![[logo] US EPA](http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
Regional Climate Model (RCM) Simulations and Analyses
Research Programs
For the CIRAQ project, DOE's Pacific Northwest National Laboratory (PNNL) has conducted Regional Climate Model (RCM) simulations using MM5 with initial and boundary conditions from the NASA GISS Global Climate Model (GCM). Simulations were completed for a reference period (e.g., 2000 ± 5 years) and a future period under climate change conditions (e.g., 2050 ± 5 years). The future GCM simulation used the Intergovernmental Program on Climate Change (IPCC) A1B greenhouse gas emission scenario. These regional climate model (RCM) simulations cover the continental U.S. domain with a 36 km × 36 km grid resolution.
Data Management and Quality Control
The RCM results from PNNL are processed into “model-ready” (i.e., MCIP) meteorology input files for the current and future CMAQ simulations. The extremely large size of the RCM datasets requires the development of automated data management and quality control tools. The raw downscaled RCM data are first processed for use by SMOKE and CMAQ using the MCIP processor. Next, the hourly gridded outputs from MCIP, and changes in these outputs with respect to time, are examined against set tolerance and temporal change limits. Values that fall outside these limits and the locations where these values occurred are stored in a database along with the hourly grid statistics. Additionally, a back-check of the meteorological data from MCIP is compared with the raw RCM output for consistency. Statistics are generated from this back-check comparison and archived in the CIRAQ database.
Climate Impact
Regional Climate ScenariosClimate Variability & Change
Emission Scenarious
Phase I (climate response only)
Air Quality Scenarios
Regional Scale Air Quality Modeling Scenarios
Evaluation of Regional Climate Model (RCM) performance and awareness of the model uncertainties is essential. Moreover, when a RCM is utilized for a future climate scenario there are no observations for comparison, so it is necessary to consider the performance of the modeling system for a reference case before results of future periods are interpreted. RCM model performance for the baseline scenario period (1995-2004) will be explored from three statistical perspectives: (a) comparison of observational and analyzed meteorological field cumulative distribution functions, (b) cluster analysis, to identify dominant and distinct or extreme patterns of synoptic-scale chemical transport, and (c) time series analyses, to identify synoptic, seasonal, and interannual time scale components. Each of these approaches are discussed in more detail below. These analyses provide information about how well the model performs in the current simulation and characterize the spatial and temporal variations in meteorological parameters that could substantially influence the forthcoming CMAQ air quality predictions.
A summary report on evaluation, spatial, and temporal analyses of downscaled regional meteorological simulations and completion of "model-ready" MCIP-processed MM5 regional climate simulations for reference period and future climate change scenario is scheduled for completion during Sept. 2005.
(a) Comparison of observational data and analyzed meteorological fields via CDFs
Model and observed distributions of hourly meteorological variables (e.g., temperature, wind, moisture) will be compared by comparing their cumulative distribution functions (CDFs). Variations in climate over a region (e.g., mountains, Piedmont, coastal plain of North Carolina) are evident in the distribution of meteorological variables such as temperature. If the RCM is indeed simulating the regional climate properly, the distributions should be similar. The CDF will be examined for a variety of climate regimes across the United States for annual and seasonal variability. Other evaluations will be performed using solar radiation measurements to determine how well the RCM is simulating the cloud coverage that impacts photo-chemistry in the air quality model. Monthly, seasonal and annual precipitation amounts will be inspected using a gridded national precipitation analysis and rain gauge observations. [primary contact: Robert Gilliam]
(b) Spatial Characterization via Cluster Analysis
This analysis evaluates RCM scenarios of synoptic scale meteorological development and behavior important to atmospheric chemical transport and reactions. Published literature suggests that patterns of 700mb u and v wind component data are good indicators of these conditions. Multivariate cluster analysis techniques described in Eder et al. (1994) and Cohn et al. (2001) have been adapted to identify and describe dominant as well as unusual modes of 700mb transport by season. These techniques were developed and verified using two different global reanalysis datasets (Kalnay, et al., 1996; Kanamitsu, et al., 2002) and, where appropriate, results were compared to published results in Eder et al. (1994) and Cohn et al. (2001). Cluster results from these two alternative observation-based datasets are compared to similarly-derived present and future RCM cluster analysis results. Specific questions to be answered include:
- Are dominant patterns of RCM transport similar in appearance across RCM and reanalysis datasets?
- Do dominant transport patterns occur with the same frequency across RCM and reanalysis datasets?
- Are there the same number of distinct or unique transport patterns in RCM and reanalysis datasets?
- Do theses unique patterns appear to describe the same kind of synoptic events (e.g., stagnation, flood, drought, etc.)? Do the unique patterns occur with similar relative frequency between RCM and reanalysis datasets?
(c) Temporal Characterization via Time Series Analysis
Meteorological conditions such as temperature and solar radiation have clear diurnal and seasonal patterns of variation, and they are driving factors for air pollution chemistry. In addition, meteorology can vary from year to year based on random differences or periodic phenomena such as the El Nino-Southern Oscillation. Time series analyses of the RCM simulation results will be used to characterize the diurnal, synoptic, seasonal, and interannual components for specific air quality-related meteorological conditions. The current and future simulations can then also be compared to determine if climate change impacts on the meteorology can be detected and to investigate differences between the current and future meteorology on the various time scales. The time series analyses for the RCM simulations, as well as the forthcoming CMAQ simulations, will be performed using a linear filter technique referred to as the Kolmogorov-Zurbenko or KZ filter (Rao et al., 1997; Eskridge et al., 1997; Hogrefe et al., 2003). Results from this analysis provide insight into the meteorological factors that influence air quality prior to the CMAQ simulations. Once CMAQ simulations are complete, time series analyses of the ozone and PM predictions will be performed and considered in light of the RCM analyses described above.
References
Cohn, R.D., B.K. Eder, S.K. LeDuc, and R.L. Dennis. A Development of an aggregation and episode selection scheme to support the Models-3 Community Multiscale Air Quality Model. Journal of Applied Meteorology, 40: 210-228 (2001).
Eder, B.K., J.M. Davis, and P. Bloomfield. An automated classification scheme designed to better elucidate the dependence of ozone on meteorology. Journal of Applied Meteorology 33: 1182-1199 (1994).
Eskridge, R.E., J. Yeong, S.T. Rao, P. S. Porter, and I.G. Zurbenko. “Separating different scales of motion in time series of meteorological variables.” Bulletin of American Meteorological Society 78: 1473-1483 (1997).
Hogrefe C., S. Vempaty, S.T. Rao, and P.S. Porter, P.S. A comparison of four techniques for separating different time scales in atmospheric variables. Atmospheric Environment 37: 313-325 (2003).
Kalnay, E., and Coauthors. “The NCEP/NCAR 40-Year Reanalysis Project,” Bulletin of American Meteorological Society 77: 437-471 (1996).
Kanamitsu, M., W. Ebisuzaki, J. Woollen, S-K Yang, J.J. Hnilo, M. Fiorino, and G.L. Potter. “NCEP-DOE AMIP-II Reanalyiss (R-2).” Bulletin of American Meteorological Society 83: 1631-1643 (2002).
Rao, S.T., I.G. Zurbenko, R. Neagu, P.S. Porter, J.Y. Ku, and R.F. Henry. “ Space and Time Scales in Ambient Ozone Data.” Bulletin of American Meteorological. Society 78: 2153-2166 (1997).