part of the Great Plains

Technical Abstract: One of the major limitations to rangeland model usage is the lack of parameter values appropriate for reliable simulations at different locations and times. In this chapter we seek to show how the GPFARM-Range, a rangeland model, which has been previously parameterized, tested and validated for the central locations of the Great Plains, could be re-parameterized to extend its domain of application to other locations of the Plains. Two main parameter determination methods are proposed: (i) manual adjustment of default parameter values and (ii) direct empirical parameter determination from some experimental data. It was recommended that in view of the level of information and expertise required for the second method, at this point the users should follow the first method; comparing the simulated output with observed data in a statistical sense such that the sum of squared deviations between the simulated and observed is minimized. It is noteworthy, however, that the published literature is a major source of data that could be used for parameterizing models. A number of texts are suggested for consultation to assist in the empirical determination of parameter values. It is shown that by varying the default values of only a few key parameters, the GPFARM-Range model could simulate forage growth under varying weather and grazing conditions at Miles City, located in northern part of the Great Plains