Unfortunately, HadGEM2 is affected by the same time line problem. In this case the shift in the time line does not comprise multiple years but only one month. As the bias-correction ensures that the historic seasonal profile is conserved consequences are less dramatic than in the IPSL case and unfortunately not readily visible in the data compared to IPSL where values had been shifted for several decades. This is why the problem went unnoticed for so long. However, although the starting level is adjusted it still means that the temporal trends are shifted, i.e. the temporal trend in the December data is interpreted as the trend in the January data. Therefore the data have to be updated. We are working on that.
At the same time is is very important for all of us to get the first data to work on the tools to do the analysis. The runs you have done so far with the current HadGEM2 data are therefore still very important for the progress of ISI-MIP. It is definitely more than a dummy data set as it will allow us to take a first look on model differences and trends in terms of global mean temperature given a plausible climate (change). Therefore we propose the following process:
A) If you are still preparing climate data inputs for your model simulations, you should move to the IPSL-CM5A-LR model, unless you need wind speed data, which is not yet available for IPSL.
B) If you have already processed the climate data inputs for your model and simulations are set up, please finalize the computations and supply the model outputs. This will allow for a first analysis and prepare the way towards the papers.
