New ISI-MIP dataset available

Posted on June 20, 2012

In this email we would like to make you aware of the new complete climate data set that is now available, and to confirm the climate model order of preferences in light of the recently discovered mistake in our HadGEM2-ES data set.

Please note that for the purposes of the ISI-MIP fast-track impact model intercomparison, the available IPSL-CM5A-LR and soon available HadGEM2-ES data are the first part of the official "ISI-MIP data set".  All fast-track ISI-MIP intercomparison excercises for IPSL-CM5A-LR and HadGEM2-ES will be based on this version (see ISI-MIP-version numbers in the attachment). Possible further updates of the climate data might be useful for the long-track or sensitivity test but will not provide the basis for the fast track exercise.

In some more detail:

1. IPSL-CM5A-LR data set now available
Data for all RCPs (including wind data) for the IPSL-CM5A-LR model are now available on the DKRZ server. 

2. HadGEM2-ES remains GCM1, IPSL-CM5A-LR is GCM2
The HadGEM2-ES model will remain GCM1 (for which all RCPs are requested under the minimal setting, as well as some sensitivity studies); IPSL-CM5A-LR is GCM2 (for which only RCPs 2.6 and 8.5 are requested under the minimal setting). Data for GFDL-CM3, MIROC-ESM-CHEM, and NorESM1-M are being processed at the moment and will be made available shortly.

3. How to proceed
a) We encourage you to finish all simulations that are currently running with the old HadGEM2-ES data and to submit the results to the ISI-MIP output archive.  These early data sets will be invaluable in helping establish data processing pipelines, to get a sense of the spread in model output, and to start the discussion about possible reasons for this spread.

b) Running RCP 2.6 and 8.5 for GCM2 is the recommended next step for those who have either finished or not started with the old HadGEM2-ES data.

c) Updated HadGEM2-ES data will be available in the next few days and should be used to do the minimal setting for GCM1.

4. Extended deadline for model results upload
In light of the data delays and replacement, the 1 July deadline for the first results data set needs to be rethought.  Taking into account the feedback from modelling teams in the teleconferences, the new deadline is 15 July, 2012.  By this date, we ask you to submit results for RCP 2.6 and 8.5 for the IPSL-CM5A-LR climate data. This new deadline ensures that the process keeps on track for the IPCC WGII (31 January, 2013) deadline.

5. Global mean temperature is available
In spite of the recent trouble with the HadGEM2 data we are starting to think about the analysis of the incoming data. Therefore we are also uploading global mean temperature projections for all climate scenarios provided so far. Global mean temperature will represent the common x-axis for the impact projections at least for the synthesis papers.

6. Summary of the CMIP5 models
The attached plot sets IPSL-CM5A-LR and HadGEM2-ES (as well as the other models in the ISI-MIP set) in the context of other CMIP5 AOGCMs in terms of land-averaged temperature and relative precipitation changes.

Once again, we would like to apologise for the inconvenience caused by the need to replace the HadGEM2-ES data set.  We are doing everything we can to avoid such mistakes in the future, and are very grateful for your patience.