From joannem at ucar.edu Mon Jul 17 11:19:33 2006 From: joannem at ucar.edu (JoAnne Martin) Date: Mon Jul 17 11:19:36 2006 Subject: [CCSM-LandWG] C-LAMP Message-ID: <44BBC6A5.8060009@ucar.edu> Hi! At the CCSM meeting you saw that we are progressing quite well in the C-LAMP intercomparison. We have some runs spun-up, observational comparisons started, and some analysis begun. However, we need to get more organized in the next stage, as well as making sure everyone who has an idea, metric or dataset to contribute is included in C-LAMP. In order to move to the next stage, we need to be more specific about metrics than we have been (http://climate.ornl.gov/bgcmip/metrics.html) We propose that the following sub-committee be formed to put together metrics that we will use to judge our models: Jim Randerson (lead), Steven Running and Dave Schimel. All interested volunteers should communicate their inputs to Jim Randerson, who will coordinate and compile the inputs. We suggest the subcommittee put together a one page document on a) philosophy of the metrics, b) list of metrics and c) prioritization of metrics. The metrics should include list of model variables or relationships among variables; their desired accuracy (desired as well as the threshold for rejection). We think it would be nice if the first draft of the metrics document were put together before we start looking at results (more than we already have). We request that a first draft be posted by August 7, 2006. Secondly, we need to organize our obs datasets. We have some ideas for putting data at ORNL DAC in the longer term (Jim is looking into this). In the meantime, we can try to collect what datasets we have into one location that is publically accessible, since our model output will be publically accessible. I can volunteer my web site, and will put what data we have compared against there (www.cgd.ucar.edu/tss/staff/mahowald/CLAMP). Any one who has put together a dataset that looks good for our comparison project, please contact Natalie or Jim to get it included in the intercomparison. PCMDI has already volunteered to hold the more cumbersome model output (need how to link to that here). Finally we need a place to put our diagnostic output and discuss things between meetings. Inez will continue to work with PCMDI and others to produce Taylor diagrams and a standard set of diagnostics related to the metrics agreed upon by the metrics sub-committee. It would be useful if the plots are posted on a wiki website so everyone can comment on the results and the website is password protected and accessible without tedious log-in procedures. This web site should be linked to Forrest's web site and the PCMDI model output web site. Inez will look into creating such a wiki website at Berkeley. Those interested in diagnosing the performance of the models or have thoughts about a wiki website (including hosting of the website), please communicate with Inez. Comments are welcome, Natalie for Scott Doney, Jim Randerson and Natalie Mahowald, co-chairs of the BGC working group. -------------------------------------------------------------------- Natalie Mahowald, Scientist, National Center for Atmospheric Research, PO Box 3000,Boulder, CO 80307, 303-497-1719, fax: 303-497-1348. Street Address: 1850 Table Mesa Dr., Boulder CO 80305. web: www.cgd.ucar.edu/tss/staff/mahowald ---------------------------------------------------------------------- -- JoAnne Martin Administrative Assistant Terrestrial Sciences Section NCAR/CGD Room 202 1850 Table Mesa Drive Boulder, CO 80307 303 497-1308 - ofc. 303 497-1695 - fax From joannem at ucar.edu Mon Jul 31 10:53:24 2006 From: joannem at ucar.edu (JoAnne Martin) Date: Mon Jul 31 10:53:26 2006 Subject: [CCSM-LandWG] Physical component biases and BGC Message-ID: <44CE3584.7010807@ucar.edu> Hello! As we discussed during the CCSM joint working group meeting, we have the opportunity to focus the attention of the physical models on particular areas that are hurting our BGC simulations, but might not be the highest priority in other groups. We can probably give the atmosphere model working group and the ocean model working group up to 10 metrics to match. I think we can think about these similar to a Taylor diagram, so we can ask them to match the mean, stddeviation and correlation of the observations. For this first iteration, let's think big picture, and make sure we are including everything we need, and then later we can define exactly the regions. To Atmosphere: Precipitation: 1. Amazon 2. Tropical Africa 3. Indonesia 4. South East US 5. High latitude (>60N) x. Some measure of extreme events also, especially for Amazon. How should we phrase this? matching the pdf of precip over 20 year time period? Temperature: 6. High latitude surface temperature (>50 N?) Winds 7. Surface winds (everywhere or some specific places?) To Ocean/sea ice: 1. Sea ice extent Arctic 2. Sea ice extent Antarctic 3. CFC simulation (we need to define exactly--this could be slightly painful for them to do, but exceedingly important for us). Others? Thanks, Natalie for Scott Doney, Jim Randerson and Natalie Mahowald, co-chairs of BGC working group.