What to be aware of
The UKCP09 projections are based on sound scientific methods provided by the Met Office, with input from over thirty contributing organisations. There are a few things that it is important to be aware of when using UKCP09 projections.
1 - The probabilities are an assessment of likelihood based on available information.
The probabilities in UKCP09 represent the relative strength of evidence supporting different plausible outcomes for UK climate, based on the climate models, physical insight, observational evidence and statistical methodology used to produce them. They may not capture all possible future outcomes, because, for example, some potential influences on future climate are not yet understood well enough to be included in climate models.
2 - Emissions uncertainty is not addressed using probabilities
One reason why uncertainty exists in climate projections is because future greenhouse gas and aerosol emissions are uncertain. UKCP09 deals with this uncertainty by providing projections for three separate emissions scenarios: low, medium and high. These scenarios correspond to the scenarios as defined in the IPCC SRES report as B1, A1B and A1FI respectively.
Uncertainty over emissions scenarios has not been explored in probabilistic terms because there is no agreed method to assess the relative likelihood of one emissions scenario against another.
More information about uncertainty in UKCP09 can be found in Chapter 2 of the Climate change projections report. Annex 1 of Climate change projections report provides more information about emissions scenarios.
3 - Downscaling method does not incorporate information from other regional climate models
UKCP09 is based on global-scale modelling of future climate change, which uses global models to simulate the regional patterns of change at scales of 300km and above. These regional patterns have been 'downscaled' to 25km-scale make them more useful for UK decision-makers.
The broad-scale changes in UKCP09 are based on information from ensembles of global climate models which capture both parametric and structural components of modelling uncertainty. The downscaling to the finer 25 km resolution is based on a perturbed physics ensemble (PPE) of the Met Office Hadley Centre regional climate model (HadRM3). This captures some of the parametric aspects of modelling uncertainty in the downscaling, but structural downscaling uncertainties are not captured, because results from other international regional models are not included.
More information is provided in Chapter 3 of the UKCP09 probabilistic climate projections report.
4 - Interpretation of information provided in the form of daily or hourly time series
The degree of confidence that can be placed in results from the current generation of global climate models is greater for long-term climate averages than it is for extreme events (such as extremely hot days or spells of sustained heavy rainfall) or for time series of daily or sub-daily values. The probabilistic projections in UKCP09 therefore focus on changes in the climatology of 30 year periods, such as monthly, seasonal and annual climate averages (for example change in August temperature or winter rainfall), or the intensity of the typical warmest day of summer or wettest day of winter.
However, UKCP09 does provide information on future daily climate through provision of a weather generator. This tool is capable of producing synthetic daily time series consistent with monthly, seasonal and annual changes in climatology sampled from the probabilistic projections. The weather generator provides plausible examples of such future time series, but does not add any climate change information beyond that supplied from the probabilistic projections.
More information about the weather generator can be found on the About the Weather Generator Report page.
The weather generator can be accessed through the UKCP09 User Interface .
5 - Probabilistic projections are not provided for all climate variables
The methodology developed for UKCP09 relies on results from both multiple runs of the Met Office Hadley Centre global climate model representing uncertainties in model parameters, and also from an ensemble of other global climate models representing uncertainties in model structure.
For most climate variables, the range of projections from both sources is similar. For some variables, such as snowfall rate, the results from other global climate models include examples well outside the range projected by the corresponding perturbed physics ensemble of the Met Office Hadley Centre model. Such inconsistencies imply that robust probabilistic projections cannot be provided in UKCP09, since the methodology requires a combination of results from the perturbed parameter and multi-model ensembles.
In other cases (such as soil moisture) it was not possible to provide probabilistic projections because data from other models was not available in a form that is compatible with the Met Office Hadley Centre model.
More information is provided in Chapter 3 of the UKCP09 Climate change projections report. Users interested in soil moisture data should refer to the 11-member RCM data.