From George Hurtt: Thanks all, I have added some comments below marked with a ">". One general comment that concerns me is the call for a comprehensive inventory in the NACP is fine print, that became clear is not on any agencies radar screen. I believe that comprehensive repeated inventories, at the end of the day, may be the most important thing we could do to reduce uncertainties about the carbon balance of ecosystems. I believe they'll be needed regardless of what we do with atmospheric data, because few will believe results calculated with a complex inversion approach until they can actually see the "beef" (account for changes in carbon stocks) locally. That was certainly the case nationally. We should make an emphatic call for making carbon inventories comprehensive (all land, all carbon pools). -----Original Message----- From: Dennis Ojima [mailto:dennis@nrel.colostate.edu] Sent: Saturday, September 08, 2001 6:26 PM To: cfield; running; dennis@nrel.colostate.edu; schimel@ucar.edu; Mac Post; Paul Moorcroft; george.hurtt@unh.edu; David Price; Peter Thornton; Changsheng Li Subject: some modeling thoughts Common data sets to initialize models for net C exchange that provide information on state conditions of C pools, land use histories, soil properties, and climate. > I am not sure what the word "Common" adds. Clearly we need datasets to initialize and test the models. These data sets should also be evaluated and subjected to peer review and widely available. This list is not bad, but also not very informative. We should mention integration with existing datasets such as FIA, Ameriflux, LTER, etc... Point to missing kinds of data that are not being collected otherwise, such as Chris' kilo-site idea, and voice support for important data missions in trouble such as VCL. Develop a terrestrial carbon modeling consortium for continental scale computation of the NEE to facilitate a common data, computational and analytical framework and to facilitate model experiments to test sensitivity of net C exchange to environmental factors including climate, N deposition, land use change, urban environmental conditions etc > This is an interesting idea. Consortiums are good for a small group of investigators working closely together. Too large and they become unwieldy and less productive. I also worry that consortiums can have a negative effect of being exclusionary, which may not be a good idea for a goal as large and important as NACP. Develop appropriate diagnostics for evaluating seasonal and interannual NEE/NBP relative to flux time-series > OK, but this should be kept in the context of the overall picture. We need more of the "integration" language and ideas that we discussed in the session. Also, where will these "flux-time series" come from? We have them at tower sites, perhaps. But models and other data sources will be needed to generate those at larger scales. Model experiments to test sensitivity of net C exchange to environmental factors including climate, N deposition, land use change, urban environmental conditions etc >Yes! Common data interface with atmospheric analysis and biospheric modeling >Again, not sure what "common" means. Different models and modelers will want to approach this problem differently, and that diversity is a good thing. The most integrative models, at the end of the day though, will need to produce gridded products of net flux at some time frequency so they can be used by the atm inversion groups. The demands on biogeochemisty models is quite different than that on biospheric models. Wetland modeling and CH4 production and consumption modeling >CH4 is very important and should be emphasized. But there are many other sources than wetlands. Suggest deleting focus on wetlands and making emphasis on Ch4 more clear. thoughts on development of insitu measurements for carbon stocks that are needed for initializing models >Different models have different needs. We should make this point, but leave it general enough to be interpreted broadly. -- >David made two good points that I agree with and want to add to. We do need emphasis o n "natural" disturbances as well as land use. This of course needs to be from both a modeling and data side. Fires, large scale insect dynamics, blowdowns, etc all need to be included better. Again, I think we want to emphasize the integration across scales. Many of these processes are large terms not sampled well by site networks. David also mentioned point sources. I brought a general version of this idea up in our discussions in the land group, and Niki Gruber raised it in plenary. Generally, the idea of carbon transfers within NA will be very important for understanding spatial-temporal patterns of carbon fluxes within the region. Most ag land for example may on average be seen as sinks because most of the production from that land decays elsewhere. Even at the US scale, this is important as the US exports more wood products than it imports. It will be even more important locally. This is a point that should be emphasized, because atmospheric considerations make it especially important, and it may not be being emphasized without this plan and an emphasis on connecting to atmospheric measurements. -------------------------------------------------------------------------- ---- Dennis Ojima dennis@nrel.colostate.edu Natural Resource Ecology Laboratory B229 Natural and Environmental Science Building (NESB) Colorado State University Fort Collins, CO 80523-1499 Telephone: (970) 491 1976 Fax: (970) 491 1965 ------=_NextPart_000_0008_01C13B6C.53B28640