| Title: | Methods to quantify and identify the sources of uncertainty for river basin water quality models |
| Authors: | van Griensven, A. and T. Meixner |
| Year: | 2006 |
| Journal: | Water Science and Technology |
| Volume (Issue): | 53(1) |
| Pages: | 51-59 |
| Article ID: | |
| DOI: | 10.2166/wst.2006.007 |
| URL (non-DOI journals): | |
| Model: | SWAT |
| Broad Application Category: | hydrologic and pollutant |
| Primary Application Category: | calibration, sensitivity, and/or uncertainty analysis |
| Secondary Application Category: | sediment loss and transport |
| Watershed Description: | 338 km^2 Honey Creek, a tributary of Sandusky River located in Ohio, U.S. |
| Calibration Summary: | |
| Validation Summary: | |
| General Comments: | Study describes an inverse uncertainty approach to identify uncertainty as a function of model input data and observed data uncertainty, using SWAT. ParaSol (Parameter Solutions), a modified version of the Shuffled Complex Evolution (SCE-UA) algorithm, SUNGLASSES, and CANOPI techniques are described and applied for an example, as well as additional residual error analysis and other statistical analyses. Confidence bounds for average daily flows and sediment loads are presented for each of the three techniques individually and two combinations. The ParaSol and SUNGLASSES techniques have been incorporated into the SWAT2005 automatic calibration routine. |
| Language: | English |
| Keywords: | Modelling; uncertainty; river basin; water quality |