Uncertainty in Regional Climate Sensitivity

updated 11 November 2003

PI: Ken Caldeira

Climate and Carbon Cycle Modeling Group
Energy and Environment Directorate
Lawrence Livermore National Laboratory
7000 East Ave., L-103
Livermore CA 94550 USA

Phone: (925) 423-4191
Fax: (925) 422-6388
Email: kenc@LLNL.gov

Web page: http://en-env.llnl.gov/cccm/caldeiracv.html (now with pdf's of papers!)

I would like to improve our understanding of the uncertainty in regional climate sensitivity using CMIP monthly-mean surface air-temperature output. Traditionally, the uncertainty in global mean warming to a CO2-doubling has been characterized as 1.5 to 4.5 C. What is the uncertainty in regional mean warming? I propose to address this question using CMIP monthly mean surface air temperature results. To separate global uncertainties in climate sensitivity from regional, at each time period I will divide the regional change in surface air temperature by the global mean change, thus expressing regional climate sensitivity as a fraction of the global climate sensitivity. I intend to

I expect to produce maps indicating the inter-model range of the ratio of regional to global mean climate change. This value can be appropriately multiplied by the global uncertainty to estimate regional uncertainty. I also plan to look at other statistics in addition to inter-model range. I think I can do this relatively quickly using CDMS and CDAT. I plan to work in coordination with Curt Covey (PCMDI). Furthermore, I plan to distribute preliminary results to Curt Covey, Karl Taylor, Ben Santer, Krishna AchutaRao, Starley Thompson, and Phil Duffy (all of LLNL) for feedback and critical suggestions. I would be delighted if others would be willing to offer guidance or assistance. This project will be conducted on a "hobby" basis, as I have no funding to pursue this. I hope it leads to a short paper in a high visibility journal.

I will also estimate from top of atmosphere fluxes the partial derivatives dFlongwave / dT, dFlongwave / dln(pCO2) and dFshortwave / dT, where T is surface temperature, and from these try to estimate climate sensitivity. For this purpose I will need only annual and global mean longwave and shortwave fluxes, pCO2, and surface temperatures.