On Sun, 1 Jul 2012 18:17:31 -0700, John Gerlach <[email protected]> wrote: > >What the agencies want is clear direction on how to plan for climate change. >Other than the obvious general tactics such as larger preserves are generally >better and connectivity is generally better there is little that the models can >do to provide the level of information that the agencies expect. For on the >ground planning the resolution of the Global models is too coarse (60-100 km >pixels) and they don't really do precipitation so you just have a temperature >increase. The Global models also don't take into effect important features such >as mountain ranges. The Regional models are also fairly coarse resolution (15 km >pixels) but do incorporate surface features. One problem with regional models >are that they are too coarse for looking at local effects which the agencies >want but they do pick up regional patterns. Another problem of usage is that the >output of the regional models is not intelligently utilized. Generally, the data >are almost always reported as the average temp or precipitation per day over a 3 >month calendar period which is not biologically relevant. Even if you were to do >the intelligent thing and lump the data by local seasons you still have to know >enough about how the climate actually works in the region to interpret the data. >Sure, most models predict more variable climate but exactly what does that mean >for a wolverine in the middle Rockies for example? PRISM data are now being >commonly used to model climate data at 800 m resolution but projecting it out to >50 years not to mention 100 gives you false precision even if the data are >accurate and there are a some known issues with the data. Finally, you have to >have baseline data to compare the model data to and that data is also modeled. >
Dear John and ECOLOG colleagues, Following on the points above, you may be interested in reading our recent paper in Landscape Ecology on high-resolution gridded historical climate (GHC) data products like PRISM. Like many others we are interested in understanding local-scale climate changes, but our initial analyses of these data led to more questions than answers. Since the historical GHC products are used in downscaling AOGCM predictions, some of the issues we found are likely relevant for the local-scale projections of future change. Beier CM, Signell SA, Luttman A, DeGaetano AT. 2011. High resolution climate change mapping with gridded historical climate products. Landscape Ecology 27(3):327-342 DOI:10.1007/s10980-011-9698-8 http://www.springerlink.com/content/b866r66086457600/ We're happy to share our climate trend and bias maps for the US Northeast upon request. Also, we recently completed a regional analysis of prediction error of the two GHC products in the paper above for the US Northeast, including 55 COOP stations common to both products. That paper is now being finalized for submission to the Int. J. of Climatology sometime this summer. Cheers, Colin Colin Beier, Ph.D. Research Associate ∙ Department of Forest & Natural Resources Management ∙ Adirondack Ecological Center ∙ Coordinator, GPES Coupled Natural & Human Systems SUNY College of Environmental Science & Forestry ∙ Affiliate Fellow, Gund Institute of Ecological Economics, University of Vermont 311 Bray Hall ∙ SUNY ESF ∙ Syracuse, NY 13210 ∙ web: www.esf.edu/faculty/beier ∙ voice: 315.470.6578 ∙ fax: 315.470.6535 ∙ skype: cmbeier
