Search site

Estimating flow in UK rivers

Objectives: The Centre for Ecology & Hydrology developed this mock case study (using dummy data) to estimate river flow changes across the river network of the UK.

How they used UKCP09 dummy data

1. Regional Climate Model (RCM) data for baseline and future time periods and for all the ensemble members were downscaling to a 1 km resolution (the required resolution) and input into the Grid-to-Grid hydrological model. The downscaling procedure used high-resolution information from a standard average annual rainfall dataset.

2. A change in river flow from baseline to future was calculated on a 1 km grid across the UK for each ensemble member.

3. A map showing estimated changes in peak annual average flow across the UK was developed for each ensemble member projection.

Next steps

The next step would be to decide how to present the data. It could be presented as graphs for a particular location showing the projected change in river flow associated with each ensemble member, or as a map of UK rivers showing the percentage change in peak river flow derived from averaging across all the ensemble members.

They could compare the regional climate model changes in the context of the full IPCC AR4 range of changes. The results could be communicated through reports, publications and refereed scientific journals.

Lessons learned

  • The RCM data were the only UKCP09 output that provided information that was transient and at the required temporal resolution while being spatially consistent across the grid cells.
  • Use of direct RCM output encouraged a critical assessment of the applicability of the resulting flow changes. The lack of probabilities associated with these projections avoided the possibility of unquestioning application of probabilities to determine absolute risks.
  • When RCM data are used in preference to the probabilistic data a simple analytical framework for assessing the likelihood of the various outcomes is not available. Expert judgment is therefore required to assess the applicability of the results.
  • There is only one emission scenario available through the RCM data (medium).
  • The sheer volume of RCM data makes the information hard to handle.
  • The RCM data are not available in what would typically be thought of as a user-friendly format. For this exercise it was only available in binary, post-launch it should also be available in .csv.
  • It was noted that the RCM models a 360-day year with 30 days in each month so it is hard to compare to observations.
  • 11 plausible results of river flow makes analysis more complex.

Find out more


  • Contact details: Vicky Bell, Centre for Ecology and Hydrology; Richard Jones, Met Office Hadley Centre (Reading Unit).
  • Downoad the full report [263kb] 
  • BADC website
  • Bell, V.A., Kay, A.L., Jones, R.G., Moore, R.J. (2007) Use of a grid-based hydrological model and regional climate model outputs to assess changing flood risk. International Journal of Climatology27, 1657-1671. doi:10.1002/joc.1539.