Sarah Raper, Climatic Research Unit, UEA.
s.raper@uea.ac.uk
I am a lead author on the IPCC Third Assessment Report Chapter 9: "Global Climate Models -Projections" and am responsible for the simpler climate models input. At the first lead authors meeting in Paris, 30th November - 2nd December 1998, we discussed a plan for the use of simple models. It was decided that the simple model of Wigley and Raper would be used in the TAR as a tool to produce results for the four recommended new SRES emissions scenarios and also to explore uncertainties associated with the climate sensitivity. Since the model is being used as a tool, the plan is to tune the simple model to the A/OGCMs featured in the CMIP2 comparison. It is hoped that the CMIP2 results, which are for a 1% increase scenario, will be sufficient for this. The fit can then be assessed using the historic/IS92a forcing A/OGCM results held at the IPCC Data Distribution Centre.
The two factors that dictate a climate model's response to a forcing change are the climate sensitivity and the ocean heat uptake. In collaboration with Jonathan Gregory at the Hadley Centre and Tim Osborn in CRU, I have made a detailed comparison between the HadCM2 model and our simple climate model. We have found that the climate sensitivity of the HadCM2 model is not a constant but increases with time. For example, when the forcing is 1% CO2 increase followed by constant 2xCO2, the climate sensitivity increases from about 2.5C over the next century to 3.8C after 900 years. The CMIP2 results include some deep ocean data which are needed to calculate the heat flux into the ocean and hence to calculate the effective climate sensitivities of A/OGCMs. This non-stationary nature of the climate sensitivity is an interesting and important result, which has implications for the use of simple models and will be featured in the TAR.
Since the simple model will be tuned to a number of A/OGCMs, there will result a large number of projections (four scenarios times the number of GCMS) which will represent the scatter produced by the different GCMs.
If the results of the analysis were of sufficient interest I would like to publish them separately if possible.
The fields required are detailed below and are needed for both the control and perturbed cases. Please note that, in addition I need the forcing change for a doubling of CO2 concentration.