Search site

Environment Agency worked example

Objectives: The Environment Agency developed this mock case study (using dummy data) to evaluate the potential occurrence of critical river water temperatures in summer that could affect upstream migration or be lethal to the Atlantic salmon stock of a southern England river.

How they used UKCP09 dummy data

1. Thresholds to salmon migration were determined using historical relationships between air (inferred from Central England Temperature record) and water temperatures in the river. The following thresholds were determined:

· At water temperatures 20°C Atlantic Salmon will not enter the river beyond the tidal limit.

· At water temperatures 23°C salmon will not migrate upstream.

· Temperatures of 28°C for 7 days is regarded as the 7-day upper lethal temperature.

2. An assessment was made as to whether there is a time lag between air temperature and water temperature. It was decided that a one-day time lag between air temperature and water temperature was statistically robust. This was achieved by assessing the events in the historical data where the specified thresholds were surpassed.

3. The change factors of projected future temperatures were extracted from the key findings statements for the upper and lower limits of the projections across emission scenarios and for the 10th, 50th and 90th probability levels of the medium emission scenario.

4. These five change factors were added to the historical data and the number of days were counted as to when the specified thresholds may be surpassed in the future (2080s).

5. The exact change in temperature required to exhibit the first lethal condition was calculated. The PDF was used to extract a probability of this occurring in the 2080s, according to the strength of evidence.

6. The plume plot was used to see at what point in the future it is projected to be more likely than not (> 50% probability level) that the lethal threshold will be crossed.

Next steps

Long-term monitoring is required for tracking environmental change, testing model projections and for evaluating the success (or not) of any interventions. This study could be replicated using the 11-member RCM data or output from the Weather Generator. There are no key findings available from either of these data sources but they both provide data at a daily temporal scale, which could be used to perturb the baseline data.

Lessons learned

  • The probabilistic information is useful for extracting change factors across a probability distribution for use in adjusting historical information. This project could have been undertaken equally successfully using the 11-member RCM data or the Weather Generator series. The appropriateness of these types of information are at the discretion of the user.
  • An aggregated area for Central England Temperature in the future projections would have been useful.

Find out more

  • Contact details: Rob Wilby, on behalf of the Environment Agency