Model parameter uncertainty analysis for an annual field-scale phosphorus loss model

Authors: Bolster, Carl

Submitted to: Kentucky Water Resources Research Institute Symposium
Publication Type: Proceedings
Publication Acceptance Date: 2/17/2015
Publication Date: 3/9/2015
Citation: Bolster, C.H. 2015. Model parameter uncertainty analysis for an annual field-scale phosphorus loss model. Kentucky Water Resources Research Institute Symposium. Proceedings.

Interpretive Summary:

Technical Abstract: Technical abstract: Agriculture can be a significant source of P loading to surface waters which can lead to water quality deterioration of P-sensitive water bodies. To mitigate the effects of agricultural activities on water quality, models are often used to assess the effectiveness of various conservation practices for reducing P loss. While model predictions of P fate and transport in the environment can provide useful information, an inherent amount of uncertainty exists with all model predictions, regardless of how complex or “physically-based” they may be. Sources of uncertainty include errors that are introduced when approximating complex physical phenomena with simplified mathematical models, the inherent amount of randomness within natural systems, measurement errors in the model input variables, and errors associated with the model parameters. The magnitude of the errors introduced from these different sources will depend on the validity of the model assumptions, the complexity of the model, the quality of the input data, and on how well the various model parameters have been estimated. This study presents results from an analysis investigating the effects of model input and parameter error on prediction uncertainties of P loss using the Annual P Loss Estimator (APLE) model, an empirically-based spreadsheet model developed to describe annual, field-scale P loss when surface runoff is the dominant P loss pathway. The specific objectives of this study were to 1) estimate the model parameter uncertainty associated with five internal regression equations used in APLE, 2) estimate uncertainties associated with model input variables based on uncertainties reported in the literature, and 3) to evaluate how the model input and parameter uncertainties affect uncertainties associated with field-scale predictions of P loss.