Click on “Download PDF” for the PDF version or on the title for the HTML version.
If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.
WARMF: Model Use, Calibration, and Validation
Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.orgCitation: Transactions of the ASABE. 55(4): 1385-1394. (doi: 10.13031/2013.42249) @2012
Authors: J. W. Herr, C. W. Chen
Keywords: Calibration, Correlation coefficient, Decision support system, Error statistics, GIS, Pollution control alternatives, River basin, Validation, Water quality criteria, Watershed simulation model
Watershed Analysis Risk Management Framework (WARMF) is a comprehensive watershed model and decision support system designed for stakeholders to formulate alternatives for point and nonpoint source pollution controls, evaluate their technical ability to meet water quality criteria, and make changes to negotiate the preferred plan. WARMF uses physically based model input parameters to describe the watershed. Many of these are known from data or watershed-specific studies, but others are determined through the calibration process. The model calibration is performed by systematically adjusting model input parameters within their normal ranges to match the simulated results to the observed data. The GUI enables users to make changes to model input parameters and/or data and graphically compare the simulated results and observed data. WARMF calculates correlation coefficient, relative error, absolute error, cumulative volume balance, and other measures to evaluate the calibration quantitatively. Relative error and absolute error are the preferred measures of accuracy and precision used in WARMF calibration. Long-term observed data may be split into two periods, one for model calibration and the other for validation. New data may also be collected for model validation. A case study is presented to demonstrate the calibration of flow and electrical conductivity (EC) for the San Joaquin River watershed in California. Soil properties were adjusted so the simulated flow would match the measured flow data during both the winter rainy season and summer irrigation season. Soil chemistry inputs were calibrated to ensure long-term stability of pore water concentrations and so the simulated water quality of the San Joaquin River followed the magnitude and pattern of in-stream monitoring data. The simulation of flow produced a 1% relative error and 13% absolute error over the eight-year calibration period. The EC calibration had relative error of 5% and absolute error of 15%.(Download PDF) (Export to EndNotes)