Please note that in the revised HadGEM2-ES data set, the period 2001-2010 is split into files spanning 2001-2004 and 2005-2010 (as opposed to 2001-2005 and 2006-2010). The reason is that the HadGEM2-ES model team, laying an emphasis on complete seasons, has started each model simulation in December, not January. Therefore this climate model moves from the /historical /period into the /RCP /period already in December 2005 (applying an interpolated radiative forcing), not in January 2006 as most other models. Thus a 2001-2005 forcing file would be slightly different for each RCP, causing a discontinuity between 2005 and 2006. Including the year 2005 into RCP-specific period ensures a consistent transition between the two periods.
For the other climate models (such as IPSL) we will keep the original partitioning (2001-2005 and 2006-2010).
Thank you very much to everybody who has already started to upload data to the server! We would like to take this opportunity to remind you to check the section 5.2 of the Simulation Protocol on file naming conventions. A file name should be all lower case, an example is as follows:
The information on which version of the variable is used should be included as comment in the net-cdf file. The files should be zipped to improve upload/download speed.