Community Atmosphere Model and Community Climate System Model Development.

Members of the Climate Modeling Section (CMS) (Byron Boville, William Collins, James Hack, Jeffrey Kiehl, Philip Rasch, and David Williamson), in collaboration with colleagues in the university and national laboratory community, contributed broadly to the development of the newly released Community Atmosphere Model (CAM), previously known as the Community Climate Model (CCM). There are a number of significant improvements and enhancements included in the physical parameterizations, dynamical approximations, and software engineering implementation of the model. The improvements to the physics include: a prognostic cloud water scheme; a generalized treatment of cloud overlap for radiation transfer calculations; a more accurate formulation of water-vapor radiative effects; a detailed thermodynamic sea ice model; and a fractional specification of ice and land consistent with the Community Climate System Model 2 (CCSM2). The improvements to the dynamics include the addition of the NASA Data Assimilation Office (DAO) finite volume dynamical core, in collaboration with Shian-Jiann Lin (NASA Goddard Space Flight Center (GSFC)), and a more flexible implementation for coupling the physics and dynamics components. The datasets used to force the atmospheric model include a new, state-of-the-art sea surface temperature (SST) and sea-ice dataset spanning 1949-2001. The CCSM Climate Variability Working Group is planning to use this dataset to create a large ensemble of CAM integrations for the last fifty years. These integrations will be used to study various modes of variability in the climate system, e.g., the North Atlantic Oscillation (NAO) and to quantify the fidelity of the simulated climate. The model source code, documentation, initial and boundary datasets, and control integrations complete with diagnostic analysis have been released to the climate community via the web. The model is capable of running on a variety of different computer architectures popular for climate research applications. Examples of major improvements in the climate simulated by CAM relative to CCM3 include: more realistic distributions of precipitable water in the tropics; much more realistic clear-sky longwave fluxes in polar regions; and an improved surface energy budget in the eastern equatorial oceans.

Hack and Rasch collaborated with Minghua Zhang (State University of New York, Stony Brook) and Chris Bretherton (University of Washington) on modifications to the CAM cloud water parameterization. Rasch and Boville also worked on modifications to the CAM moist physics and radiation parameterizations to address temperature, cloud, and water vapor biases in the polar and tropical upper troposphere. In collaboration with John Bergmann (NOAA), Rasch explored the role of cloud overlap in the representation of CCM3 cloud-radiation interactions.

Hack and John Truesdale (CMS) completed development of a simple hydrometeor evaporation scheme for deep convection in CAM2. The incorporation of Collins' improved treatment of water vapor absorption in the longwave parameterization introduced severe systematic biases in the global simulation for the prototype CAM. Through a study of the total diabatic heating budget, using the Single-column Community Climate Model (SCCM) simulations to identify physics interactions, Hack and Truesdale devised a modification to the deep convection parameterization that significantly improved the simulated hydrological cycle.

Hack and Truesdale introduced a fractional land and fractional sea ice formulation into CAM2. This capability allows for a more realistic specification of the underlying surface. The resulting simulation changes represent improvements, especially in regions like the maritime continent and Carribean. There is also some suggestion that there may be physically and statistically significant improvements to the storm track regions. A more rigorous evaluation of the simulation response is underway.

Hack and Truesdale, in collaboration with Cecilia Bitz (University of Washington), incorporated and tested a version of the CCSM Sea Ice Model (CSIM) ice thermodynamics into CAM2. This change clearly improves the high-latitude simulation in the uncoupled model and makes the simulation properties much closer to those of the fully coupled model.

Rasch moved the aerosol parameterization developed for the Model for Atmospheric Transport and Chemistry (MATCH) into CAM. He worked closely with the software engineering group on issues associated with decomposition strategies for parameterizations in connection with his work with Warren Washington (Climate Change Research Section, CCR) on a sulfate parameterization for use in their climate change calculations. Rasch also modified numerous components of CAM associated with transport of tracers and aerosols.

Hack, James Hurrell (Climate Analysis Section (CAS)), James Rosinski (CMS), and Julie Caron (CMS) developed a new boundary dataset for forcing the uncoupled atmospheric model. This dataset, a blended Hadley Ice and SST (HadISST)/Reynolds product, presently covers the period from 1949 to 2001, and can be extended as new Reynolds data becomes available. A major portion of this effort dealt with the specification of sea ice concentration, which required the development of several new quality control procedures to deal with problems in the raw datasets.

Hack, Bruce Briegleb (Oceanography Section), Rosinski, and Bitz have collaborated on the development of a Slab Ocean Model (SOM) capability in CAM2. This model extension is based on the slab ocean and thermodynamic ice formulations employed by the CCSM Polar Climate Working Group. This modeling capability will be an essential tool in quantifying the climate sensitivity of the uncoupled model, and identifying the underlying feedbacks associated with the physical approximations.

Williamson and Jerry Olson (CMS) continue to work on developing the capability to apply the CAM2 in forecast mode. The objective of this work is to better understand parameterization methods by directly comparing parameterized variables (e.g., clouds, precipitation) with observations early in the forecast while the forecast state is still near that of the atmosphere, but after initial transient computational modes are damped. They are working to develop methods that involve less effort than what would be required to develop a complete forecast-analysis system. The approach runs CAM2 in a forecast mode using an ensemble of historical analyses from very recent years. The main effort is then to map the high-resolution analyses to the coarse climate model grid. This is reasonably straightforward for atmosphere state variables, but more involved for parameterization variables, which carry time history. They have developed two procedures to spin-up land and atmospheric parameterized variables referred to as "forecast-analysis cycle" and "nudging." The forecast-analysis cycle involves updating atmospheric state with analyses periodically (e.g., 6 hourly) and letting the land model respond to the forcing from the atmosphere. The nudging involves adding terms to the model to relax predicted state variables toward analyses, and again letting the land model respond. Experiments have been performed with a "perfect model" assumption indicating that these approaches hold promise. They have modified the atmosphere, land, and sea-ice initial datasets so they contain all data needed to initialize the models. Previously these datasets contained only relatively slowly varying variables needed for climate applications, but now include the faster time scale variables needed for forecasting. Experiments are now underway with European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis data.

Hack and Caron have lead an effort to develop configurations of CAM2 that run at resolutions other than T42. The two principal examples are the T31 configuration, which was developed in collaboration with Christine Shields (CCR) for paleoclimate applications, and a T85 configuration, which is being explored for climate change and assessment applications. The high-resolution configuration, which is based on the Eulerian spectral dynamical core, employs a T85L26 spectral truncation, twice the horizontal resolution used for the standard atmospheric configuration. The bulk of the high-resolution development activity has been aimed at modifying free parameters in the cloud fraction and cloud microphysics schemes to achieve a global radiation budget that is in agreement with observations. This work has revealed limitations in CAM2 cloud fraction parameterization, which demonstrates a strong sensitivity to the horizontal resolution. Although the T85L26 configuration exhibits a reasonable Top of Atmosphere (TOA) energy budget that is balanced in the global annual mean, there are regional deficiencies that will need to be addressed by reformulating specific components of the cloud scheme. Preliminary work at T170L26 has indicated that it will not be possible to assemble a realistic configuration at this very high resolution without fundamental changes to the formulation of the cloud scheme. Work on reformulation of the cloud parameterization is currently underway. Hack has also continued work with external collaborators (Patrick Duffy Lawrence Livermore National Laboratory) to explore ultra high-resolution climate configurations (i.e., greater than T213 truncations) versions of the CCM3.

Hack and Truesdale completed the development of a beta version of the new CAM2 Single Column Model (SCAM). This modeling tool has become a central means for developing new parameterization techniques for the atmospheric model. Members of the CCSM Atmosphere Model Working Group (AMWG) intend to use the SCAM to better understand the current physical CAM2 parameterization suite.

Climate Research

Kiehl, in collaboration with Eric Maloney (CMS), carried out simulations with a version of the CCM3 coupled to a slab ocean to investigate the role of ocean coupling on eastern Pacific intraseasonal variability. This study indicates that interaction between the atmosphere and ocean plays an important role in determining the phase of intraseasonal variability in this region of the tropics. Kiehl, in collaboration with V. Ramanathan (University of California, Santa Barbara/Scripps Institution of Oceanography) and Chul Chung (University of California, Santa Barbara/Scripps Institution of Oceanography), used the CCM3 to investigate the importance of absorbing aerosols in the Indian subcontinent climate. These simulations show that the presence of absorbing aerosols lead to a cooling of the surface and a shift in precipitation patterns over this region. A cooling in this region has actually been observed and may be linked to the absorbing aerosols.

Kiehl, in collaboration with Ramanathan and Washington, has carried out simulations with the Parallel Climate Model (PCM) to consider the effects of Asian South East Asian haze on the coupled climate system. These simulations indicate a significant climate impact of haze in this region. The conclusion of these studies indicates that any future studies of climate change will need to include absorbing aerosols.

Williamson and Olson have conducted aqua-planet simulations with Eulerian and semi-Lagrangian dynamical cores coupled to the NCAR CCM3 parameterization to better understand the underlying reasons for the formation of the double Intertropical Convergence Zone (ITCZ). These experiments show that the two dynamical cores produce results with very different zonal average precipitation patterns, one with a narrow single precipitation peak centered on the SST maximum, and the other with a broad double-peaked structure straddling the SST maximum. Their work has indicated that the different structures are caused primarily by the effect of the different time stepping procedures adopted by each dynamical core, rather than by different truncation errors introduced by the dynamical approximations. Different diffusive smoothing associated with different spectral resolutions is a secondary effect influencing the strength of the double precipitation structure. When the semi-Lagrangian core is configured to match the Eulerian with the same time stepping approximation, a three-time-level formulation, and same spectral truncation, it produces precipitation fields similar to those produced by the Eulerian configuration. They argue that the physical reasons behind the double structure are related to the longer time step, where more water vapor is evaporated into the lower troposphere over a longer discrete time step, the evaporation rate being the same. The additional water vapor in the region of equatorial moisture convergence results in higher amounts of the Convective Available Potential Energy (CAPE) farther from the equator, which results in an earlier initiation of convection. The resulting heating drives upward vertical motion and low level convergence away from the equator, resulting in much weaker upward motion at the equator. The feedback between the convective heating and dynamics reduces the instability at the equator and decreases the convection there. Williamson and Olson are designing additional experiments to further explain this behavior. Williamson is also organizing an international aqua-planet intercomparison activity with members of the University of Reading and members of the Program for Climate Model Diagnosis and Intercomparison (PCMDI). The intercomparison is sponsored by the Working Group on Numerical Experimentation (WGNE) and will hopefully help further identify the physical mechanisms responsible for large-scale climate simulation features like the double ITCZ.

Climate and Chemistry

Collins and Rasch have continued their research into the global climate effects of aerosols. This year they created the first ever global climatology of major aerosol species based upon assimilation of aerosol data. The climatology was created using MATCH coupled with a chemical assimilation package developed by scientists in the NCAR Atmospheric Chemistry Division (ACD). Rasch worked on the development of aerosol parameterizations for MATCH in collaboration with Collins, Charles Zender (University of California, Irvine), Natalie Mahowald (University of California, Santa Barbara UCSB), Chao Luo (UCSB), and Lynn Russell (Princeton University). Modifications to the modeling capabilities included: revisions to the wet deposition of aerosols, revisions to the emission inventories for aerosols, and the development of a tagged aerosol to characterize the length of time since emission from given regions to help understand measurements made during the Aerosol Characterization Experiments (ACE)-Asia field program.

The global climatology was generated by forcing MATCH with the NCAR/National Centers for Environmental Prediction (NCEP) reanalysis meteorology for the period 1980 through 2001. In order to constrain the model simulation, Collins and Rasch assimilated daily satellite retrievals of aerosol optical depth over the same time period. This model data is now being used in the operational Surface and Atmosphere Radiation Budget (SARB) subproject in the NASA Clouds and the Earth's Radiant Energy System (CERES) program, where the data has been shared extensively with colleagues at other NASA branches. Collins and Kiehl have developed methods for reading this climatology into CAM to compute aerosol radiative forcing and to study the effects of aerosols on the simulated climate. The initial calculations using this new technique are underway and will be reported at the Fall 2002 meeting of the American Geophysical Society. The longer-term objective is to include this climatology as a standard dataset in future releases of CAM to the climate community.

The 2002 summer colloquium for the NCAR Advanced Study Program (ASP) addressed "Interactions among Aerosols, Climate, and the Hydrological Cycle." The colloquium was organized by Collins, Yoram Kaufman (NASA/GSFC), and Ramanathan. Thirty-three students pursuing doctoral or postdoctoral research on aerosols attended, eight of which were from overseas locations. The sixteen lecturers included experts on aerosol chemistry, global and chemical modeling, measurement and characterization, and remote sensing. The speakers were invited from leading institutions in the U.S., Canada, U.K., and France. This year, for the first time the complete audio-visual record of the talks is available for the participants and other students to download from the ASP website. The students also worked on group projects to quantify some of the climatic effects of aerosols. Using aerosol properties derived from new NASA satellite data, each group studied the aerosol radiative forcing using the NCAR SCM developed by Hack and Truesdale. The students presented their preliminary results on the last day of the colloquium. The success of the colloquium is due to the efforts of many people in ASP, CGD, the NCAR Scientific Computing Division (SCD), and the NCAR Mesoscale and Microscale Meteorology Division (MMM), as well as support from the NCAR Director's Office and from the University of Colorado at Boulder.

WACCM Development

Boville and Sassi continued their collaboration with Rolando Garcia (ACD), Douglas Kinnison (ACD), and Raymond Roble (High Altitude Observatory, HAO) on the development of a Whole Atmosphere CCM (WACCM). WACCM is motivated by the scientific appreciation of the importance of coupling between vertically adjacent atmospheric regions. Vertical propagation of atmospheric waves and the transport of minor species from the troposphere are known to play important roles in the dynamics and chemistry of the middle and upper atmosphere. There is also increasing awareness that changes in the propagation characteristics of planetary waves in the stratosphere, due to natural or anthropogenic factors, may play a role in tropospheric climate variability by influencing such phenomena as the Arctic Oscillation. The longer term WACCM objective will be to represent the atmosphere from the surface to ~500 km, including fully interactive chemistry.