Get access

Identification of rainfall–runoff model for improved baseflow estimation in ungauged basins

Authors

  • Jos Samuel,

    Corresponding author
    1. Department of Civil Engineering and School of Geography and Earth Sciences, McMaster University, 1280 Main Street West, Hamilton, Ontario, Canada L8S 4L7
    • Department of Civil Engineering and School of Geography and Earth Sciences, McMaster University, 1280 Main Street West, Hamilton, Ontario, Canada L8S 4L7.
    Search for more papers by this author
  • Paulin Coulibaly,

    1. Department of Civil Engineering and School of Geography and Earth Sciences, McMaster University, 1280 Main Street West, Hamilton, Ontario, Canada L8S 4L7
    Search for more papers by this author
  • Robert A. Metcalfe

    1. Aquatic Research and Development Section, Ontario Ministry of Natural Resources, c/o Trent University, DNA Building, 2140 East Bank Dr., Peterborough, Ontario, Canada K9J 7B8
    Search for more papers by this author

Abstract

Baseflow is an integral component of environmental flow prescriptions to mitigate the impacts of flow regime alteration on the ecological condition of rivers and often the most contentious issue in project planning, particularly for waterpower development. Baseflow prescriptions to meet ecological objectives are increasingly derived from a natural reference condition, often a simulated streamflow time series for an ungauged basin. Accuracy of baseflow estimates is important for identifying ecological water requirements and possible ecosystem and economic tradeoffs with confidence. Estimating baseflow time series in ungauged basins to accurately quantify a natural reference condition in all regions of Ontario is particularly difficult given the heterogeneity in landscape, climate, and size of the basins where simulated time series are required. To identify the optimal hydrologic model for baseflow simulation in all regions of Ontario, five variants of McMaster University-Hydrologiska Byråns Vattenbalansavdelning (MAC-HBV) were tested using a combination of possible model parameters, linear and nonlinear storage-discharge relationship in deep soil layers, and different criteria for optimizing model parameter sets. It was found that the hydrologic model which used a nonlinear storage-discharge relationship, a larger range of model parameters and included low flow criteria in the optimization procedure showed the best performance for baseflow estimates. The optimal model was then combined with a coupled regionalization method to improve baseflow estimates in ungauged basins compared with the original MAC-HBV model. This included a ∼20% increase in the median of Nash Sutcliffe efficiencies and a 50% reduction in the volume errors for the tested basins. The resulting model provides a tool that can be used reliably throughout Ontario to simulate streamflow and baseflow time series in ungauged basins and to calculate baseflow indicators and criteria for aquatic ecosystem assessments of planned flow regime alterations. Copyright © 2011 John Wiley & Sons, Ltd.

Ancillary