date: Tue, 24 Feb 1998 17:52:37 -0800 from: "Williams, Larry" subject: RE: COSMIC summary -Reply to: 'Joel Smith' , 'Michael Hulme' Joel, What you have described for SCENGEN seems identical to my procedure. ie. my tn3 IS the ratio of (2xCO2-1xCO2)-at a specific location-to change in global average temperature for a specific GCM . The same is true for pn3 which is precip at a specific location normalized by the global average temp for that specific GCM. The deltaTglobal(time) that I refer to below is provided by the Energy Balance Climate/Upwelling Diffusion Ocean model which provides the global average temperature. Regards, Larry PS. Mike, Hugh Pitcher has suggested that I make you aware of our COSMIC model. Since Joel has provided your email address, let me use this opportunity to direct you to Michael Schlesinger's web page that has an announcement letter and a few sample screens from the model. The url is: http://crga.atmos.uiuc.edu If you would like a copy of the model just let me know. It is available with a no-cost license. Larry Williams |voice: + (650) 855-2695 EPRI |fax: + (650) 855-2950 3412 Hillview Ave | email:ljwillia@epri.com Palo Alto, CA 94304-1395 USA >-----Original Message----- >From: Joel Smith [SMTP:JSMITH@habaco.com] >Sent: Tuesday, February 24, 1998 5:23 PM >To: Williams, Larry >Cc: m.hulme@uea.ac.uk >Subject: RE: COSMIC summary -Reply > >Larry, > >I believe SCENGEN does it somewhat differently, but I am cc'ing Mike Hulme on >this to be sure. > >I believe SCENGEN uses the ratio of temperature change at a specific location >(2xCO2 - 1xCO2) to change in global average temperature for a specific GCM >and >holds that ratio constant while changing global average temperature. I think >precipitation is measures as the ratio of 2xCO2 to 1xCO2 for the given >location >and the ratio is normalized by the change in global average temperature. The >precipitation is slightly different, but still generally similar to COSMIC. > >Mike, please comment if I misinterpreted SCENGEN. > >Joel > >>>> "Williams, Larry" 02/24/98 06:03pm >>> >Joel, > >It looks ok--but I am not 100 percent sure how SCENGEN does the scaling. >Here is how I do it: >To get the "normalized" GCM data I divided the country specific GCM data >as follows: >* tn3 = "normalized" temp = [t2x -t1x]/(deltaT in ) = dimensionless >* pn3 = "normalized" prec, = [p2x-p1x]/(deltaT in ) = in/(month- ) >* tn3(i,j,k) and pn3(i,j,k) where i=1:158 countries, j=1:12 months, >k=1:14 GCMs >Then I multiply by the deltaTglobal(time) to get country specific temp >and precip. If that >is how SCENGEN does it, then everything is ok. > >Larry > >Larry Williams |voice: + (650) 855-2695 >EPRI |fax: + (650) 855-2950 >3412 Hillview Ave | email:ljwillia@epri.com >Palo Alto, CA 94304-1395 >USA > >>-----Original Message----- >>From: Joel Smith [SMTP:JSMITH@habaco.com] >>Sent: Tuesday, February 24, 1998 4:26 PM >>To: Williams, Larry >>Subject: COSMIC summary >> >>Hi Larry, >> >>I wonder if you would not mind reading the following brief summary of COSMIC >>to >>see if it is accurate. This will appear in a UNEP handbook which is an >>overview >>of options for assessing climate change impacts. >> >>thanks a lot, >> >>Joel >> >>********************************* >> >>In addition, the University of Illinois at Urbana-Champaign and the Electric >>Power Research Institute developed a scenario generator that runs on a desk >>top >>PC, called the Country Specific Model for Intertemporal Climate (COSMIC). >>It >>allows the user to choose between seven sulphate emissions scenarios, ten >>greenhouse gas stabilization scenarios (based on the IPCC (Schimel et al., >>1996) >>and the *WRE* stabilization scenarios (Wigley et al., 1996), outputs from >>fourteen GCM models. The model uses a >>energy-balance-climate/upwelling-diffusion-ocean model to calculate changes >>in >>mean global temperature and sea level on an annual basis out to 2200. >>COSMIC >>scales the GCM outputs to 0.5o cells and averages the changes in each half >>degree cell for each of 158 countries (Williams et al., in press). The >>scaling >>of the GCMs is done in the same fashion as SCENGEN (Larry Williams, EPRI, >>personal communication). Results for changes in temperature and >>precipitation >>are given for each month in up to the year requested by the user. >