FastFind »   Lastname: doi:10.1029/ Year: Advanced Search  

Water Science and Applications

 

Keywords

  • Watersheds—Mathematical models
  • Watershed management—Mathematical Models
  • Hyrdrologic models

Index Terms

  • 1860 Hydrology: Runoff and streamflow
  • 1836 Hydrology: Hydrologic budget
  • 1863 Hydrology: Snow and ice

Article

WATER SCIENCE AND APPLICATION, VOL. 6, PP. 255-266, 2003

Use of a priori parameter-estimation methods to constrain calibration of distributed-parameter models

G. H. Leavesley, L. E. Hay, R. J. Viger, and S. L. Markstrom

Over-parameterization, ungauged basins, and the assessment of the impact of land-use and climate change are a few of the problems that limit the use of calibration techniques for distributed-parameter models. One approach to addressing these problems is the use of a priori parameter-estimation procedures to minimizing the number of parameters to be calibrated, or to obtain parameter values where calibration is not possible. A set of modeling and analytical tools is being developed using the US Geological Survey's Modular Modeling System to facilitate the development and evaluation of objective a priori methods. Initial testing of these tools was conducted on basins in the Rocky, Sierra-Nevada, and Cascade Mountain Regions of the United States, a priori parameter estimates were made for the USGS distributed-parameter model PRMS using available digital datasets of terrain, soils, vegetation type and density, and climatological data. Only the Rocky Mountain basin had an acceptable uncalibrated performance. Model performance for all basins improved as the parameters calibrated were increased incrementally from none, to those affecting the water balance, then hydrograph timing, and then all soils and vegetation related parameters. Problems were identified in the use of the forest-density dataset as a surrogate for forest cover density. A full evaluation of the soils dataset for determining available water-holding capacity was not possible due to the insensitivity of the model to this parameter in these snowmelt basins. Key issues in a priori parameter estimation for this limited application were identified to include regional climatic and physiographic differences, dataset limitations, and selection of measures of parameter and model performance. These will be addressed in an expanded research effort using tens of basins in different climatic and physiographic regions of the United States and the world.

Citation: Leavesley, G. H., L. E. Hay, R. J. Viger, and S. L. Markstrom (2003), Use of a priori parameter-estimation methods to constrain calibration of distributed-parameter models, in Calibration of Watershed Models, Water Sci. Appl., vol. 6, edited by Q. Duan et al., pp. 255–266, AGU, Washington, D. C., doi:10.1029/WS006p0255.

Cited By

Please wait one moment ...