Community Climate System Model Narrative

Scientific Objectives

The scientific objectives of the Community Climate System Model (CCSM) project are as follows:

develop and continuously improve a comprehensive climate modeling system that is at the forefront of international efforts to understand and predict the behavior of Earth's climate;

use this modeling system to investigate and understand the mechanisms that lead to interdecadal and interannual variability in Earth's climate;

explore the history of Earth's climate through the application of versions of CCSM suitable for paleoclimate simulations; and

apply this modeling system to estimate the likely future of Earth's climate to provide information required by governments in support of national and international policy determination.

The project is organized and conducted so as to involve a large community of scientists and stakeholders in the scientific and technical design, evaluation, and use of the ultimate product, CCSM.

Overall Activities and Accomplishments of the CCSM Project

The major objectives of the CCSM project during FY04 were to develop and validate a new version of CCSM called CCSM3. This model is designed to support the CCSM project's contributions to the Fourth Assessment Review (AR4) of the Intergovernmental Panel on Climate Change (IPCC). The development was completed by the late fall of 2003 through an intensive effort by the CCSM scientists and software engineering teams. The project then conducted and evaluated a large series of control runs to establish the basic characteristics of the climate as simulated by CCSM3. The Climate Change Working Group performed an extensive series of emission scenario experiments using CCSM3 for submission to the IPCC. The scientists and engineers generalized the CCSM3 code to produce realistic climate simulations across a wide range of spatial resolutions. There are also extensive control runs for each of the configurations to provide the basic documentation of the simulated climate system via the Web (www.ccsm.ucar.edu/experiments). The code, documentation, and standard experiments for CCSM3 were released to the community on June 23, 2004.

The CCSM Project Office organized and supported several working group meetings, CCSM Scientific Steering Committee and CCSM Advisory Board meetings, and the 9th annual CCSM Workshop. This workshop was held in Santa Fe, whereas all previous workshops had been held in Breckenridge. The 2004 Workshop set a new record for numbers of participants with a total of 331 attendees, 237 from universities and laboratories and 94 from NCAR.

This figure shows the growth in participants (NCAR and non-NCAR) at the annual CCSM workshops from 1996-2004.

The project office was also responsible for the administration of a number of Small Grants for Exploratory Research (SGERs) from NSF. These small grants produce major benefits to the CCSM project, particularly by attracting new talent and investment in the CCSM enterprise. Results from the SGER program include an outreach grant that was crucial in supporting and re-energizing a coordinated regional climate modeling program at NCAR. Another grant supported a PhD student (Lucinda Shellito) from the University of California, Santa Cruz, investigating paleoclimate regimes using CCSM. The SGER awarded to Ilana Wainer (visitor, University of Sao Paulo) supported short-term visits to NCAR for collaborative analysis of the behavior of the southern Atlantic Ocean in the coupled model. She has published her findings in several papers in both Brazilian journals in Portuguese, as well as English-language refereed journals. The SGER support provided to James Hack (CMS) and David Bader (visitor, Department of Energy, DOE) was used to hire a student assistant and to purchase a Linux dual-processor Xeon computer system for the purpose of running the cloud model and analysis software. The project successfully ported the University of Wisconsin Non-hydrostatic Modeling System to NCAR for use in parameterization studies with the CCSM Community Atmosphere Model (CAM) single column model. The group developed analysis procedures for examining output from the cloud model and the infrastructure required to allow for direct comparisons of cloud resolving model simulations with CAM3 single column model simulations. Follow-on work will continue under separate funding. A grant was awarded to Natalie Mahowald (TSS) and Andrew Gettelman (CMS) to hold a community workshop on the modeling and use of isotopes in CCSM. The workshop was held in January 2004, and the participants included 24 scientists from universities and other research institutions and 16 scientists from NCAR. The workshop attendees reviewed the science of isotopes as a tool to understand climate processes, the state of the science of isotope transport, and the transformation in climate models. They also addressed the issue of modeling isotopes in the CCSM and formulated a work plan and priorities for the development and testing of the capability in CCSM. A copy of the workshop report is available from the CCSM Project Office.

Present Scientific Capabilities of the CCSM Modeling System

The new version of the model, CCSM3, includes a number of important scientific and technical enhancements. These improvements are designed to facilitate a greater range of scientific experimentation by a larger community of climate modelers than any previous release. CCSM3 can be integrated over a wide range of spatial resolutions depending upon the goals of a specific application. The model can be run at a low resolution suitable for lengthy simulations of paleoclimate regimes and biogeochemical interactions, an intermediate resolution for development, and a high resolution for climate change assessments.

The major enhancements for the atmospheric model include new treatments of cloud and ice-phase processes; improved representation of the interactions among water vapor, solar radiation, and terrestrial thermal radiation; new treatment of the effects of aerosols on shortwave and longwave radiation; and new dynamical frameworks suitable for atmospheric chemistry. The updates to the land model are new methods to enable simulation of the terrestrial carbon cycle, dynamic vegetation, and improvements in land-surface physics to reduce temperature biases. The ocean component incorporates improvements to the representation of the ocean mixed layer, inclusion of solar heating by chlorophyll, and a new infrastructure for studying vertical mixing in the ocean. The enhancements for the sea ice model include a more accurate representation of surface stresses on ice, inclusion of salt in ocean-ice mass exchanges, and significant improvements in the numerical ice advection scheme.

In general, the climate for present-day conditions produced by CCSM3 has greater fidelity to the observed climate than simulations from previous generations of CCSM. The magnitude of the cold bias in the equatorial east Pacific has been halved in the new version. The warm bias in the winter land-surface temperatures at high northern latitudes has also declined, which will facilitate studies of the polar amplification of global warming using CCSM3. The thickness and spatial distribution of the Arctic sea ice are much more realistic in this new version due in part to better gradients in sea level pressure across the Arctic basin and to a much more realistic polar surface radiation budget. The snow climatology has improved because of better low-level moisture transport and atmospheric temperatures. The greater fidelity of the snow and sea ice for present-day conditions are important for studies of climate feedbacks in the cryosphere. The ocean exhibits better sub-surface zonal currents in the eastern and equatorial Pacific and smaller biases in shallow-ocean mixing and ocean heat uptake. Several aspects of the simulated El Niño-Southern Oscillation (ENSO) cycles are also more realistic, including the teleconnections between the equatorial and northern Pacific and the cloud radiative feedbacks.

Despite the improvements in the model fidelity, several significant challenges remain. The next CCSM workshop in June 2005 will include an integrated assessment of the systematic biases in the climate simulated with CCSM3. The excessive mass transport through the Drake Passage is directly related to the overestimate of surface wind stress over the southern oceans. The warm sea surface temperature (SST) biases off the western coasts of Africa and South America are due to an underestimate of stress and an overestimate of surface insolation. Preliminary studies indicate that the coastal SST errors affect the surface state over large areas of the Pacific and Atlantic Oceans. Other issues in the tropical climate include a double Intertropical Convergence Zone (ITCZ) in the Pacific and Atlantic, excessive precipitation in the western Indian Ocean, a semi-annual SST cycle in the eastern Pacific, and representation of the major modes of variability including ENSO. With the introduction of biogeochemical cycles in CCSM, it has become more urgent to address long-standing biases in continental precipitation and temperature, particularly over tropical Africa and the Amazonian basin. The CCSM Scientific Steering Committee is committed to focusing future development and experimentation on amelioration or elimination of these systematic errors.

Control Runs Accomplished and Data Available

The new version of CCSM3 was released to the community in June 2004. All source code and input data are freely available via the CCSM Web site (www.ccsm.ucar.edu/experiments). This model is significantly different from any previous version with major changes made to all the CCSM components. CCSM3 can also run on a much larger number of machines and associated platforms than previous releases, including vector machines and Linux clusters. The portability to Linux clusters, which are widely available in the university community, should help users run CCSM3 at their home institutions. The reintroduction of a vector capability will allow CCSM scientists to perform demanding simulations on the fastest computers available to the climate community.

New user-friendly scripts for building and running the system were introduced, producing easy-to-use methods to run the IPCC climate change experiments, as well as the flexibility to simulate climate over a wide range of spatial resolutions. The new build/run scripts also permit the straightforward porting of the model system to new machines by providing the user with the ability to easily tailor the scripts for their particular machine environment. The release was also accompanied by new built-in test facilities suitable for validating installation and verifying some types of model changes. CCSM3 was accompanied by a completely new flux coupler with new infrastructure that improves its performance and scalability on parallel supercomputers and accelerates multi-way communication among the component models. CCSM3 also includes the introduction of "dead" models, which provides the ability to easily test the CCSM3 software infrastructure. In conjunction with the release of the new CCSM3 model, the project produced the most comprehensive suite of control runs ever provided to the CCSM community. The following table indicates the number of simulated years for a variety of standard forcing scenarios. Each scenario was run at three resolutions (low, medium, and high). In addition, there are ensembles for some of these scenarios. For example, the T85_gx1v3 20th century control has a seven-member ensemble. The total volume of control simulations is approximately 4,000 simulated years. A large subset of output data from each control run has been made available on the Web via the Earth System Grid (www.earthsystemgrid.org ), which is a DOE project to distribute large collections of climate research data.

Run Length In Years
Scenario	T85_gx1v3	T42_gx1v3	T31_gx3v5
1990 control	700	1000	880
1% increasing CO₂	150	150	180
2xCO₂	150	150	150
4xCO₂	150	150	150
1870 control	750	300	150
20th century	100	100	100

The scenarios are:
1990 control	constant 1990s forcing
1% increasing CO₂	starting from a 1990 control, applies 1% annual increasing CO₂ forcing
2xCO₂	starting from where a 1% increasing CO₂ reaches 2 times 1990 levels, CO₂ held constant
4xCO₂	starting from where a 1% increasing CO₂ reaches 4 times 1990 levels, CO₂ held constant at 4x levels
1870 control	constant 1870s forcing
20th century	starting from an 1870 control, a time series of 20th century forcing is applied

The IPCC Simulation Production Process

The application of CCSM to IPCC represents a multi-institutional, distributed effort by scientists and software engineers at NCAR, Oak Ridge National Laboratory (ORNL), National Energy Research Supercomputing Center (NERSC), Lawrence Berkeley National Laboratory (LBNL), and partners in the Kyosei Consortium using the Japanese Earth Simulator (ES).

This figure shows IPCC CO₂ concentrations. In addition to CO₂, the simulations are run with solar, volcanoes, sulfate aerosols, ozone, and other greenhouse gases for the period 1870-2000. A spread of CO₂ concentrations for future forcing scenarios is used from years 2000-2200 that also include sulfate aerosols and ozone.

CCSM3 (T85 version) has been run to perform a suite of 20th, 21st, 22nd, and 23rd century climate simulations. These simulations are the CCSM contribution to the IPCC AR4. The experimental design and list of experiments was suggested by IPCC Working Group 1 and coordinated by the World Climate Research Program/Climate Variability and Prediction Program (WCRP/CLIVAR) Working Group on Coupled Models (WGCM).

The IPCC scenarios that are being run can be categorized into two groups: a set of "commitment scenarios" and a number of "science scenarios" that explore alternative future emissions outcomes. The IPCC Special Report on Emissions Scenarios (SRES) that were recommended to the global coupled climate modeling groups by IPCC Working Group 1 include B1 (a low forcing scenario), A1B (a medium forcing scenario), and A2 (a large forcing scenario). The A2, A1B, and B1 scenarios consist of eight-member ensembles run from the year 2000 to the year 2100. Then, commitment scenarios are run with the concentrations of all atmospheric constituents in A1B and B1 held constant at year 2100 values, and the model continues to the year 2200 with five-member ensembles. One of the ensemble members from each is continued for an additional 100 years to the year 2300 with constant year 2100 concentrations. There is also a 20th century commitment scenario that freezes concentrations at levels observed in the year 2000, and the model is run to the year 2100. A limited number of experiments performed on the ES examined the implications of future reductions in CO₂ concentrations.

This figure shows the CCSM3 IPCC Run Plan: The IPCC scenario runs have been done at NSF/NCAR (blue), DOE/ORNL (green), and DOE/NERSC (red). Parallel runs on the ES (pink) were made with the scientific and technical assistance of the Kyosei Consortium.

The CCSM3 and Parallel Climate Model (PCM) IPCC runs are being performed at NSF/NCAR, DOE/ORNL, and DOE/NERSC supercomputing sites. An additional set of parallel runs, using the same version of CCSM3 with forcing data sets prepared by NCAR, was run on the ES. Focused model development and significant dedicated computing resources provided by NCAR, DOE, and the Kyosei Consortium have kept the IPCC runs on track for completion in time to meet IPCC data submission and paper publication deadlines. Most recently, NERSC consultants assisted the CCSM team in running the NERSC A2 five-member ensemble as one single 1020 processor massively-parallel run.

The data products from CCSM3 and PCM IPCC simulations will be used for climate change impact studies and as boundary conditions to drive regional scale models. To support NSF and DOE regional modeling efforts, we are outputting two additional sub-daily data streams for each of the CCSM IPCC scenarios. By design, the raw and post-processed data products from the different IPCC scenarios are being kept at DOE and NSF sites where they were computed. This large distributed data set is being made freely available to the U.S. climate research and education community via the DOE Earth System Grid (ESG).

This figure shows CCSM3 DOE/NSF IPCC scenario runs as of October 4, 2004. Observed forcings (solar, volcanoes, greenhouse gases, sulfate aerosols, carbon aerosols, and ozone) are used during the historical period, from years 1870-2000. A variety of future forcing scenarios (20th century freeze, B1, A1B, and A2) are used from years 2000-2200 to simulate the most likely range of future climates. Two of the commitment scenarios (A1B and B1) will be executed out to year 2300.

The IPCC experiments are intended to address several outstanding climate simulation issues. First, how well can the CCSM3 simulate 20th century climate if most of the dominant forcings are included over that time period? In this regard, forcings used previously included natural forcings (solar and volcanoes) and anthropogenic forcings (greenhouse gases, sulfate aerosol direct effect, and ozone). It also included an important new forcing by carbon aerosols, using a present-day distribution of carbon aerosols that scaled the geographical pattern over the 20th century by population as an approximation of their varying effect with time. Future experiments already planned will examine the role of carbon aerosols in much more detail. The 20th century simulations with the CCSM3 show that the time evolution of globally averaged temperatures track the observed time series very well, indicating that the forcings we have included are relevant for simulating 20th century climate.

Second, given three SRES's for 21st century climate change, what possible changes could occur by the end of the 21st century? Results show that the global temperature increases are proportional to the forcing as could be expected, with the pattern of temperature changes showing greatest warming at the high latitudes of the Northern Hemisphere and over land areas, with least warming over the North Atlantic and Southern Oceans.

A third major issue is the nature of climate change that is committed by previous emissions and by emissions occurring during the 21st century. The analysis is based upon idealized experiments where all concentrations are held fixed, first at year 2000 values, and then at year 2100 values for the A1B and B1 scenarios. Results show that global temperatures continue to rise on the order of a couple of tenths of a degree for the next 100 years after concentrations are fixed, but the much greater commitment comes from sea level rise. Subsequent sea level increases from thermal expansion alone 100 years after fixing concentrations are on the order of a couple hundred percent. These results demonstrate that the threat from future climate change after stabilization comes more from sea level rise we are already committed to, rather than future increases of temperature.

The CCSM3 experiments for the IPCC AR4 involve the most extensive and highest resolution multi-member ensemble ever performed with a global coupled climate model at NCAR, with five-member ensembles for all of the experiments. We also have been coordinating a parallel set of runs with CCSM3 on the ES. Combining the multi-member ensembles run on the ES, the CCSM3 will have by far the largest multi-member ensemble data set of any of the international global coupled models run for the IPCC AR4. Thus, the CCSM3 has assumed a leadership role in the performance and analysis of these high profile climate change simulations for the AR4. The model simulations have been completed and a subset of requested model data is being shipped for archival at DOE's Program for Climate Model Diagnosis and Intercomparison (PCMDI). This data will become part of the largest multi-model climate analysis project ever attempted, as over 200 researchers from around the world have registered to analyze the multi-model data set, of which CCSM3 is a part, for the IPCC AR4. After the deadline for submission of manuscripts for the Journal of Climate special issue on CCSM3 in fall 2004, the entire CCSM3 data set will become available for analysis to the CCSM3 community.

Some Highlights of Scientific Results Using CCSM Models

a) Estimating Climate Sensitivity from Last Glacial Maximum

The second phase of the Paleoclimate Modeling Intercomparison Project (PMIP2) is coordinating simulations and data syntheses for the Last Glacial Maximum (LGM, 21,000 years before present) to improve the assessment of climate sensitivity. The important forcing for the LGM is not the direct effect of insolation changes, which are minor, but instead results from the large consequential changes in various elements of the climate system, including greenhouse gases, aerosols, ice sheets, and vegetation change. The changes in greenhouse gases are fairly well known but the others are not. Global climate sensitivity can in principle be estimated from the difference between LGM and present, from Q=a; DT, where Q is the radiative forcing, DT; the global average surface air temperature change, and a; the climate feedback parameter. The climate feedback parameter is related to equilibrium climate sensitivity according to D;T_2x=Q_2x/a;. While Q and DT;are uncertain, fortunately the signal associated with the changes between LGM and present-day conditions are large.

Using measurements from ice cores, atmospheric CO₂ concentrations at the LGM are estimated to be 185 ppmv, approximately 50% of present-day values. Global, annual mean surface temperature simulated by the slab ocean version of the CCSM3 shows a cooling of -2.8°C for LGM CO₂ levels and a warming of 2.5°C for a doubling of CO₂ (see figure below). Slab and coupled CCSM3 simulations that include the reductions of the other atmospheric trace gases and the large ice sheets covering North America and Eurasia at LGM give cooling in agreement with proxy inferences. The simulations indicate that atmospheric CO₂ concentration change explains about half of the global cooling at LGM.

This figure shows global, annual surface temperatures as simulated by slab ocean and fully-coupled versions of CCSM3 for past and future climate states.

Regional signatures of the climate response to changed LGM forcing are also an important measure of climate sensitivity and are key for understanding the processes governing global sensitivity. Patterns and amplitudes of annual surface temperature change are similar but opposite in sign between the LGM CO₂ and 2xCO₂ simulations with the CCSM3 slab ocean model (see figure below, top vs. middle). Cloud feedbacks, particularly feedbacks involving low clouds, play an important role in determining the surface temperature response in the model and show comparable patterns of change in the 2xCO₂ and LGM CO₂ simulations. Additionally, a LGM CCSM3 slab ocean simulation that includes the reductions of the other atmospheric trace gases and the large northern ice sheets indicates that the cooling over the oceans in the tropics and Southern Hemisphere subtropics are primarily (~85%) due to the reductions of atmospheric CO₂ (see figure below, middle vs. bottom). Comparisons of PMIP2 model simulations of surface temperature changes for these regions and paleoclimatic proxy data may allow the LGM to be used as a metric to identify estimates of climate sensitivity that are outliers. Other regions of scalable response will also be identified and quantified as part of the PMIP2 model-data comparisons for the LGM.

This figure shows the surface temperature change simulated by the slab ocean version of CCSM3 for doubled CO₂ compared to present (top), LGM CO₂ compared to present (middle), and Full LGM compared to LGM CO₂ (bottom).

b) The ENSO Phenomenon in CCSM3

The ENSO phenomenon is a natural irregular oscillation of the coupled tropical ocean-atmosphere system on interannual time scales. ENSO is manifest primarily as a warming of the equatorial Pacific Ocean and a redistribution of precipitation within the tropics. However, the effects of ENSO are felt worldwide due to the global atmospheric circulation response to tropical rainfall changes.

This figure shows the variance of monthly SST anomalies from (top) observations during 1900-1999 from the Hadley Centre Sea Ice and SST (HadISST) data sets and (bottom) the T85 CCSM3 control run during model years 400-499.

The ENSO phenomenon in the CCSM3 at T85 resolution exhibits a realistic spatial pattern and amplitude of SST variability in the tropical Pacific and Indian Oceans, with maximum variability in the eastern equatorial Pacific and along the South American coast. This represents a considerable improvement over the previous version of the model (CCSM2), which greatly overestimated the amplitude of the SST variability and its zonal extent (not shown). The T42 version of CCSM3 exhibits a similar pattern and amplitude of SST variability as the T85 version (not shown).

A commonly used index of ENSO is the area-averaged monthly SST anomaly (defined as the departure from the long-term monthly mean) in the equatorial eastern Pacific (5ºN-5ºS, 170ºW-120ºW). A power spectrum of this index shows that the ENSO period in the model is approximately 2 to 3 years (also in the T42 version of CCSM3), considerably shorter than the observed period (approximately 3 to 8 years based on years 1900-1999). The peak amplitude of the power spectrum is comparable in the model and observations.

This figure shows the power spectrum of the monthly equatorial eastern Pacific SST index from observations during 1900-1999 (black curve) and the T85 CCSM3 control run during model years 400-499 (red curve).

The spatial distribution of the tropical precipitation changes during December-February for a composite "warm-minus-cold" ENSO event is shown in the figure below, where "warm" and "cold" years are identified using a one standard deviation criterion of the equatorial eastern Pacific SST index. In the model, warm events are accompanied by rainfall increases in the central and western equatorial Pacific and decreases directly to the north and south; rainfall decreases are also seen in the tropical Atlantic and eastern Indian Oceans (a similar pattern and amplitude of precipitation anomalies is seen in the T42 version of CCSM3; not shown). The simulated rainfall changes show good overall agreement with the observed changes, although nature exhibits more pronounced rainfall decreases over the western Pacific and maritime continent.

This figure shows the warm-minus-cold ENSO composite of precipitation anomalies during December-February from (top) observations during 1979-2000 based on the Xie-Arkin data set and (bottom) the T85 CCSM3 control run.

The global atmospheric circulation responds to the redistribution of tropical precipitation during ENSO events via Rossby waves and their interaction with the midlatitude storm tracks. The global sea level pressure (SLP) response to ENSO during December-February is well simulated in the model compared to observations (see figure below). Below normal SLP is found in the eastern tropical Pacific, with above normal SLP in the remaining tropical areas. In the extra tropics, a deepening of the Aleutian Low over the North Pacific is well simulated in the model, with maximum values of approximately 8-10 hPa for the warm-minus-cold ENSO composite. SLP teleconnections to the Southern Hemisphere are also well simulated both in terms of location and amplitude. The T42 version of CCSM3 exhibits a similar SLP response to the T85 version over the tropics, but the amplitude of the Aleutian response is considerably weaker (~5 hPa) than the T85 version (not shown).

This figure shows the warm-minus-cold ENSO composite of sea level pressure anomalies during December-February from (top) observations during 1948-2003 based on the National Centers for Environmental Prediction (NCEP)-NCAR Reanalysis and (bottom) the T85 CCSM3 control run.

c) Simulating the Climate of the Late Permian

The boundary between the late Permian and the Early Triassic at 251 Ma is one of the most remarkable time periods in Earth's history. At this point in time, the largest known extinction of life took place, with approximately 95% of species extinguished. Geologic data indicate this was a time of global warming and very low oxygen levels in the oceans. Jeff Kiehl (CCR) and Christine Shields (CCR) have configured the CCSM3 to simulate the climate of this period. To study ocean conditions at depth, the fully coupled model has to be run for thousands of years for the system to come to a new equilibrium state. The Late Permian simulation is forced with a ten-fold increase in present-day CO₂ concentrations. The simulation has been run for over 2000 years, the longest integration of CCSM to date.

This figure shows the time series of the net energy flux at the top of the atmosphere (red) and surface (blue) for the Late Permian simulation. The initial forcing to the system was 14.5 Wm^-2 due to a 10x increase in atmospheric CO₂.

This figure shows the time series of the global mean annual mean energy flux (Wm^-2) into the oceans.

Given the large heat capacity of the ocean, the long time scale for the model to attain equilibrium is governed by the heat uptake of the oceans as shown in the figure above. Note that around year 1900, the energy flux into the system has reached a very small value (-0.04 Wm^-2). This flux of energy is proportional to time rate of change of the volume averaged ocean temperature, shown in the figure below.

This figure shows the time series of the volume averaged ocean temperature (°C) from the Late Permian simulation.

The near equilibrium solution indicates a dramatic decrease in the overturning circulation in the deep ocean. The present-day high latitude pathways of deep water formation, i.e., in the North Atlantic and off the coast of Antarctica, are shut off in the warm climate of the Late Permian. This is the first comprehensive coupled simulation of this time period, and it indicates the need for substantial computational resources to study scientific questions involving deep ocean circulations.

d) Coupled Carbon Cycle Results

A suite of multi-century fully coupled carbon-climate integrations have been performed using a modified version of the Climate System Model version 1.4 (CSM1.4). Several century-long spin-up runs were done to adjust the modeled carbon cycle incrementally to the coupled model climate. These spin-up runs led to a stable multi-century (1000 years) carbon-climate pre-industrial control run.

Additionally, multiple 180-year historical and 100-year future scenario fossil fuel emissions experiments have been completed. These experiments are being analyzed to understand connections and feedbacks between the carbon cycle and the physical climate.

This figure shows the CSM1.0 simulation of atmospheric carbon dioxide in peta grams of carbon from the pre-industrial to future climate using the SRES A2 scenario. This simulation includes both ocean and terrestrial carbon response to climate and changes in carbon dioxide.

Plans for the Use of CCSM3

The CCSM project will continue to follow a research and development program to enhance the capabilities of the modeling system. The research will be directed toward providing a more complete, more useful, and more accurate CCSM. Representations of new physical and chemical processes will be formulated and tested within the CCSM framework to make the modeling system closer to what is termed an earth system model. Development will take place to provide increased resolution of climate change projections for impact assessment. As well, increased emphasis will be placed on improving the simulations of the hydrological cycle on continents. Software development will be directed to make the modeling system more portable, more reliable, and easier to use. Considerable effort will be put toward making the modeling system more accurate through a concerted attack on the long standing biases and continued examination of the fidelity of the representations of the physical and dynamical processes of the basic components of the model, atmosphere, ocean, ice, and land.

The recently released version of the model, CCSM3, will be used to tackle a large number of interesting scientific questions. One of the most interesting questions has to do with the physical processes that can lead to abrupt climate change (ACC). In general, climate change is viewed as a gradual process taking place over centuries or millennia. However, observations of past climates indicate that large changes can occur over a few decades. There are hypotheses about physical processes involving ice, air, and ocean interaction particularly in the North Atlantic that, when perturbed, appear to have the potential to lead to rapid changes as a new equilibrium is sought by the climate system. CCSM3 will be used to investigate these physical mechanisms from a basic science point of view. The ad hoc CCSM group for ACC will also attempt to reproduce the characteristics of periods of ACC that have been observed in the paleoclimate record.

CCSM3 will be used to address a rather important question about the effects of human-induced land-use change on climate. Most people think of the emission of greenhouse gases as the substance of human effects on climate. However, in the past 150 years human activity has transformed the land surface. For example, farms have replaced forests in most of the eastern half of the U.S. and southern Canada. There is instrumental evidence of the changes in local climates produced by changes in land use. It is important to address the question of the impact on the climate on at least the continental scale by land-use changes in the 19th and 20th centuries.