- It is not available in a user-friendly manner
The 11-member RCM information was not originally intended to be made publically available. User demand has resulted in it being made available via the Climate Impacts LINK project website. Hence, it is not considered part of the UKCP09 package and has less user support associated with it.
Users will not be able to access the 11-member RCM information via the UKCP09 User Interface, nor be able to create graphical output for the 11-member data.
- There are model biases from the historical simulations
The 11-member RCM simulations provide projections of absolute future climate rather than changes relative to the baseline (the UKCP09 probabilistic projections provide changes relative to the baseline). As with all model data, that from the RCM will contain biases, due to systematic errors of various sorts.
The effect of historical model biases can be removed by differencing a baseline period from the future projections to get a climate change – this can then be used to modify an observed climatology to obtain absolute climate for a future period. This does not mean that the future projections will be error free due to the uncertainty in modelling, but the historical model bias will have been removed. Note that this method only corrects biases in the mean and that biases in other statistical aspects of the data e.g. variance, extremes still remain.
- Enhanced detail in the projections should not be taken to imply higher accuracy
Although an RCM is recognised as a sophisticated downscaling tool, as with any downscaling it inherits all the uncertainties of the Global Climate Model (GCM) that drives it. Hence, the detailed fine-scale projections from the RCM should not be seen to provide greater accuracy. We recommend that any variable of interest should be validated by comparing its baseline statistics with the relevant observed values wherever possible.
- The RCM output cannot be used in conjunction with the UKCP09 probabilistic output
It is not robust to use a variable from the 11-member RCM projections in conjunction with other variables from the UKCP09 probabilistic projections. A user must choose whether they want to use the 11-member RCM projections or the UKCP09 projections for their analysis but not use them both together. This is because the two projections have been created in totally different ways. The 11-member RCM does not incorporate the large range of Hadley centre model projections nor the IPCC models which are incorporated into the UKCP09 projections. Hence the two sets of projections are not comparable to one another.
They are absolute values of future climate, not relative to the baseline.
- They do not incorporate the full range of uncertainty available through the UKCP09 projections
Unlike the UKCP09 probabilistic projections, the 11-member RCM only represents projections from the Hadley Centre model and does not take into account any other IPCC model projections. This is a primary limitation of the 11-member RCM information in comparison to the UKCP09 probabilistic projections, and should be kept in mind.
There are also structural processes and uncertainties that are not taken into account in the HadRM3 model, but which are in the UKCP09 probabilistic projections. The RCM does not include a full ocean or a carbon cycle.
Additionally only one emissions scenario (medium) is available for the 11-member RCM output. Users should be aware that the use of a single emissions scenario may not result in a robust decision-making process.