Battelle Pacific Northwest Laboratories
Integrated assessment projects underway at Battelle Pacific Northwest Laboratories (PNL) stress collaboration between interdisciplinary teams within PNL, and networks of researchers from other institutes who broaden the range of disciplinary and regional expertise represented in the project. PNL's major integrated assessment projects include parallel development of two integrated-assessment models, PGCAM and MiniCAM (Mini Climate Assessment Model), and a comprehensive review of the current state of social-science understanding relevant to global environmental change. This project, the State of the Art Report (SOAR), is being prepared by a broad international network of authors, editors, and advisors.
The two modeling projects both involve linking and reconciling modular submodels, some developed at PNL and some elsewhere Edmonds, Wise, and MacCracken 1994). PGCAM is a large, computationally demanding, process-based model, while MiniCAM is a reduced-form model intended to be fast and accessible enough to run on desktop personal computers. For a detailed description of MiniCAM, see the MiniCAM Model Guide.
MiniCAMMiniCAM integrates three existing models. The Edmonds-Reilly-Barns (ERB) energy-economic model represents long-term trends in economic output, energy use, and greenhouse gas emissions for nine world regions through detailed submodules representing energy resources, primary energy supply and demand, energy markets including world trade and electricity conversion, and fuel-specific emissions factors. Edmonds and Reilly (1985) and Edmonds et al. (1986) contain more details on the ERB. (For further documentation of ERB, see the Edmonds-Reilly-Barns Energy Model section.)
From these projections of greenhouse gas emissions, MAGICC (Model for the Assessment of Greenhouse-gas Induced Climate Change) (Wigley and Raper 1992) and SCENGEN (regional climate change Scenario Generator) provide estimates of atmospheric concentration, climate change, and sea level rise. (For further documentation of MAGICC, see the Model for the Assessment of Greenhouse-gas Induced Climate Change section.) These, in turn, yield market and non-market damages using simple damage functions drawn from the MERGE model (Manne, Mendelsohn, and Richels 1993). Recent results from MiniCAM include an analysis of the effects of advanced energy technologies on climate change (Edmonds et al. 1994) and an analysis of strategies (using results from MiniCAM and MERGE) to stabilize carbon dioxide concentrations (Richels and Edmonds 1994). A simple graphical spreadsheet-based interface has been developed for MiniCAM to permit easy real-time exploration of model runs and scenarios without having to directly modify either source code or data files.
PGCAMPGCAM is a large model under development as a collaboration between PNL, the University Corporation for Atmospheric Research (UCAR), and Texas A&M University. Like MiniCAM, PGCAM has a modular structure with major components representing human activities and emissions, atmospheric concentrations, climate and sea-level change, and ecosystems and impacts, but each component will have substantially greater process detail. The human activities component is represented by a set of computable general-equilibrium models being developed in parallel for 16 world regions through collaborative projects with researchers from each region. Collectively called the "Second-Generation Model" (SGM) to denote their familial relation to the ERB model, these retain ERB's detailed seven-sector representation of the energy sector, but add sectors for agriculture and other industruial production, and factor markets for land, labor, and capital. There is substantial technological detail on the supply side, permitting policies to influence technological development or investment to be modeled in substantial detail. A separate demographics module permits specification of regional trends in age-specific fertility, mortality, and migration levels. Labor-supply and savings decisions are specified externally, and expectations can be varied among several formulations. Planned developments include allowing the emission coefficients of particular activities to vary over time, and disaggregating the agricultural sector (Fisher-Vanden et al. 1993).
The atmospheric component of PGCAM uses two highly parameterized models, one for atmospheric chemistry and one for climate dynamics. These generate projections of global average temperature change from calculation of change in radiative forcing, equilibrium temperature change, and dynamic adjustment. Given a specified global temperature change, the model will look up archived results from past Global Circulation Model (GCM) runs to generate distributions of climate variables on a scale of 5-degree by 5-degree grid points.
The impacts component of PGCAM is based on further analysis of some of the models employed in the MINK project. The farm-level analysis using the EPIC crop model is being repeated for a much larger variety of crops and locations to generate a set of crop-specific productivity response surfaces as functions of four or five climate variables, using some form of sensitivity analysis at particular points on the response surface to represent adaptations. Parallel analyses are being conducted using detailed sectoral models for forestry and water. Impact estimates are being developed for the United States first, to be followed by an integrated North American study of agriculture and water by 1994-95, and subsequently by global studies.
For representing non-market ecosystem impacts, two approaches are being considered: one (as in the Massachusetts Institute of Technology project) that projects how the function of existing vegetation types in their present locations might change under changed climate and CO2, and an alternative approach that seeks to employ physiological process modeling to explain the location and dominance of particular plant types as a function of climatic variables. This approach has been employed to generate a prediction of the distribution of vegetation types under the present climate that agrees more closely with the observed distribution than the traditional approach of Holdridge zones, but has not yet been employed to predict vegetation shifts under higher CO2 and changed climate (Prentice et al. 1992).
Uncertainty will be incorporated into the project by developing probability distributions of future emission paths and a few key parameters such as climate sensitivity and response time. The team is also seeking plausible ways of representing "threshold case" probabilities, low probability, high-consequence events such as massive methane release from clathrates.
ERB, SGMs for six regions (U.S., Canada, Japan, India, South Korea, and Western Europe), and MiniCAM are operational and nonproprietary. SGMs for remaining regions, and the first regional PGCAMs (USA and North America) are projected for late 1995 and 1996. For further information on PNL's integrated assessment activities, including the modeling projects and the State of the Art Project, contact the following:
Global Environmental Change Program
ATTN: Jae Edmonds, Technical Leader
Battelle Pacific Northwest Laboratories
901 D Street, S.W., Suite 900
Washington, D.C. 20024
The next section is IMAGE 1.0 .