As I have discussed on my weblog; eg. see
and papers, e.g. see
Pielke Sr., R.A., R. Wilby, D. Niyogi, F. Hossain, K. Dairuku, J. Adegoke, G. Kallos, T. Seastedt, and K. Suding, 2011: Dealing with complexity and extreme events using a bottom-up, resource-based vulnerability perspective. AGU Monograph on Complexity and Extreme Events in Geosciences, in press
the presentation of multi-decadal predictions of changes in regional climatoogy has no underpinning of skill.
There is an article that is highlighted on the front page of the November 29 2011 issue of EOS which perpetuates the erroneouns claim that such predictions are accurate. I do not doubt the sincerity of the authors, nor the details of their analyses, except for the fundamental issue that their starting point, the multi-decadal global model predictions have shown no skill at predicting changes in regional climate statisitics in the coming decades.
These studies are misleading policymakers.
The article is
Gangopadhyay, S., T. Pruitt, L. Brekke, and D. Raff (2011), Hydrologic projections for the western United States, Eos Trans. AGU, 92(48), 441, doi:10.1029/2011EO480001.
The abstract reads [highljght added]
Motivated by a common interest in establishing data access for climate change impacts analysis, the U.S. Department of the Interior’s Bureau of Reclamation (referred to hereinafter as Reclamation) has collaborated since 2007 with federal and nonfederal entities to provide monthly gridded precipitation and temperature data from 112 contemporary climate projections (Coupled Model Intercomparison Project Phase 3 (CMIP3)) over the contiguous United States. The grid size resolution of this downscaled data archive (publicly available at http://gdo-dcp.ucllnl.org/downscaled_cmip3_projections/) is 1/8° latitude x 1/8° longitude (approximately 12 x 12 kilometers) and covers the period 1950–2099 [Maurer et al., 2007]. Downscaling is necessary to develop hydroclimate data (e.g., precipitation and temperature) from a coarse- resolution climate model grid to a higher-resolution grid, and the CMIP3 archive was downscaled using the statistical method of bias correction. Although approximately 1000 unique users to date have downloaded the precipitation and temperature information contained within the archive (commonly referred to as the bias corrected spatially downscaled, or BCSD-CMIP3, archive), these temperature and precipitation projections have not been used to consistently generate hydrologic projections over the United States and at fine enough scale to perform hydrologic impacts analysis and support local adaptation assessments. Without available hydrologic projections, planners typically develop and apply their own site-specific and local hydrology models to fill this information gap. However, this makes consistent regional intercomparisons of hydrologic impacts of climate change difficult.
As the text writes
To address this challenge, Reclamation has collaborated with the University of Washington and the Colorado Basin River Forecast Center (CBRFC) of the National Oceanic and Atmospheric Administration’s (NOAA) National Weather Service (NWS) to generate 1/8° gridded hydrologic projections over the western United States using the variable infiltration capacity (VIC) macroscale hydrology model [Liang et al., 1994]…….
The development of these BCSD-CMIP3 hydrologic projections (a total of 112 covering the period 1950–2099), including analysis for the eight major river basins, was documented [Bureau of Reclamation, 2011a] in support of Reclamation’s first Science and Engineering to Comprehensively Understand and Responsibly Enhance (SECURE) Water Act report to Congress [Bureau of Reclamation, 2011b]; both reports are available at http:// www .usbr .gov/ climate/.
It is a puzzle why the users of this information are not demanding that the downscaled results be able to skillfully predict changes in local and regional hydrologic statistics for the period 1950 to the present when real world data is available. This does not, of course, ensure skillful predictions for the coming decades (out to the year 2099 in the article), but it is a first step in developing some confidence in them. If they are going to have any real world value, the models must be able to predict changes in local and regional climatology. Until, and unless they do, these types of studies are not only a waste of money, but are leading policymakers to make decisions which do not have a sound scientific basis.