cc: mnoguer@meto.gov.uk date: Fri, 19 Jun 1998 10:47:36 -0600 (MDT) from: Linda Mearns subject: Re: Regional models and the IPCC DDC to: m.hulme@uea.ac.uk (Mike Hulme) Mike, Maria, Yes, I know the PIRCS program and have been at some of their meetings. They will be using my stats package to do some of their comparisons. I think they are definitely doing some very good comparisons, but I also know that they are rather behind schedule, and I rather doubt they will have what Ray says they will have by the end of the year. However, we are all being pushed rather hard by the US Assemssment demands, so maybe they will. At any rate, the regional modelers I talked to (who are also involved with PIRCS) didn't indicate that they were participating in these particular runs. One of the problems with PIRCS is that they don't have resources to pay people to do runs, so this of course affects people's motivations, and priorities. (Not that the IPCC is about to provide any funding either). At any rate, I discussed doing some runs over western Canada, with Machenhuaer, Rene Laprise, John McGregor, and Filippo. They were all interested in doing this. We chose western Canada for its complex coasline, topography, and also because it isn't the US. And it would help the Canadians a bit, since they have already worked in the region, and they are having some difficulties with their model. So this would involve four different regional models. I think we should follow up on this. We (at NCAR) already have a smaller project going with John McGregor comparing the Australian regional model with the RegCM2 in the Southeastern US. Also, Machenauer showed some interesting results at a workshop I was at in Canada a couple of months ago. They were results from two different regional models using transient boundary conditions from, I believe, the ECHAM4. The two models produced different responses. Perhaps we (Maria, Filippo, Mike, me) could have a discussion about all this in Germany. Cheers, Linda >Ray, > >Thank you for your email. The IPCC Task Group set up to oversee the >establishment of the DDC has been preparing its summary report to go before >the IPCC Bureau in July. Regional modelling and the issue of selecting >regions for IPCC focus has been part of that discussion. > >At present, we have not felt it appropriate for the DDC to incorporate >results from RCMs for a number of reasons that I won't elaborate on here. >Clearly, IPCC will be reviewing the science of regional climate modelling >in the TAR and the Guidance Material on climate scenarios being prepared by >the TGCIA (Tim Carter co-ordinating) will also address the question of how >valuable RCM results are for impacts analysis. > >If the TGCIA and DDC win a renewed mandate from IPCC this summer then over >the next 1-2 years we may well look at criteria for the inclusion of RCM >results in the DDC and at that stage your offer may well be important. For >now, it remains a little premature. > >I look forward to meeting you again in September. > >Regards, > >Mike > > >At 16:58 02/06/98 -0500, you wrote: >>Dear Mike, >> >>It was a pleasure to meet you at last month's LINK conference. You may >>recall our conversation over dinner when we briefly discussed the PIRCS >>project. One of the prime motivations for our regional modeling work under >>PIRCS is to provide high resolution climate change projections for the >>impacts assessment community, especially in the context of the the IPCC >>TAR. Thus we are very interested in furnishing our results to the IPCC >>data distribution centre. >> >>By the end of this year, we expect to have at least 10 years of control >>climate and CO2 scenario results from at least three different regional >>models over North America, driven by HadCM2 output. There is a strong >>possibility that this will be expanded to 20 years of output from up to >>five different regional models. Below is appended an abstract for the >>upcoming University of East Anglia "Climate and History" conference which >>gives an overview of these runs. Negotiations are presently underway to >>fund a similar set of regional projections driven by a different GCM. >> >>Please let me know if you think there is a place in the DDC for these >>regional scenarios as a supplement to the global models. We look forward >>to providing the regional output for wider use. >> >>Regards, >>Ray Arritt >> >> >> >>Regional Climate Change for Impacts Analysis: >>Evaluating Ensemble Projections >> >>R.W. Arritt, E.S. Takle, W.J. Gutowski Jr. and Z. Pan >>Iowa State University, Ames, Iowa 50011 USA >>Telephone +1-515-294-9870, fax +1-515-294-2619 >>email rwarritt@iastate.edu >> >>J. Hesselbjerg Christensen >>Danish Meteorological Institute, Lyngbyvej 100, >>DK-2100 Copenhagen Ø Denmark >> >>Oral presentation is preferred; poster presentation is acceptable. >> >>The discussions following the Kyoto Protocol have emphasized the need to >>understand better the potential regional impacts of global climate change. >>Impacts on agriculture, water resources, human health, and ecosystems are >>tied to climate change over regions (provinces, counties or states) rather >>than to changes in broad continental or global averages. It is therefore >>imperative to understand and, ideally, to predict how global climate change >>is manifested at these regional scales. >> >>The horizontal resolution of present-day numerical general circulation >>models (GCMs) is too coarse to provide satisfactory depictions of regional >>climate change. Studies of regional climate-change impacts thus require >>methodologies to translate GCM output to smaller scales. Using the GCM >>output to drive limited-area atmospheric simulations of climate change on >>regional scales is a promising but relatively new approach and research is >>actively underway to assess its merits. In response to a challenge from >>the Working Group on Numerical Experimentation (WGNE) of the World Climate >>Research Programme (WCRP), we are coordinating a community-based project to >>examine the validity of the regional climate modelling approach and to >>assess its added value relative to GCM predictions. >> >>An international team of nine modeling groups from North America, Europe >>and Australia is conducting the first stage of this research, in which >>sub-seasonal (60 day) simulations are being performed for two extreme >>events in the central United States. Preliminary results indicate >>strengths and weaknesses common to these regional models as well as >>occasional substantial differences between models. The second stage is >>being performed using three of these regional models to simulate 10 years >>of climate for the central United States. These simulations are being run >>with three sets of lateral boundary conditions: >> >>· First, climate of the recent past (1979-1988, included within the AMIP >>period) is being simulated using NCEP/NCAR reanalysis output to provide the >>same initial and boundary conditions for all three of the regional models. >>These results can be compared directly to observed data in order to >>establish the models’ skill at replicating regional climate and its >>variability. Using three distinct regional models gives a preliminary >>indication of the robustness of regional climate projections. >> >>· Second, these three models will perform 10-year simulations of regional >>climate for the same location but with initial and boundary conditions >>supplied from Hadley Centre GCM simulations of present climate. By >>comparing these results with the preceding simulations and with >>observations we can identify biases that are attributable to driving the >>regional models with large-scale information from this GCM rather than from >>observations. >> >>· The third set of simulations will be identical to the second except that >>initial and boundary data will be taken from Hadley Centre output for >>future climate perturbed by increased greenhouse gases and by aerosol >>effects (around the year 2050). Comparison of these results with the >>preceding two sets allows us to interpret the projected regional climate >>change in light of the errors induced both by the regional models and by >>the use of the GCM-supplied initial and boundary conditions. >> >>We are coordinating our efforts with statistical downscaling research and >>other regional climate modeling projects such as the European Community >>MERCURE project. Output from these simulations will be prepared for use by >>impacts analysts for the Third Assessment Report of the Intergovernmental >>Panel on Climate Change. >> >> >>_____________________________________________________________________ >>Raymond W. Arritt tel (515) 294-4758 >>Associate Professor fax (515) 294-2619 >>Department of Agronomy, Iowa State University, Ames, Iowa 50011 USA >>e-mail: rwarritt@iastate.edu web: http://www.mesoscale.iastate.edu >> > -- ****************************************************************************** Quote of the week: ``Diplomacy is the art of saying "nice doggy" until you can find a rock'' Dr. Linda O. Mearns Phone: 303 497 8124 Scientist Fax: 303 497 8125 Environmental and Societal Impacts Group e-mail: lindam@ucar.edu NCAR P.O. Box 3000 Boulder, CO 80307