|
|
 |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
ERL/ETL MM5 The workshop concluded that eventually the detailed chemistry of atmospheric aerosols and gases (anthropogenic and otherwise) should be incorporated into prediction models, but it was also agreed that it is premature to begin this work in earnest because many of the important chemical transformation mechanisms are not yet known well enough. This is especially the case for semi-volatile chemicals. Moreover, deficiencies in current mesoscale models preclude accurate modeling of the chemistry. For example, many important chemical reactions are known to involve cloud water, yet cloud liquid and cloud ice are only recently being included in the more advanced mesoscale models. It was also concluded that an initial focus should be on improving treatment of clouds and surface boundary conditions. Full coupling of atmospheric and chemistry models must await further maturation of both. Furthermore, it was also concluded that a possible early focus for a cooperative effort between atmospheric modelers and chemists is the problem of visibility. Many agencies are interested in this problem: the Environmental Protection Agency, the National Park Service, the Federal Aviation Adminstration (from an aviation standpoint), and the National Weather Service. It was pointed out that much is known about the major contributors to visibility reduction, but the group agreed that a concerted attack even on this problem would not be easy. Several scientists expressed enthusiasm for ensemble prediction and applying this tool for air quality. Regular meetings are needed between the NOAA operational and research modeling groups. The research groups need to become more familiar with the operational models (Eta or RSM). Future research developments must consider the needs of the operational models. Specialized subgroups concluded that:
-- improved moisture modeling in the surface (e.g.: soil moisture, runoff) and surface boundary layer, -- the modeling and/or data assimilation of non-precipitating clouds, -- aggregating fine-scale surface features (physical and vegetative), -- the benefits of hydrostatic vs. non-hydrostatic modeling, -- the need to match process parameterization and model scale resolution -- the benefits of data assimilation vs. dynamic initialization of meteorological fields and, -- improved techniques and data sets to evaluate fine resolution models Finally, it was recommended that a NOAA mesoscale
modeling World Wide Web page should be set up, to help
keep NOAA scientists up to date on latest developments
and findings. 1. PURPOSE OF THE WORKING GROUPA preliminary meeting, held at the NOAA Air Resources Laboratory (ARL) Atmospheric Sciences Modeling Division (ASMD) in January 1996, focused on a plan to improve cooperation between NOAA groups working on environmental modeling. The discussions also addressed methods to improve the interactions between the NOAA modeling research communities and the NOAA National Centers for Environmental Prediction (NCEP) operational environmental predictions. The results of these discussions have led to a formation of a NOAA Mesoscale ModelerÃs Working Group Meeting in May 1996, to help foster collaborations between NOAA scientists working on these issues. Since the meeting was held at ASMD, many of the discussions addressed applying mesoscale modeling techniques to environmental assessment and prediction (in this case air pollution). For air quality prediction, three modeling components should be addressed:
Research and operational modeling within NOAA has traditionally emphasized item two, the meteorological modeling system. However, an integral component of NOAA's strategic mission is to improve environmental prediction that would incorporate all three items. Further, many NOAA modeling groups emphasize either research aspects of meteorological or air quality modeling, or operational forecasting. It was proposed that representatives from NOAA's modeling communities meet to discuss current developments and future collaborations. The first meeting was held atASMD for an overview of their state-of-the-art air quality modeling program. Twenty-two modelers from nine NOAA labs participated in the meeting and have formed an ad-hoc Mesoscale Modeler's working group. This document summarizes the current activities of NOAA's modeling community and attempts to make recommendations for future research to meet NOAA's missions. A specific goal of this group might be to identify (a) areas of overlap, (b) products that might be shared rather than reinvented, and opportunities for joint development in the future, provided the interest exists. The goal is not to generate new programs, nor to create forced marriages. This enhanced cooperation is especially needed now that the questions being asked are becoming more complicated while the people who are working on the problems are becoming fewer as funding is reduced. For this cooperation to work, the practicing modelers should get to know each othersà views, capabilities, and intentions and it is hoped that this working group will be a vehicle for future interactions. To reach these goals, the following suggestions were made:
This approach may be an effective means for the NOAA modeling
communiy to learn from each others activities while continuing
to pursue our individual research goals. The NOAA modeling
community has become increasingly aware that we must join
forces with colleagues so that we can draw on their experience
and detour around problems that have already been identified.
For this to work, the group will be attempting to generate
a good, healthy rapport at the working level, so that the
interactions do result in accelerated developments of the
improved models that we all want. That is why the present
focus is on the practicing scientist level. 2. CURRENT ACTIVITIESCurrently, several groups in NOAA are using mesoscale Eulerian meteorological models at hydrostatic and non-hydrostatic scales. These models are applied for operational forecasting, meteorological research and to drive pollutant transport and dispersion models for assessments against historical conditions and for forecasting. The table below summarizes the works of these groups. During day one of the meeting, the participating scientists reported on current activities and provided brief written summaries which are included below. Some of the mesoscale modeling issues commonly being addressed were:
Once these works were summarized, the working group attempted
to identify the most critical research issues where interactions
could be fruitful. A summary of these recommendations are
given in sections 3 and 5. More detailed summaries of the
mesoscale modeling work being done at each laboratory was
provided by the working group scientists in Appendix
A.
ERL/ETLMM5 Coastal Oceans81-15 kmatmosphere-ocean dynamical
interactionsERL/ GLERL RAMS-CHARM 3. DISCUSSION OF CRITICAL RESEARCH ISSUES AND RECOMMENDATIONS3.1 Summary of Group Discussion on Data Assimilation(Facilitator and author: Tom Schlatter, NOAA Forecast Systems Lab) As a stimulus for discussion, the following proposition was advanced: The most basic long-term goal of a cooperative effort between atmospheric modelers and atmospheric chemists (more generally, those who deal with air quality) should be to merge sophisticated mesoscale models with atmospheric chemistry models for aerosols and gases. If this proposition is accepted, the following questions arise:
The group seemed willing to accept the basic proposition that eventually the detailed chemistry of atmospheric aerosols and gases (anthropogenic and otherwise) should be incorporated into prediction models, but we agreed that it was premature to begin this work in earnest because:
These process are still poorly modeled, almost exclusively through parameterization because of the small scales at which mass exchange occurs. The bases, tops, and areal coverage of clouds are still poorly predicted and not even well analyzed. Finally, surface conditions, especially soil moisture, must be more accurately tracked so that the weather in the surface boundary layer, where many pollutants reside and have their source, may be more accurately predicted. Biogenic emissions depend strongly upon soil moisture. For now, mesoscale modelers should work to improve their treatment of clouds and surface boundary conditions. For now, chemists are content to use highly detailed analyses from data assimilation systems as input to their own models. In effect, they are assuming that models are "meteorological preprocessors," describing the "true" atmosphere at the temporal frequency and spatial resolution of the analyses, and they ask how that atmosphere will affect the chemical constituents that they insert. Though we know much about emissions on an annual and, in some cases, a monthly basis, information on daily emissions is still sparse. Full coupling of atmospheric and chemistry models must await further maturation of both. From a regulatory standpoint, analyses of atmospheric observations are gaining credibility, whereas numerical predictions have virtually no legal standing. Whether atmospheric models applied to air quality problems should be hydrostatic or nonhydrostatic was debated but without resolution of the issue. Some argued that the very conditions that lead to pollution episodes are those most likely to be hydrostatic: quiescent atmospheric conditions and weak flow regimes. Others argued that a sudden release of toxic substances would not necessarily occur in quiescent conditions and that a realistic prediction of the transport and dispersion of these substances might depend strongly on nonhydrstatic processes such as tropopause folding, penetrative convection, and mountain waves. A possible early focus for a cooperative effort between
atmospheric modelers and chemists is the problem of visibility.
Many agencies are interested in this problem: the Environmental
Protection Agency, the National
Park Service, the Federal
Aviation Adminstration (from an aviation standpoint),
and the National Weather Service. It was pointed out that
we know a lot about the major contributors to visibility
reduction: sulfur compounds arising from combustion, mostly
from fixed point sources; nitrogen compounds arising from
combustion, especially motorized vehicles; ammonia from
feedlots, swamps, and vegetative decay (poorly measured);
and certain organic compounds of biogenic origin. The group
agreed, however, that a concerted attack even on this problem
would not be easy. 3.2 Short-Range Ensemble Forecasting(Facilitator and author: David Stensrud, National Severe Storms Laboratory, Norman, OK) Main recommendation: NOAA needs to put more effort into investigating ensemble approaches to short-range forecasting Summary: Twelve members of the workshop met to discuss the concept and possible uses of short-range ensemble forecasting (SREF). The meeting was opened by the facilitator noting that there are two extremes to weather prediction in ten years: one very high-resolution (3 km) numerical forecast over North America, or one thousand mesoscale (20 km) forecasts over North America. The question yet to be answered is which extreme, or blend of these two extremes, provides the most valuable guidance to the public. It is important to note that forecasts have no value until they are used. Therefore, while many end users are comfortable with deterministic forecasts, they could learn to use probabilistic forecasts to their economic advantage through cost-benefit analyses. And forecasters can learn to issue probabilistic forecasts with the proper training. Much discussion centered around the pilot program on SREF being conducted by the Environmental Modeling Center (EMC) at the National Centers for Environmental Prediction. While the ensemble approach has been used mainly with medium-range forecasts, the large uncertainties in the initial conditions at smaller-scales makes the ensemble approach a valid one for short-range forecasts as well. The eta ensembles in this pilot program are produced using six different operational analyses and two pairs from the breeding of growing modes technique used in medium range forecasts. Similarly, the five ensemble members from the regional spectral model are from the control run plus the same two pairs from the breeding of growing modes technique. Even though having only fifteen ensemble members is small and undoubtably does not sample well the solution space, the results from the precipitation analyses are impressive and indicate that this area of research needs to be explored as a viable option for making numerical forecasts in the future. While the initial work on SREF has dealt exclusively with the meteorological implications, the potential value to air quality research is large. Not only is there uncertainty in the meteorological variables, but there is an even larger source of uncertainty in the emission data used in studies of air quality. When analyzed meteorological fields are used to drive air quality models, small errors in the windfields can accumulate along derived trajectories of air parcels. The resulting trajectory displacements can mean either assimilating or not assimilating a particular emissions source(s) along the trajectory. Thus air quality predictions can be quite sensitive to apparently small errors or uncertainties in the meteorological analysis. While clients typically prefer a single answer or number, it is likely that the use of probability information could become accepted and prove to be more useful and informative than single values, with undefined uncertainties, when dealing with air quality and hazardous emissions research and forecasts. Not only is SREF of potential value owing to providing information on the variety of possible solutions, but there is a direct economic benefit to NOAA as well. As massively parallel processors (MPPs), and code to run on them, become more and more commonplace, an advantage arises to using ensemble forecasts, since it takes less memory to run one thousand mesoscale model forecasts than it takes to run one high-resolution simulation. This reduction in memory requirements implies a reduction in the cost of purchasing the MPPs used to make the forecasts. The discussion concluded by highlighting the many areas
of research that need to be addressed in order to use ensemble
techniques to their best advantage. The creation of ensemble
members for mesoscale modeling is an area in which little
is known. How do you create the mesoscale ensemble members
when convection is such an important process and feedback
mechanism? Can you use different model physical parameterization
schemes to help generate the ensemble members? How do we
realistically add information to a model initial condition
at spatial scales smaller than those of the observational
network? Even assuming we can create an ensemble that provides
a complete picture of the possible solutions available,
then one still must find methods for using all this information
to our benefit. What are the best ways for depicting ensemble
information? And once we determine good ways to interrogating
the ensemble data, we still must then educate our customers
on the uses of the ensemble information. It is clear that
we have much to do in addressing the questions of SREF
if NOAA is to use this
approach for weather prediction in the next decade. 3.3 Air Quality Prediction(Facilitator and author: Jason Ching, ARL/ASMD, Reserach Triangle Park, NC) Overview: The discussion below represents an attempt to synthesize the relatively limited notes, and even more limited recollections of the working group discussion dealing with the air quality predictions. Also attempted was a discussion of some major issues and problems, as well as a listing of some current capabilities, and suggested opportunities for cooperation among the NOAA components with the goal of achieving environmental forecast tools on various time frames. The various processes which need to be modeled for air quality prediction are shown in Figure 1. The following thoughts are meant to be a stimulus for further discussions at and prior to our next meeting(s). Figure 1. A representation of the conplex interactions involved with air quality prediction, including the emissions, meteorological, chemical and air-surface exchange processes. 1. General considerations: With the ever increasing industrialization of society, the recognition of the value and importance of clean air, and for cost-effective management of air quality, there is a great need to develop accurate air quality prediction tools. Air quality predictions are considered to be either (a) retrospective or (b) forecast. Retrospective (sometimes also known as prospective) models are used to determine the changes to air quality for different emission control strategies for a set of adverse meteorological conditions in which pollution levels have been historically observed at high levels. Such weather conditions may be episodic or diurnal, or seasonal. Due to the transport processes, the modeling scale is at least regional. Retrospective modeling provides guidance toward achieving National Ambient Air Quality Standards (NAAQS) for a variety of pollutants such as ozone, CO, SO2, NO2, and particulate matter. NAAQS primary standards are designed to protect human health, while the secondary standards are targeted primarily to protect the environment. In contrast, air quality prediction in the more traditional forecast sense simulates future air quality changes and distributions for which emissions source strengths and their distributions are known. In this mode, forecasts of air quality would provide advanced warning of pollution conditions that would pose exposure risks to the public. Forecasts of pollutants such as ozone and fine particles can be prepared and utilized in a manner similar to the situation with forecasting of UV-B radiation. Armed with such information, the public can take the necessary precautions or modify their activities to reduce their exposure and risk. Potentially, such information could also be used to provide guidance on adjusting emissions source strength for forecasted adverse air quality situations. Both the retrospective and the forecast type models have numerous similarities and requirements. The primary one is the need for accurate meteorological predictions or forecasts and for gridded day-hour specific emissions inventories. For both types of models, pollutant concentration fields are predicted. In addition, air quality related values such as dry deposition, visibility, and precipitation scavenging leading to wet deposition (acid rain) are of primary interest. 2. Specific Considerations Air quality modeling systems require emission inventories of anthropogenic pollutant, detailed (modeled) inputs on biogenic sources, and of course the basic meteorological forecasts. With these boundary and initial conditions, and other types of data, these model systems simulate with state-of-science parameterizations, all the important transport, transformation and loss processes acting upon these airborne pollutants. Important linkages between emissions (both anthropogenic and biogenic) and meteorology including processes governing plume dispersion, dry deposition of pollutants to the surface and emission of organic pollutants from vegetation, and soil NOx emissions are reorganized and performed. Episodes to be simulated are at least multi-day in duration and regional in scope. With the exception of precipitating systems yielding wet deposition patterns, the critical human exposure situations are typically correlated with conditions of weak to stagnant synoptic flows, where temperatures and humidity values are high and for which the presence and role of fair weather clouds are significant. Thus, it is important that features associated with the weakly forced synoptic conditions need to be accurately forecasted or characterized. Great improvements in meteorological models have been achieved recently due to (1) improved parameterizations arising from better understanding of atmospheric processes, (2) improved computational methods from advances in computational sciences, (3) incorporation of techniques that assimilate observed meteorological fields using either dynamic initialization for forecast models, or four dimensional data assimilation using Newtonian nudging techniques in the retrospective mode, and with (4) advances in the power and performance of computational systems. Thus, we now enjoy improved accuracy in the quantitative predictions of precipitating systems, in increased confidence and accuracy in extended forecasts. Experience with the Regional Acid Regional Model (RADM) demonstrated that the conversion of precursors to acidic species, the formation of secondary particulate in the fine size mode and some of the photo oxidants such as peroxides and radical species are dominated by aqueous phase chemistry. Thus, accurate predictions of non-precipitating cloud fields (coverage, type, spatial and horizontal distribution) are essential and predictions of the subtle differences in cloud coverage is essential for accurate air quality predictions. It follows that models must incorporate the feedback between cloud fields and boundary layer processes. Further, air pollution events tend to be multi-day in duration allowing for a pollution buildup through weak, recirculating flow fields. Accurate modeling of such scenarios will require increased understanding of the linkages between complex land surface conditions at different spatial resolutions and the atmospheric boundary layers. The modeling of important processes for air quality prediction will require additional research and modeling on most of the processes listed above. 3. Current Capabilities The U.S. EPAModels-3 framework provides a community-based computational framework for air quality predictions and/or assessments (projections). Its computation framework uses an object oriented programming for plug-and-play capability on meteorology, emissions processing and gas and aqueous atmospheric chemistry processes within a high performance computing environment. The Models-3 framework uses as its meteorology preprocessor, the MM5-FDDA system with an emission processing system. It is in general extensible for air quality forecasting with meteorological preprocessors such as the ETA model used in the forecast mode and emissions processing for future emission inventory with hourly projections. Models-3 could provide a basis for near as well as long term development of air quality forecasting tools. It would allow the introduction and testing of improved parametric formulations that are developed as scientific insights involved in air quality modeling are advanced. Features such as nesting to handle regional and city specific forecasts are, in principle, possible. In order to achieve the increased accuracy and confidence of tools to be used to predict air quality, important scientific advancements and improvements are needed in several key areas. These include: improved moisture modeling in the surface and surface boundary layer, the modeling and/or data assimilation of non-precipitating clouds fields, aggregating and special treatment to sub grid surface (roughness, vegetative, land use, terrain) features, hydrostatic vs. non-hydrostatic modeling, matching process parameterization to model scale resolution, and data assimilation vs. dynamic initialization of meteorological fields. Also, the accuracy of emissions inventory in the projection mode has not been assessed. Therefore, the current emission inventory models will require significant effort for its adaptation to air quality forecasting. 3.4 Parameterizing the Ground Surface Interactions(Facilitator and author: Ken Mitchell, NCEP) Sixteen workshop members met for about one hour to discuss approaches, experiences, and technical issues related to the physical modeling of ground surface interactions with the atmosphere in mesoscale models. The facilitator opened with some introductory comments noting the literal explosion of interest worldwide over the last five years in advancing the science of coupling modern-era land-surface schemes to mesoscale (and global) atmospheric models, via such research programs as GCIP, PILPS, and ISLSCP. NOAA in particular is strongly sponsoring land-surface modeling research through the NOAA GCIP Program Office, represented at these meetings by Rick Lawford. Both NCEP/EMC and ERL/FSL have actively collaborated over the last two years in their respective GCIP-sponsored land-surface modeling initiatives. During the discussions, other NOAA Labs such as ARL/ASMD and ERL/NSSL were encouraged to consider fostering their NOAA land-surface modeling collaborations within the existing NOAA GCIP umbrella. After the above introduction, the facilitator presented an outline of the following five major land-surface modeling topics:
Regarding the first topic, it was noted from the previous day's models overview, that most of the participating modeling groups had avoided complex vegetation canopy treatments such as SSiB and BATS, which require the specification of 20 or more vegetation and soil parameters, in favor of more moderate treatments along the lines of that of Noilhan and Planton (1989), which require on the order of about one dozen vegetation and soil parameters. Thus participants strongly encouraged each other over the next months to share experiences regarding their respective parameter sensitivity experiments, for such key parameters as green vegetation fraction, leaf area index, minimal stomatal resistance, soil porosity, soil moisture capacity, root distribution, and soil hydraulic conductivity. The development and testing of any chosen land-surface scheme requires a concerted validation effort (topic two). The group discussion clearly emphasized that validation efforts require both a) site specific validation (against surface based measurements) and b) regional validation over the mesoscale model domain (utilizing satellite data, for example GOES retrievals of hourly surface skin temperature available from the GCIP archive). In the site-specific area, the group agreed that collaboration would be greatly enhanced if the modeling groups could agree on a short list of sites for which the groups would routinely generate detailed time series output from their models of surface water and energy fluxes, surface state variables, and related surface quantities. It was pointed out that the Project for Intercomparison of Land-Surface Process Schemes (PILPS), co-sponsored by GCIP, has over the last five years provided a convenient formal structure in which to conduct collaborative site specific model validation and sensitivity experiments, by providing common surface forcing and surface validation data sets at various field sites for land-surface model testing. The PILPS Special Issue of the June 1996 issue of Global and Planetary Change (e.g. Henderson-Sellers, 1996) in particular provides a set of seminal papers on a wide-ranging, site-specific, land-surface model intercomparison and sensitivity study, involving 14 land-surface schemes. A fundamental conclusion of the above PILPS study is that validation of land-surface schemes cannot only consider the warm (growing) season, but must consider the full annual cycle and must pay more attention to the physical parameterization of runoff. In particular, over the long annual time scales embodied in continous NWP assimilation systems such as the ERL/FSL MAPS assimilation or the NCEP/EMC Eta EDAS assimilation, the PILPS results showed that the partitioning of incoming radiative energy between sensible heat flux and latent heat flux (bowen ratio) is highly tied to the partitioning of incoming precipitation between evaporation and runoff. Hence errors in runoff modeling (say through a poor surface infiltration parameterization) will eventually result in evaporation errors. Conversely, errors in evaporation modeling (say through a poor canopy resistance formulation) will eventually result in runoff errors. The above longer time scale issues (6-12 months) are crucial to mesoscale modeling groups because of the problem of soil moistue initialization (topic 3). It was agreed that the NOAA mesoscale modeling groups need time-dependent initial soil moisture rather than a soil moisture climatology (although further sensitivity studies are needed in this area). Other than empirical soil moisture initialization methods, such as various forms of Antecedent Precipitation Index (API), continuously cycled soil moisture in 4-D Data Assimilation Systems (4DDA) is the leading candidate for soil moisture initialization. Present and near-term satellite remote sensing of soil moisture (e.g. microwave SSM/I retrievals) are still plagued by the inability to retrieve soil moisture a) below a vegetation canopy (except say in extremely wet, ponded or flooded cases) or b) to a significant depth in the case of bare soil (say beyond 2-3 cm). One somewhat hopeful satellite approach is to infer an effective soil moisture from GOES retrievals of morning rise in surface skin temperature. Thus soil moisture from 4DDA systems appears to hold the most imminent promise for soil moisture initialization, but even here the challenges are daunting. These challenges include the inevitable drift in 4DDA soil moisture due to precipitation and solar insolation biases in the assimilating atmospheric model, as well as errors in canopy resistance and runoff formulations, as discussed earlier. Two approaches being considered by various groups to attack the drift problem were discussed. One is to execute an "uncoupled" land-surface assimilation, wherein the land-surface scheme is continuously integrated using observed precipitation (e.g. from national WSR-88D network) and observed net surface solar insolation (e.g. the NESDIS GOES retrievals on a 0.5 degree grid of hourly net surface insolation), thereby avoiding model biases in precipitation and surface insolation. Another approach is to nudge the soil moisture state variable (either toward drier or wetter) according to differences between observed and assimilating model's 2-m temperature and dewpoint. In sharing soil moisture fields among modeling groups, the PILPS experience has shown that the only entity that has any hope of being shared is soil moisture "fraction" (generically defined as the fraction from the low to high end of a model's effectively realizable dynamic range of soil moisture), since the various land-surface schemes have widely varying dynamic ranges of absolute soil moisture. Finally, the group discussed at some length several collaboration opportunities in the area of acquiring and gridding various land surface characteristics databases, including the USGS 1-km vegetation classification database and the Penn State 1-km STATSCO soils classification database. Considerable development and testing is needed to identify reasonable algorithms to represent these high-resolution classification databases on coarser resolution (5-50 km) model grids, including compaction of the number of classes. To specify the seasonal vegetation cycle, NESDIS/ORA has recently released an exciting new 5-year climatology of NDVI-based monthly green vegetation fraction globally at 0.15 degree resolution. NCEP/EMC is implementing the latter operationally in the Eta model land surface scheme and can make it available to any interested NOAA modeling group. In summary, the potential for collaboration among NOAA regional modeling groups in land-surface parameterization seems good in the following areas:
4. SUMMARY OF QUESTIONNAIRESQuestionnaires were given out during the discussion session to help the participants address the goals of the meeting. Questions were asked which would determine the participants views on the most important recommended future study topic, its feasibility and the participants interest in collaborating. Questions were also asked regarding mesoscale modeling research topics which are not currently being addressed by NOAA. The questionnaires were used by the facilitator in helping to produce a consensus and making recommendations. The following summarizes the results of individual responses for each discussion group (scale 1-10 used, 1 being not interested or feasible). 4.1 Overall RecommendationsAll groups were asked what mesoscale modeling topic would be most valuable for NOAA to work on, irregardless of resources. The following summarizes the topics chosen with the number in parenthesis indicating the number of participants who chose this topic.
4.2 Four-dimensional Data Assimilation
(4)
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Recommended future study topic | Feasibility | Collaborative interest |
|
7,9,3,7 | 8,8,9,3 |
Research areas not addressed:
|
||
| Recommended future study topic | Feasibility | Collaborative interest |
|
8,10,5,10,8,8,7 | 8,8,8,5,4,5,8 |
Research areas not addressed:
|
||
| Recommended future study topic | Feasibility | Collaborative interest |
|
10,5 | 8 |
|
8 | - |
|
8 | 10 |
Research areas not addressed:
|
||
| Recommended future study topic | Feasibility | Collaborative interest |
|
8 | 10 |
| 10 | 10 |
|
3 | 5 |
Research areas not addressed:
|
||
This first meeting of the NOAA mesoscale modelers working group served as a good vehicle to exchange information on mesoscale modeling at the working level. Many participants found that our work was more closely linked than previously thought which helped to begin identifying areas of intersection and potential further interaction. While many labs are using mesoscale models to support individual goals, the tools being developed were often similar (for example: improved methods to parameterize the boundary layer, visualization of model outputs). The primary outcome of this meeting, however, was that it served as a vehicle for information exchange between NOAA labs. The information exchange worked well since the meeting format was structured to be informal and required little preparation. The workshop also brought together scientists from diverse communities (operational forecasting, severe storm research, air pollution assessments) but who have common interests and needs for mesoscale modeling. Future meetings, however, should provide time for more questions when current activities are being presented.
Several key areas were identified by the working group as deserving more support by NOAA and which would be strongly related to the NOAA mission of environmental assessment and prediction :
Several labs reported on useful products and in-house
expertise which could be shared by other NOAA labs
to address their goals :
| ARL: | Air quality assessment and prediction tools. Several atmospheric chemistry models, visualization, land surface parameterization. |
| ETL: | Air-water/ocean interactions parameterizations, visualization. |
| FSL: | RUC and LAPS 4DDA analysis tools; parallel processing techniques, visualization, PBL processes. |
| GFDL: | Non-hydrostatic effects, tropical convection. |
| GLERL | Air-water/lake interactions. |
| NCEP: | EDAS archive, the workstation version of the RSM, 4DDA, land surface parameterizations, ensemble forecasting. |
| NSSL: | Ensemble forecasting techniques, moist convective processes. |
| OFCM: | Interagency modeling coordination. |
Some follow-up activities were suggested for the group
to consider. OFCM had
offered to host a follow-up meeting in the next year which
would finalize the identification of future research areas
in mesoscale modeling and begin structuring formal interactions
between labs. A NOAA mesoscale
modeling internet web page could also be initiated which
would help to update NOAA scientists on latest developments
and findings. Participants also expressed a need to better
communicate the results of this meeting to the lab directors,
therefore, feedback from the administrators is necessary.
Bender, M. A., R. J. Ross, R. E. Tuleya, and Y. Kurihara, 1993: Improvements in tropical cyclone track and intensity forecasts using the GFDL initialization system. Mon. Wea. Rev., 121, 2046-2061.
Bender, M. A., R. E. Tuleya, and Y. Kurihara, 1987: A numerical study of the effect of island terrain on tropical cyclones. Mon. Wea. Rev., 115, 130 - 155.
Benjamin, S. G., D. Kim, and T. W. Schlatter, 1995: The Rapid Update Cycle: A new mesoscale assimilation system in hybrid theta-sigma coordinates at the National Meteorological Center. Preprints, Second International Symposium on Assimilation of Observations in Meteorology and Oceanography, Tokyo, Japan, 13-17 March, 337-342.
Benjamin, S. G., G. A. Grell, K. J. Brundage, T. L. Smith, J. M. Brown, T. G. Smirnova, and Z. Yang, 1995: The next version of the Rapid Update Cycle -- RUC II. Preprints, 6th Conference on Aviation Weather Systems, 15-20 January, Dallas, AMS, 57-61.
Berbery, E., E. Rasmusson, and K. Mitchell, 1996: Studies of North American continental-scale hydrology using Eta model forecast Products, J. Geophys. Res., 101, 7305-7319.
Betts, A., F. Chen, K. Mitchell, and Z. Janjic, 1996: Assessment of the land-surface and boundary-layer models in two operational versions of the NCEP Eta model using FIFE data, Mon. Wea. Rev., (Submitted).
Binkowski, F. and U. Shankar, 1995: The Regional Particulate Matter Model 1.Model description and preliminary results. J. Geophys. Res., 100, 26191-26209.
Black, T., 1994: The new NMC mesoscale eta model: Description and forecast examples. Wea. Forecasting, 9, 265-278.
Burk, S.D. and W.T. Thompson, 1989: A vertiically nested regional numerical weather prediction model with second-order closure physics. Mon. Wea. Rev., 2305-2324.
Carter, W. P. L., 1990: A detailed mechanism for the gas-phase atmospheric reactions of organic compounds. Atmos. Environ. 24A, 481-518.
Chang, J. S., R. A. Brost, I. S. A. Isaksen, S. Madronich, P. Middleton, W. R. Stockwell, and C. J. Walcek, 1987: A three-dimensional Eulerian acid deposition model: Physical concepts and formulation. J. Geophys. Res. 92, 14681-14700.
Chen, F., and co-authors, 1996a: Modeling of land surface evaporation by four schemes and comparison with FIFE observations, J. Geophys. Res., 101, 7251-7268.
Chen, F., Z. Janjic, and K. Mitchell, 1996b: Impact of atmospheric surface layer parameterization in the new land-surface scheme of the NCEP mesoscale Eta Model, Boundary-Layer Meteor., (Submitted).
Draxler, R. R., G. D. Rolph, J .T. McQueen, J. L. Heffter, and B. J. B. Stunder, 1993: NOAA Washington GDPS Regional Specialized Meteorological Center on the provision of transport model products for environmental emergency response. RSMC Meeting, 19-24 September, 1993, Montreal, Canada.
EPA, 1992: Interagency Workgroup on Air Quality Modeling (IWAQM) Work Plan Rationale. EPA-454/R-92-001. U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Research Triangle Park, North Carolina. 23 pp.
EPA, 1993: Interagency Workgroup on Air Quality Modeling (IWAQM) Phase I Report: Interim Recommendation for Modeling Long Range Transport and Impacts on Regional Visibility. EPA-454/R-93-015. U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Research Triangle Park, North Carolina. 113 pp.
EPA, 1995: Interagency Workgroup On Air Quality Modeling (IWAQM): Assessment of Phase I Recommendations Regarding the Use of MESOPUFF II. EPA-454/R-95-006. U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Research Triangle Park, North Carolina. 134 pp.
Gery, M. W., G. Z. Whitten, J. P. Killus, and M. C. Dodge, 1989: A photochemical mechanism for urban and regional scale computer modeling. J. Geophys. Res. 94:12925-12956.
Gross, B. D., 1996: The effect of compressibility on barotropic and baroclinic instability. J. Atmos. Sci., (In press).
Gross, B. D., 1994: Frontal interaction with isolated orography. J. Atmos. Sci., 51, 1480-1496.
Henderson-Sellers, A., 1996: Soil moisture simulation: Achievements of the RICE and PILPS intercomparison workshop and future directions, Global and Planetary Change, 13, 99-115.
Irwin, J. S., J.S. Scire, and D.G. Strimaitis, 1995: A comparison of Calpuff modeling results with CAPTEX data results. 21st International Technical Meeting on Air Pollution Modeling and Its Application, November 6-10, 1995, Baltimore, MD. 8 pp.
Janjic, Z., 1990: The step mountain coordinate: Physical package. Mon. Wea. Rev., 118, 1429-1443.
Janjic, Z., 1994: The step-mountain eta coordinate model: Further developments of the convection, viscous sublayer, and turbulence closure schemes. Mon. Wea. Rev., 122, 927-945.
Kumar, A., M. Hoerling, M. Ji, A. Leetmaa, and P. Sardeshmukh, 1996: Assessing a GCMÃs suitability for making seasonal predictions. J. Climate, 9, 115-129
Jeffries H.E and S. Tonnesen 1994: A comparison of two photochemical reaction mechanisms using mass balence and process analysis. Atmos. Environ., 28, 2991-3003.
Kurihara, Y., and M. A. Bender, 1980: Use of a movable nested mesh model for tracking a small vortex. Mon Wea. Rev., 108, 1792-1809.
Kurihara, Y., M. A. Bender, and R. J. Ross, 1993a: An initialization scheme of hurricane models by vortex specification. Mon. Wea. Rev., 121, 2030-2045.
Kurihara, Y., M. A. Bender, R. E. Tuleya, and R. J. Ross, 1990: Prediction experiments of Hurricane Gloria, 1985, using a multiply-nested movable mesh model. Mon. Wea. Rev., 118, 2185-2198.
Kurihara, Y., M. A. Bender, R. E. Tuleya, and R.J. Ross, 1995: Improvements in the GFDL hurricane prediction system. Mon.Wea. Rev., (Submitted).
Kurihara, Y., C. L. Kerr, and M. A. Bender, 1989: An improved numerical scheme to treat the open lateral boundary of a regional model. Mon. Wea. Rev., 117, 2714-2722.
Lamb, R.G. 1983: A Regional Scale (1000 km) Model of Photochemical Air Pollution, Part-1: Theoretical Formulation. EPA-600/3-83-035, U.S. Environmental Protection Agency, Research Triangle Park, NC.
Lofgren and Croley , 1995: Validation of the Coupled Hydrosphere-Atmosphere Research Model (CHARM). Proceedings of the Conference on Hydrology, American Meteorological Society, 15-20 January 1995, Dallas, Texas, 26-30.
McQueen, J. T., R. R. Draxler, and G. D. Rolph, 1995: Influence of grid size and terrain resolution on wind field predictions from an operational mesoscale model. J. Appl. Meteor., 34, 2166-2181.
McQueen, J. T., R. R. Draxler, B. J. B. Stunder, and G. D. Rolph, 1996: An overview of the Regional Atmospheric Modeling System (RAMS) as applied at the NOAA/Air Resources Laboratory. NOAA Technical Memorandum ERL ARL- , pp. (In press)
McQueen, J.T., R.A. Valigura, and B.J.B. Stunder, 1996: Evaluation of the Regional Atmospheric Modelling System for estimating nitric acid pollution deposition onto the Chesapeake Bay. Atmos. Environ. (Submitted)
Mellor, G. L., and T.Yamada, 1974: A hierarchy of turbulence closure models for planetary boundary layers. J. Atmos. Sci., 31, 1791-1806.
Noilhan, J., and S. Planton , 1989: A simple parameterization of land surface processes for meteorological models. Mon. Wea. Rev., 117, 536-549.
Orlanski, I., and J. P. Sheldon, 1995: Stages in the energetics of baroclinic systems. Tellus, 47A, 605-628.
Orlanski, I., and B. D. Gross, 1994: Orographic modification of cyclone development. J. Atmos. Sci., 51, 589-611.
Orlanski, I., and J. P. Sheldon, 1993: A case of downstream baroclinic development over western North America. Mon. Wea. Rev., 121, 2929-2950.
Persson, P. O. G., P. J. Neiman, and F. M. Ralph, 1996: The role of a topographically generated potential vorticity anomalies in initiating a coastal wind reversal. Preprints, Conf. on Coastal Oceanic and Atmospheric Prediction, Jan. 28 - Feb. 2 1996, Atlanta, GA., pp. 120-125.
Persson, P. O. G., P. J. Neiman, and F. M. Ralph, 1995: Topographically generated potential vorticity anomalies: A proposed mechanism for initiating orographically trapped disturbances. Preprints, 7th AMS Conference on Mountain Meteorology, 17-21 July 1995, Breckenridge, CO.
Pleim, J. E. and A. Xiu, 1995: Development and testing of a surface flux and planetary boundary layer model for application in mesoscale models. J. Appl. Meteorol., 34, 16-32.
Pleim, J. E., J. F. Clarke, P. L. Finkelstein, E. J. Cooter, T. G. Ellestad, A. Xiu, and W. M. Angevine, 1995: Comparison of measured and modeled surface fluxes of heat, moisture and chemical dry deposition. Presented at the 21st NATO/CCMS International Technical Meeting on Air Pollution Modeling and its Application, Baltimore, MD, November 1995.
Rogers, E., D. Deaven, and G. DiMego, 1995: The Regional Analysis System for the operational "Early" Eta Model: Original 80-km configuration and recent changes, Wea. Forecasting, 10, 810-825.
Snook, J.S., J.M. Cram, and J.M. Schmidt, 1995a: A nonhydrostatic mesoscale modeling system configured for operational use. Tellus, Vol. 47A, No. 5:1, 864-875.
Snook, J. S. and R. A. Pielke, 1995b: Diagnosing a Colorado heavy snow event with a nonhydrostatic mesoscale numerical model structured for operational use. Weather and Forecasting, Vol. 10, No. 2, 261-285.
Stauffer, D. R., and Seaman, N. L., 1990: Use of four-dimensional data assimilation in a limited-area mesoscale model. Part I: Experiments with synoptic-scale data. Mon. Wea. Rev., 118, 1250-1277.
Stauffer, D. R., N. L. Seaman, and F.S. Binkowski, 1991: Use of four-dimensional data assimilation in a limited-area mesoscale model. Part II: Effect of data assimilation within the planetary boundary layer. Mon. Wea. Rev., 119, 734-754.
Stockwell, W. R. 1986: A homogeneous gas-phase mechanism for use in a regional acid deposition model. Atmos. Environ., 20, No. 8:1615-1632.
Tuleya, R. E., 1994: Tropical storm development an decay: Sensitivity to surface boundary conditions. Mon. Wea. Rev., 122, 291-304.
Tuleya, R. E., M. A. Bender, and Y. Kurihara, 1984: A simulation study of the landfall of tropical cyclones using a movable nested-grid model. Mon. Wea. Rev., 112, 124 -136.
Wilczak, J. M, W. D. Neff, D. Ruffieux, and J. Schmidt, 1994: Atmospheric and oceanic mesoscale flows in the Monterey Bay, California region. Preprints, Second International Conference on Air-Sea Interaction and on Meteorology and Oceanography of the Coastal Zone, 22-27 Sept., 1994, Lisbon, Portugal, pp. 194-197.
Wilkinson, J. G. and B. Emigh. 1994: The Geocoded Emissions Modeling and Projections (GEMAP) System Training Workshop. Sacramento, CA. January 19-21, 1994, Presented to the California Air Resources Board by Alpine Geophysics, 270 pp.
Yarosh, E., C. Ropelewski, and K. Mitchell, 1996: Comparisons of humidity observations and Eta model analyses and forecasts for water balance studies during GIST, J. Geophys. Res.,(In press).
Ying, R., V. M. Canuto, and R. M. YPMA, 1994: Numerical simulation of flow data over two-dimensional hills. Boundary-Layer Meteorology, 70, 401-427.
Ying, R., and V. M. Canuto, 1996: Turbulence modelling over two-dimensional hills using an algebraic Reynolds stress expression. Boundary-Layer Meteorology, 77, 69-99.
Zhao, Q., and F. Carr, 1996: A prognostic cloud scheme
for NWP models, Mon. Wea. Rev., (Submitted).
(Jason Ching, ARL/ASMD, Research Triangle Park, NC)
The Atmospheric Sciences Modeling Division (ASMD) of NOAA's Air Resources Laboratory (ARL) is assigned to provided meteorological support to the U.S. Environmental Protection Agency (USEPA). It is located in Research Triangle Park, NC and is an integral component of the USEPA National Exposure Research Laboratory (NERL). Some of the major activities of ASMD are to: 1) develop and evaluate air quality and deposition modeling systems as research tools for investigating air pollution phenomena; 2) predict concentration and deposition fields; 3) provide retrospective and prospective analyses for use in promulgating national ambient air quality standards (NAAQS); and 4) examine various and optimal source control strategies. Recently, regional scale modeling of air pollution has been developed by ASMD including the Regional Oxidant Model (ROM) (Lamb); the Regional Acid Deposition Model (RADM) (Chang et al., 1987); and the Regional Particulate Model (RPM) (Binkowski and Shankar, 1995).
These prototypical systems will soon be replaced by Models-3,
a new major modeling system that incorporates state-of-the-science
formulations of critical atmospheric-biospheric processes
controlling air pollution into high performance computing
environments to serve as a predictive tool for air quality
research and it management. When introduced, Models-3 will
greatly increase the functionality of current systems in
the following ways: (1) Models-3 is designed to be a community
based framework that will accommodate participation by
a wide user community from scientists to policy makers.
System functionalities include control strategies decision
support, model building and development, and analyses tools
including visualization and model evaluation. Study planning
and implementation using the method of Directed Acyclical
Graphs (DAG), a flexible graphical user interface (GUI),
will provide a powerful mechanism for system usage by a
wide variety of users including scientists and policy makers.
When fully implemented, use of object oriented programming
will make model building and model/process module evaluation
open to the scientific community for applications and for
scrutiny. Model quality will be facilitated with automatic
documentation. Each oestudy" will be unique, reproducible
and traceable. A set of 11 volumes (several still in preparation)
provides documentation of the Models-3 framework. Volume
VII describes the system requirements in great detail.
(2) Models-3 will allow great flexibility for scientists
to build and examine scientific capabilities and credibility
into the system. Such features will provide an efficient
way to evaluate and improve the state-of-science of the
modeled processes affecting pollutant fields.
1. Models-3 Air Quality Modeling System
The air quality modeling system is at the core of the Models-3 system framework. The system framework consists of six processing layers including the user interface, a system manager, a UNIX environment, the computational program layer, data access, and data storage. The air quality modeling system resides in the computational program layer and consists of subsystems such as meteorology, emissions, chemistry-transport model (CTM), and analysis/visualization. Models-3 utilizes a state-of-the-art input/output system to facilitate efficient and disciplined subsystem communication. Each subsystem briefly summarized below is itself a complex modeling system.
Meteorology model: The in-situ meteorological observational network provides basic meteorological parameters at locations designed to detect synoptic weather systems. However, air quality modeling usually requires much finer scale information than the measurements can provide. For this reason a predictive numerical meteorological model is used as a driver for a CTM. The meteorology model is expected to show the following characteristics: (a) Faithful representation of atmospheric dynamics; (b) Best estimation of derived meteorological parameters that are essential for air quality predictions (e.g., mixing height, momentum and heat flux, etc.); and (c) Scalability to handle different resolution, grid system, and topography. While several options exist, the prototype will incorporate the Meteorological Mesoscale Model Version 5 (MM5) model and include a four dimensional data assimilation of observations to provide maximally accurate "data". Additionally, this system will contain nesting and the necessary hydrostatic/nonhydrostatic modes. Improved soil-moisture and vertical diffusion processes will provide the MM5 system with contemporary boundary layer science.
Coupled Land Surface and Dry deposition Model: A new dry deposition model coupled to a new land surface scheme which has been incorporated into the Fifth Generation Penn State/NCAR Mesoscale Model (MM5) has been developed as part of Models-3. The new land surface model includes explicit soil moisture and vegetative evapotranspiration to better estimate surface fluxes of latent and sensible heat in the meteorology model (Pleim and Xiu 1995). Key parameters produced by this model which control surface-atmosphere exchange processes, namely aerodynamic resistance and canopy resistance, are then used in a resistance analog dry deposition model. The canopy resistance is an aggregation of leaf based stomatal resistance which is parameterized as functions of environmental factors including soil moisture. Therefore, coupling the dry deposition model to the land surface model results in a better estimate of canopy resistance which is responsive to dynamic soil moisture conditions. Also, the land surface model in the modified MM5 includes an indirect nudging scheme which uses forecast errors of lowest model level temperature and humidity compared to observed analyses to nudge model soil moisture. This scheme results in more realistic canopy conductances than can be estimated by stand-alone dry deposition models. A preliminary evaluation of these modeling techniques has been made through comparison to field measurements over a corn field in Bondville, Illinois during August, 1994 (Pleim et al. 1995). Model estimates of surface fluxes of latent and sensible heat, soil heat flux, net radiation, and boundary layer height, all compared quite well to field measurements. Also, comparisons of dry deposition velocity of ozone as well as canopy conductance showed good agreement. Further evaluations of the model through comparison to other field experiments with more varying conditions are being conducted.
Emissions model: Models-3 targeted emissions modeling subsystem (EMS) will be capable of (a) speciating for different chemical mechanisms; (b) providing temporal and spatial allocation of emissions factors; (c) processing point source and area source (mobile, non-mobile, biogenic) emissions; (d) developing control strategy projections, including "across-the-board, of by source type, locality, economic sectors etc.; and (e) generating appropriate emissions inputs to urban and regional Eulerian air quality models. The EMS should allow easy updates and modifications to the system as the science, methodologies, or air pollution problems change. An adaptation of the GEMAP (Geocoded Emissions Modeling And Projection System) (Wilkinson and Emigh, 1994) is currently the targeted initial emissions model with appropriate GEMAP input processor (GIP) and output processor (GOP) to link with meteorology and CTM subsystems. The domain currently includes most of North America, and grid resolution may be one square kilometer or larger. Tests are planned using grid resolution ranging from 4 km to 80 km resolution. Temporal resolution for episodic modeling is determined by a series of source category specific files which reflect current knowledge.
The GEMAP contains basic emission data processing functionality, including spatial allocation (gridding) of point and area source emission inventory data, temporal allocation of emission data for episodic modeling purposes, and chemical speciation of emission data for ozone air quality modeling. The capability to model mobile source and biogenic emission data with EPA models has been added. Mobile source emissions are addressed with the EPA Mobile 5a model using meteorological data from MM5 and vehicle miles traveled. Biogenic emissions are modeled using the Biogenic Emission Inventory System-2 (BEIS-2) model. The pollutant species are computed to operate with different or generalized chemical speciation mechanisms .
Projection of a base year inventory to future or past years from a base inventory year for modeling and regulatory analysis is accomplished using a combination of modified tools previously developed for EPA. The tools are a part of a projection module for the emission system, called the Strategy Manager. The Strategy Manager is intended to allow iterative "what-if" approaches to determine optimal pollution control strategies. The projection module interacts with GEMAP by accepting base year annual emission data and provided projected data back to GEMAP for further processing. The user may specify the emission data of interest and the module estimates future year emissions by using economic growth factors. Biogenic emission data may be projected only by changing the land cover data. When control cost information is known, the economic costs of applying controls may be evaluated.
Chemistry-Transport Model (CTM): The CTM is the symbolic engine of Models-3. The design attributes require the CTM to: (a) be a state-of-the-art air quality modeling system capable of handling multi pollutant issues (e.g., oxidants, acid deposition, visibility, and particulate matter); (b) provide a standard interface that facilitates interchange of science modules with an intent to be a community modeling system; and (c) provide advanced modeling capabilities with the flexibility to operate at a spectrum of spatial scales. Additionally, the system should be easily extensible in providing means to incorporate additional pollutants and processes. A CTM solves a set of atmospheric mass continuity equations coupling chemistry with advection and diffusion processes. The various atmospheric processes currently modeled include gas and aqueous phase chemistry, dry deposition, wet scavenging, and transport/advection. Models-3 will provide two optional solution methodologies, an operator splitting and a flow-through-reactor, to achieve efficient numerical solutions of terms in the governing equations. In the current configuration, the chemistry-transport model (CTM) currently includes current and extensible formulations for the following processes or model tasks:
Analysis and visualization: A measure of the quality
and degree of credibility of model simulations is the verisimilitude
or degree to which the result conforms to reality. Customary,
model results are compared against very sparse and intermittent
observations of pollutant species concentrations and depositions.
However, for certain air quality problems such as tropospheric
ozone, direct comparison of concentrations is too limiting
to predict confidently, the results of future emissions
projects. Since the intentions for model applications involve
a variety of emissions control projections, it is critical
that the modeling system provide the evaluation capability
to investigate and understand the inter-connectedness of
the physical-chemical system and of the feedback contained
in the highly nonlinear chemistry-dynamics simulations
of the secondary pollutant concentrations. To achieve this
level of understanding, Models-3 will
implement the IRR/MB (Integrated Reaction Rates and Mass
Budget) processor analysis routine. A description of the
chemical component of the processor is given by Jeffries
and Tonnesen, 1994. Other postprocessing routines such
as a trajectory analysis for studying source-receptor relationships,
pointer-flyer techniques for comparing model output with
aircraft observations, and a grid-cell time series analysis
for evaluating model diagnostics with hourly surface observations.
A visualization subsystem will provide an effective and
powerful tool for handling massive amounts of model generated
evaluation data.
2. Multi-scale Air Quality Modeling
The Models-3 coordinate
systems will be generalized to allow transformation among
various vertical coordinates (e.g., physical height, pressure)
and among various horizontal coordinates, especially map
projections (e.g., spherical, rectangular, Mercator, polar
stereographic) by simple changes in a few scaling parameters,
boundary conditions, map origin, and orientation. This
will permit the adaptation of each of the component science
models including the CTM and the meteorological models
to any of the coordinate systems commonly used in atmospheric
modeling. Nesting capabilities and plume-in-grid treatment
of major pollutant point sources in Models-3 will provide
the means to resolve fine scale features such as found
in urban areas, in regions with adjoining urban areas,
and in complex terrain conditions.
3. Comprehensive Multi Pollutant Predictions
Many different air pollution problems are recognized
by scientists and the public. They are: acid deposition,
tropospheric ozone, stratospheric ozone, global warming,
particulates, and toxics, etc. Even today, new air quality
problems are discovered by scientists. The involved chemical
reactions of these air pollution problems are tremendously
different. To transcend from old practices of developing
a one air-pollution-issue air quality model, the Models-3 system
will include generalized chemistry mechanism readers and
solvers. The prototype version of Models-3 will include
in the CTM, the three important atmospheric chemical mechanisms:
the Carbon Bond Mechanism IV (CBM-IV) (Gery et al., 1989),
the RADM mechanism
(Stockwell, 1986) and SAPRC-90 (Carter, 1990). The chemistry
components are usually hand-coded to represent a specific
reaction mechanism, thereby requiring considerable recoding
when the mechanism must be replaced or enhanced. In the
Models-3 system a generalized representation of the chemistry
will be possible. Models-3 will enable consideration of
pollution in both gas and particle phase. Pollution as
aerosols, their emissions, size distribution and model
dynamics, chemistry (gas, heterogeneous and aqueous phase)
evolution, and transport and deposition are incorporated
into the science framework.
Discussion:
In each of these models, meteorological preprocesors
provide the winds, temperature and moisture fields from
which gridded hourly fields of pollutant transport, transformation
and deposition are predicted. These air quality predictions
would relate to historical meteorological events or episodes,
and therefore would not represent a true forecast in the
meteorological sense but calculated fields based on well
characterized synoptic events conducive to conditions of
high pollution. In the lexicon of EPA,
such model predictions provide retrospective analyses of
base source emissions inventories, and a prospective for
emissions projections, using the same meteorological events.
In principle, a meteorological and emissions preprocessor
providing forecasted conditions for future weather and
emissions could provide the basis for air pollutant forecasting.
This mode of modeling has not had previous or current utility
for implementing NAAQS strategies.
A.1.2 Preparation of Meteorological Data for Air Quality Modeling: IWAQM
(John Irwin, U.S. EPA, Research Triangle Park, NC)
The Interagency Workgroup on Air Quality Modeling (IWAQM) was formed to provide a focus for development of technically sound, regional air quality models for regulatory assessments of pollutant source impacts on Federal Class I areas. Meetings were held with personnel from interested Federal agencies, viz. the Environmental Protection Agency, the U.S. Forest Service, the National Park Service, and the U.S. Fish and Wildlife Service. The purpose of these meetings was to review respective regional modeling programs, to develop an organizational framework, and to formulate reasonable objectives and plans that could be presented to management for support and commitment. The members prepared a memorandum of understanding (MOU) in November of 1991 (EPA, 1992) that incorporated the goals and objectives of the workgroup and obtained signatures of management officials in each participating agency. Although no States are signatories, their participation in IWAQM functions is explicitly noted in the MOU.
In the early discussions the workgroup recognized that formal guidance was needed on the use of atmospheric dispersion models suitable for characterizing transport on the order of 50 to 200 km. These models are needed to estimate pollutant concentrations, including the individual and cumulative impacts of proposed and existing sources on Air Quality Related Values (AQRVs), Prevention of Significant Deterioration increments, and National Ambient Air Quality Standards, with emphasis on Federal Class I areas. Therefore one of the first efforts by the workgroup was to review existing modeling techniques in order to recommend an interim approach for long-range transport modeling capable of providing the information to asse
