Meteorological Applications & Assessment Division Meteorological Observational and Analysis Support 1996-1997 Northern Front Range Air Quality Study (NFRAQS)

The Desert Research Institute has been selected to carry out the Northern Front Range Air Quality Study (NFRAQS) so as to address the general goals of the Colorado House Bill, HB95-1345, and, in the context of limited resources, to meet three goals:

1. Source apportionment of carbonaceous material in airborne particles in the NFRAQS study region.
2. Determination of whether the Denver area is ammonia-rich (or anion-deficient) with regard to the formation of secondary ammonium nitrate and ammonium sulfate.
3. Apportionment of sources leading to the formation of non-carbonaceous portion of PM_2.5, with emphasis on nitrate and sulfate.

The study has three experimental phases. The first phase was a preliminary air chemistry monitoring program operating from mid-January through the end of February 1996 at the Welby monitoring site. This site was used in the 1987-1988 Denver Brown Cloud Study (DBCS) and was proposed as a core site for the 1996-1997 summer and winter field programs. The second phase, the 1996 Summer Study involved not only the core site at Welby but also three satellite sites along the Front Range where data were obtained on a forecast basis. Finally, the third phase, a major field study will occur in the winter of 1996-1997 from late November through mid-February.

Underlying each of these experimental phases is the need for meteorological observations and analysis. Because of the distributed nature of primary emissions along the Front Range and the importance of moisture to many chemical processes, characterization of horizontal transport and vertical mixing processes as well as moisture sources, is essential to understanding the spatial and temporal evolution of Front Range pollution episodes. In addition, the nature of daily weather regimes determines to a large extent the relative importance of various pollution sources. For example, stagnant dry conditions may emphasize the role of carbonaceous materials whereas moist circulations may favor secondaries and multiday episodes may create conditions for both. In addition, some geographical regions, because of local topography and sources, may be dominated by one particular source category. Thus, documenting the meteorology of the Front Range during NFRAQS is critical to meeting the goals listed above and interpreting the results from the research that addresses them.

Following a long history of participation with remote sensing devices in studies of Front Range air quality that extends back to 1973, (including the meteorological observations and analysis for the 1987-1988 Denver Brown Cloud Study), NOAA's Environmental Technology Laboratory (ETL) has played a cooperative role in the first two phases of this program and in the planning for the major field phase in the winter of 1996-1997. This proposal addresses the major meteorological goals for the winter study and outlines the cooperative support we are providing for first two field phases.

In the original study plans for the winter of 1995-1996, there was little time allowed for the analysis of the large volume of data that might have been available. However, with the delay of the major field program until the winter of 1996-1997 and a greater period of time provided for the analysis of the resulting data, ETL will be able to provide more in-depth case study analyses of the data in addition to providing the meteorological data set to the DRI data base for wind field visualization and analysis by STI.

The broad scientific goals of the NOAA/ETL effort include

• Documentation, through case studies and climatological analyses, of the meteorological conditions conducive to pollution episodes through the South Platte River basin via integration of meteorological, light scattering and absorption, and fine-particle mass data.
• Provide a meteorological data set suitable for classification of episodes and diagnostic wind field modeling in the current study.
• Documentation of site-specific micrometeorology associated with pollution episodes at each core-observing site.
• Further revision and/or refinement of the conceptual model and classification methods originally developed by NOAA/ETL for the 1987-1988 Denver Brown Cloud Study Final Report. [This model has been expanded in subsequent scientific publications that used the final fine-particle mass concentration, visibility, and meteorological data sets provided by DRI and NOAA to SCENIC Denver as well as the more detailed meteorological research data sets available from later studies of transport and dispersion along the Colorado Front Range.]

2. Observational Strategy:

During the 1987-1988 DBCS, meteorological support, particularly profile measurements, was limited to sites within about 30 km of downtown Denver. As a result, much of the interpretation of those data was flavored by the limited domain of the observations. It was later realized (e.g. Neff and Watson, 1990; Neff and King, 1991a; Neff, 1996) that an understanding of Front Range air quality meteorology would only be gained by simultaneous observations covering a much larger segment of the South Platte airshed. Figure 1 shows the large scale terrain surrounding the South Platte river. In this figure, Denver and Fort Collins are separated about 90 km while Fort Morgan is about 110 km to the east. Greeley is located at the confluence of a number of drainages as can be seen more clearly in Figure 2. The Saint Vrain and Big Thompson rivers flow through Longmont and Loveland respectively and join the South Platte river north of Platteville. The Cache la Poudre joins the South Platte river near Greeley and then flows eastward past Kersey and on through Fort Morgan. In the 1987-1988 DBCS the Brighton and BAO Tower sites (Figure 2) formed the northern boundary of the study area.

2.1 Guidance from the 1987-1988 DBCS

As a result of preliminary measurements obtained during the winter of 1986-1987 a conceptual model for Denver air pollution episodes was developed by NOAA/ETL to assist the State in the forecasting of high carbon-monoxide episodes and in the preparation for the 1987-1988 Denver Brown Cloud Study. The model was then further refined and is summarized in Section 2.3 of Volume III of the DBCS Final Report contributed by NOAA (Appendix A, this proposal). This conceptual model provided the basis, using meteorological observations from Welby, for a reduction of the original chemistry data set into subsets for coal- and gas-burning periods, with comparable transport and dilution characteristics as described in Sections 8.1, 8.2, and 8.3 of the Report (Appendix B) and was the basis for the discussion of concentration data by DRI in Section 8.4. Finally, in Section 8.5 of the DBCS Final Report, NOAA assessed the DBCS results against the conceptual model with emphasis on the multiday episodes described further in Sections 8.6.1. and 8.6.2 of the Report (Appendix C). This was the point at which the DBCS project terminated further analysis of the data as part of the study. However, a number of scientifically interesting puzzles emerged from the DBCS that stimulated further inquiry: As a result of further analyses of the final SCENIC Denver data set and data collected as part of the 1991 Winter Validation Study at Rocky Flats (Neff, 1994, Levinson and Banta, 1995), the conceptual model was refined in the following manner (Neff and Watson, 1990; Neff et al., 1990; Neff, 1990; Neff and King, 1991b; Neff, 1994, 1996):

Figure 1

Figure 1 GIF Image

Figure 2

Figure 2 GIF Image

2.1.1 Meteorological Regimes

From Neff (1996), four meteorological regimes were identified as important for further research:

• Nocturnal drainage flows that follow the South Platte River from the southwest to the northeast through Denver. [ Hypothesis: The nocturnal drainage jet structure (Neff et al., 1990), because of a nearly laminar layer that forms between 100m and 200m, may result in the trapping of urban surface emissions in a thin layer below the jet and may isolate elevated emissions (from point sources) in the air flow above the wind maximum: This drainage system extends well into northeastern Colorado during summer (Toth and Johnson, 1985), including regions with significant ammonia sources . However, there were no observations of vertical mixing or transport processes outside of the Brown Cloud Study area.]
• Thermally and/or dynamically driven northeasterly winds (upslope, toward the foothills), often associated with a shallow front-like or surge structure only a few hundred meters deep, that can transport cool air from the lowlands of the South Platte, northeast of Denver, southwestward into the foothills. During the Brown Cloud Study, these winds were most likely to occur during the afternoon but were also observed at many other times of the day and night (Crow, 1973; Neff, 1990; Neff et al. 1990) and sometimes as a result of mesoscale eddies that form along the Front Range (e.g. Levinson and Banta, 1995). [ Hypothesis: These upslope and recirculating flows enable aged aerosol and/or precursor gases such as ammonia to return to Denver and may contribute to a rapid degradation of visibility (Sloane et al., 1990). The stability of the shallow air mass limits vertical mixing and allows further buildup of pollution. When alternating with a nocturnal drainage wind, they may lead to a day-to-day recycling of the same airmass.]
• Moist, cool northeasterly upslope winds, usually in response to lee cyclogenesis southeast of Denver and/or cold, surface high pressure developing over the Great Plains to the northeast of Denver, sometimes result in snowfall along the base of the mountains, but also in fogs and low clouds. [ Hypothesis: Such conditions can support rapid chemical transformations, such as SO₂ to sulfate, that depend on the presence of clouds (McHenry and Dennis, 1994). A related area that merits further investigation is melting and evaporation into the shallow boundary layers that often follow snowstorms.]

• Downslope westerly winds that usually are strongest near the foothills west and north of Denver and which are associated with falls of pressure along the foothills, contributing to shallow upslope flows along the Platte River. [ Hypothesis: Warm westerly winds several hundred meters aloft and light, cool easterly winds near the surface enhance the low-level temperature inversion creating strong trapping conditions unless there is a strong differential acceleration of the wind across the inversion layer. During the Brown Cloud field study, the inversion often proved remarkably resistant to erosion by the strong westerly winds above it.]

2.1.2 Air Quality Regimes

These meteorological regimes were then related to three characteristic air quality regimes. In particular, the 1987-1988 Brown Cloud field study resulted in source contributions to fine-particle mass concentrations divided into six major categories: 1) primary geological, 2) primary boiler, 3) primary vegetative burning, 4) primary mobile, 5) secondary ammonium nitrate, and 6) secondary ammonium sulfate. To a large extent, the difference in the character of the visibility reduction depends on the relative contribution of secondary ammonium nitrate and sulfate. Figures 3a,b,c show a range of three brown cloud events with the following fine-particle mass concentrations:

with the following chemical distribution for the Federal Building site:

¹ Daytime Sample

² Nighttime Sample prior to photograph (daytime samples were not analyzed in the original study).

These examples reflect three air quality regimes. The first is one dominated by primary emissions, the second by secondaries, and the third by a mixture of the first two. The first two reflect similar total fine-particle mass concentrations in the downtown area: however, the event dominated by primary emissions is characterized by considerable spatial inhomogeneity where that dominated by secondaries is quite homogeneous. The third case, with by far the highest concentrations, also reflects the greatest inhomogeneity in both carbonaceous and secondary material resulting from the meteorological conditions on that day (Neff, 1996).

Figure 3

Figure 3 GIF Image

REGIME I, Dry Conditions. Air masses lying over Denver are often fairly dry, thus minimizing the effect of many chemical transformations that depend on the presence of moisture to produce secondary aerosols, such as ammonium nitrate and ammonium sulfate. In these dry cases, light winds and temperature inversions can lead to an accumulation of high concentrations of primary emissions with relatively small contributions from secondary aerosols. Figure 3a, a view looking south toward downtown Denver at 1600 MST on 3 December 1987, shows a shallow brown cloud episode with late afternoon B_scat of 174 Mm^-1 at Auraria and 83 Mm^-1 at the top of the Federal Building (70 m). The brown cloud has a thick, opaque, brown appearance in this case; no snow is present on the ground at this time. For this daytime period, the DBCS reported major contributions to optical extinction from mobile sources and vegetative burning. There were virtually no secondary particle contributions, consistent with a relative humidity of less than 30%. This case followed a windstorm on 2 December and was characterized by light southerly winds underlying downslope winds from the west-northwest above 500 m, a condition that often leads to persistent daytime temperature inversions, as discussed earlier. The fact that the light, near-surface winds were from the south, rather than the east, may account for the low relative humidity. The nature of shallow temperature inversions in cases such as these lead to localized high concentrations in low-lying terrain as indicated in Table I. In this case Welby shows the highest concentration where measurement on more elevated outlying terrain (BAO Tower and Brighton) show the lowest.

REGIME II, Moist Upslope, without Precipitation. Another meteorological event that provides increased humidities involves nonprecipitating, moist, upslope circulations. These events, frequent along the Front Range during some seasons, bring shallow layers of high-humidity air laden with fog or low clouds from the eastern plains that persist for a few hours to several days. Such was the situation on the night of 19 December and the day of 20 December 1987, when the picture in Figure 3b was taken. At the time of the picture (1200 MST, 20 December), surface relative humidities of 65% had followed nighttime humidities of almost 100%. Note the presence of clouds in the photograph, which often result from such upslope flows. Buildings in downtown Denver in this photograph are barely visible at a range of just under 15 km, a standard visual range corresponding to B_ext of about 300 Mm^-1. In events such as these, the secondary aerosols are the principal contributors to haze, as evidenced in the BCI analysis of source-type contribution to visibility extinction shown in Table 2. Later analysis (for a subsequent modeling study) of filter material from the daytime period on 20 December showed the following concentrations:

¹ Includes Front and Back Filters.

² According to the DRI report these showed losses compared to the original filter analyses.

³ Daytime samples.

The lower levels of carbonaceous particles (particularly elemental carbon) in this type of event suggest that the air mass had not been exposed for a prolonged time to normal urban emissions, except possibly in the area of Welby. However, at issue in many such events is not so much the chemical composition during the initial haze or fog phase but the subsequent evolution of the haze in multiday events. For example, does the majority of the conversion to secondary aerosol occur only during the moist event (or in the aftermath of snowstorms as in regime III) and then remain at a constant level, despite decreasing humidity, as new primary emissions are added to the brown cloud for several days of the episode?

REGIME III, The Aftermath of Snowstorms. Snowstorms are a major source of moisture for the Colorado Front Range during the winter as moist air from the eastern plains moves upslope toward the foothills of the Rocky Mountains. Vigorous winds usually result in a rapid ventilation of the Denver area. However, after such storms, rapid evaporation of moisture from streets and other urban features usually occurs in the presence of relatively strong temperature inversions and light winds, resulting in a moist boundary layer underlying dry air aloft; such was the case on 29 December 1987 following a heavy snowstorm on 26 to 27 December. [On this day, Denver's haze, which had a white appearance, was confined to the 200-m height of the buildings in the downtown area ( Figure 3c, taken at 1600 MST). Extensive snow cover still remained.] The BCI found major contributions from vegetative burning, mobile sources, and secondary aerosols. The relative humidity above the haze was less than 30%, but within the haze it ranged from 90% to 100% at night, to 80% in the morning, to 50% in the afternoon, as measured at the surface at Auraria. There was no documentation of whether fog developed at night.

2.1.3 Network Requirements

Based on these hypotheses and conceptual models of different air quality regimes the following requirements should apply to the meteorological network that supports the winter NFRAQS:

• Documentation of the presence of clouds, fogs, and high boundary-layer relative humidity.
• Characterization of the potential for vertical mixing at various sites throughout the South Platte airshed.
• Documentation of boundary-layer and aloft winds throughout the domain with higher resolution in low-lying areas such as Welby and Greeley where shallow nocturnal flows may provide a significant transport mechanism but not be well resolved in diagnostic or prognostic models.
• Documentation of synoptic conditions that underlie episodes within different portions of the NFRAQS study area.
• Ability to trace the history of aerosol-laden air masses, particularly during periods of recirculation over mesoscale domains.
• Documentation of the micrometeorology of high-aerosol concentration regions such as those in the Welby-Adams City area and other areas in low-lying terrain within the South Platte air basin.

As described below, a number of instruments are available with which to document the meteorological conditions associated with NFRAQS pollution episodes. For the most part these instruments can document winds (profilers and Doppler sodars), temperature profiles (RASS), and near-surface winds, temperature, and humidity (weather stations). Clouds can be identified using satellite images, ceilometers, radar reflectivity profiles, as well as photographic or video devices. Mixing processes can be identified using sodar and radar reflectivity data. Humidity is also a central issue in the study but is not a parameter that can be remotely sensed easily at the present time. However, surface measurements give an indication of the water vapor mixing ratio in the boundary layer. The BAO tower will produce routine continuous profiles whereas the NWS sounding will give a twice a day sounding. Alternative methods include airsondes and tethered balloon systems. However, they may not be possible in the new Terminal Control Area for DIA and add a significant cost component to the experiment for labor and expendables.

2.2 Outline of the Meteorological Network

2.2.1 Rationale Underlying the Choice of Observation Sites

Because of the limited resources available to the study, we used the following guidelines to locate meteorological sites:

1. Locate instruments at Core Chemistry Sites so as to characterize local winds and boundary-layer behavior.
2. Locate instruments near major sources outside of Denver such as those near Greeley and Fort Morgan.
3. Locate instruments throughout the South Platte River basin so as to characterize regional wind, temperature, and moisture regimes so as to test, extend and refine existing conceptual models for interpreting Northern Front Range air quality.

At a minimum, this network will be able to provide the data necessary to i) classify periods dominated by ground-based, elevated, or mixed source types, ii) identify significant local source influences if any, iii) document the role of local microclimates on production of high fine-particle concentrations, and iv) define regional transport and recirculation regimes.

2.2.2 Existing Network Data

Three primary surface networks spanning the South Platte River basin will be most useful in support of Front Range studies. These include the NOAA MESONET, a network of 21 weather stations recording wind, temperature, RH, solar and net radiation (available for the Winter-1996 and Summer-1996 studies only), the CDPHE network, an array of 10 weather stations recording wind and temperature but not RH, and the Colorado Department of Transportation network. Other weather-station data may be available but should only be acquired if a specific analysis requirement is identified and resources made available to provide for the ingest of these data. The NOAA MESONET will only operate through September 1996. However, discussions are underway to form a regional observing cooperative. Although these data will be available for the early 1996 and summer 1996 field programs, the availability of surface data for the winter study requires serious consideration. However, the combination of CDPHE and ETL surface stations will provide a minimum network of 16 to 20 surface sites. If RH measurements are added to the CDPHE sites, the network should be adequate for the analyses that will be carried out in 1997.

Upper-air data are available from several sources. These include the NWS rawinsonde at Stapleton and the Wind Profiler Demonstration Network (WPDN) 404-MHz profiler east of Platteville (this profiler has a minimum range of 500-700 m, well above the surface inversion). Other profilers include the 404-MHz profiler operated by CSU north of Greeley. The 915-MHz profiler operated by NOAA/ETL at Stapleton has been decommissioned because of Federal budget cuts.

2.2.3 ETL Resources

The Front Range meteorological network operated by ETL has evolved during 1996 as resources and time have become available. This process started in January 1996 as ETL profilers with RASS were installed at the Brighton Ridge site and Front Range Airport. At the same time Radian Corporation was installing and debugging a new multi-beam profiler at the BAO purchased by ETL for an EPA-sponsored project. This later project includes the integration of a Doppler sodar and ceilometer with the profiler to provide winds and mixed-layer depth input into a plume dispersion model under development for the EPA in support of the Strategic Environmental Research and Development Program (SERDP) of DOD. This system will be ready for full field testing during the 1996-1997 winter experiment. At the beginning of March a profiler was installed at Fort Morgan. In addition, an ETL profiler at Granby continued to operate through the summer of 1996. Additional data will be available from the BAO 300-m meteorological tower just east of Erie, along I-25. Four levels of wind, temperature, and RH (10, 50, 100, and 200m) were available by late June 1996. During July we installed an experimental profiler system at Elitch's along the South Platte near Confluence Park: This system is being deployed to test new radar clutter removal algorithms that will allow such instruments to operate in dense urban environments. Data from this system were available during the summer experimental period. The remainder of ETL's profiler, meteorological tower, and sodar network is now deployed on the U.S. west coast in a study of coastal meteorology. When this experiment ends in the October some of these instruments will be returned to Colorado and may be able to support the winter phase of the NFRAQS.

2.2.4 Instrumentation Summary

The suite of ETL instruments available for the NFRAQS Study Periods is described below with potential siting indicated in Figure 4 and described in more detail in Section 2.3. The minimum number of instruments for each phase is indicated in brackets: [Jan-Feb, 1996; Summer, 1996, Winter, 1996-97]. Data from the preliminary winter and summer studies is being provided as an in-kind contribution on a time-available basis. The instruments specified for the winter 1996-1997 study is the minimum number committed to NFRAQS. If other instruments are available, they will be deployed as suggested in Section 2.3 below.

2.3 Site-Selection and Field Deployment

Instruments have been in the process of deployment through 1996. Sites for the 1996-97 winter NFRAQS experiment are identified below. A subset of these sites operated for the summer 1996 phase of NFRAQS. All core sites and supplemental sites would operate through at least 15 February 1997. Contingency sites, if installed, would operate through the end of December and then be moved for possible programs in Colorado Springs in mid-January and/or Mexico City in mid-February.

Core Sites:

Downtown Site (ELH): A site in the Auraria campus area was used during the BCI study and included a monostatic sodar, Doppler sodar, tethersonde/airsonde system, and a surface weather station. The Auraria site is not suitable for profiler/RASS operation so an alternate site has been located at Elitch's as indicated on Figure 4. The Elitch's site is removed from the immediate blocking effects of high-rise buildings in downtown Denver. A wind profiler and 10-m meteorological tower was installed at this location during July 1996. There is a possibility that CDPHE will locate their new Doppler sodar at Auraria. Figure 5 provides a detailed view of the downtown Denver area relative to the Welby, Adams City, and CAMP monitoring sites as well as the elevated transmissometer path from the Federal Building to Cheeseman Park. It should be noted that the Elitch's and Auraria surface sites lie near the confluence of the South Platte River and Cherry Creek whereas Welby and Adams City lie just northeast of the confluence of the South Platte River and Clear Creek. [Operating]

Adams City Site (ADC): The Welby CDPHE site, just across the South Platte from Adams City and slightly lower in elevation, included a Doppler sodar and a monostatic sodar during BCI and provided the primary meteorological characterizations for the BCI chemical analyses. The same instruments are proposed for the 1996-1997 study. ETL will make available a 20-m meteorological tower with wind speed, direction, temperature and sigma theta measured at 2 and 18 m using 20-min averages. The 20-min averaging is required for satellite transmission. In addition we will install separate instruments at 10 m with wind speed, direction, temperature, and RH measured at 1-min averages. A high-resolution laser ceilometer will be located at this site to document the presence of clouds as well as the depth of shallow aerosol layers.

Brighton Ridge Site (BRI): During the DBCS the Brighton water-tower site (BWT) included a Doppler sodar, 3-m weather station and a rawinsonde sounding unit. Because of the proximity to the approach to DIA, it is unlikely that balloon soundings will be allowed at this site. A radar wind profiler with RASS and 10-m tower now operates at the water treatment facility site just to the south of the water tank. This site has been operating since January 1996. It is anticipated that the Brighton Core Monitoring site will be located at the water tank facility. If this is the case we will locate a monostatic sodar and 3-m meteorological tower at the core chemistry location also. [Operating]

Greeley (La Salle-Evans area) (GLY): We are proposing a site, collocated with a core chemistry site, to monitor vertical-mixing and transport processes to and/or from the Greeley area along the South Platte River. The site would include a wind profiler together with monostatic sodar and a 10-m meteorological tower. Past work near Greeley by Andre Erasmus of the Department of Earth Sciences at the University of Northern Colorado has suggested a shallow (<100m) drainage following the Cache la Poudre underlying a deeper flow from the south-southwest along the axis of the South Platte.

Supplemental Sites

Front Range Airport (FRA): This site lies east of DIA along I-70 and includes a wind profiler with RASS and a 10-m tower. This site has been operating since January 1996 and followed a recommendation of the peer-review committee for a site on the southeast side of the South Platte drainage. [Operating]

Fort Morgan (FTM): This site is on the eastern boundary of the study area and complements the Granby site on the western "upwind" side of the domain. It is also near a major elevated source. A site was found just south of Ft. Morgan for installation of a wind profiler with RASS, 10-m meteorological station. This site has been operating since early March 1996. [Operating]

Boulder Atmospheric Observatory (BAO): This site includes the SERDP system under development for the EPA. It includes a wind profiler with an integrated RASS/Doppler sodar system, a ceilometer, and surface flux measurements; the 300-m tower itself will be instrumented with Temp, RH, and wind at four levels and a sonic anemometer at a minimum of one level. [Operating]

We are expecting that the Greeley and Longmont (see below) sites would be established by DRI and available to ETL without additional costs for power and phone. Installation for the winter study would be scheduled for late October 1996 following the return of equipment from California: these sites would thus need to be selected and acquired earlier than others. Additional sites (depending on the availability of equipment) that may be possible include:

Contingency Sites (in-kind, if instrumentation is available)

Figure 4

Figure 4 GIF Image

Figure 5

Figure 5 GIF Image

Longmont/I25 area (LGM): This site will include a radar wind profiler and a 10-m tower: this was a satellite chemistry site for the summer program and may be also for the winter field program.

Granby(GBY): Our profiler located at Granby would be kept in its current location so as to characterize winds over the mountains to the west of the study domain. [Operating]

Boulder (BOU): This profiler would be located east of Research Laboratory Number 3 on the University of Colorado East Campus next to Boulder Creek. When not being used for profiler development efforts, it would routinely collect data for NFRAQS.

Notes:

1. The Profiler is located at Elitches just north of Auraria. CDPHE is anticipating locating their Doppler sodar at the actual Auraria site.

2. The Profiler is located at the Water Treatment Facility (WTF) whereas the actual sampling will be done at the Brighton Water Tower (WT) site that ETL used during the 1987-1988 DBCS.

3. This site also includes a 300-m tower with four levels (50, 100, 150, and 200 m wind, temperature, and RH). Plans for November include the addition of wind, temperature, and relative humidity at 10 and 300 m, turbulence measurements at 100, 150, and 200 m, and a videocam at 250 m.

4. Contingency Sites if other ETL profilers are available.

5. The Platteville 404 mHz Profiler is one the National Demonstration Network whereas the CSU profiler is used for educational activities and may or may not be available.

6. The 10-m meteorological towers include solar radiation as well as wind speed and direction at 10 m, temperature and relative humidity at 2 m with net radiation, and surface pressure. Existing towers at Elitch's, Brighton, and Front Range Airport have relative humidity and temperature at both 2 and 10 m.

7. The 20-m tower at Adams City would have two levels of wind speed, direction, and temperature with 20-min averaging at 2 and 18 m. A third level at 10 m would have 1-min averages of wind speed, direction, temperature, and relative humidity.

8. A 3-m tower would be used to verify differences, if any, from the 10-m tower at the profiler site 0.5 miles to the south.

2.4 Field Program Schedule

The following graph summarizes field operation schedules for NFRAQS. Instrumentation is as noted in Table IV.

Field Program Notes:

1. Equipment will be returned from another experiment at the end of October. It will be most effective if sites have been acquired by late October so that we can off load equipment directly at field sites.

2. Adams City will require additional time to optimize equipment and cross-calibrate sensors with Welby so as to compare the micrometeorology of the two site.

2.5 Supporting Data Products

Documentation of cloud features and snow cover in eastern Colorado will be important at times during the experimental periods. Satellite data routinely available include GOES-8 and 9 visible and infrared images for Colorado from the CSU/NOAA CIRA WWW site:

http://www.cira.colostate.edu/earthsta/overview.htm

and via FTP anonymous access at

merak.cira.colostate.edu/jpeg/ (GOES-9), and

spica.cira.colostate.edu/jpeg/ (GOES-8)

The archive includes high-resolution visible images of Colorado including county boundaries. In addition, NOAA's Forecast Systems Laboratory (http://www.fsl.noaa.gov) operates a local analysis and prediction system (LAPS) for the Colorado Front Range. Currently, ETL profiler data are being fed to this data assimilation system and gridded analyses have been available to ETL beginning in late summer of 1996. The gridded data outputs are sent to ETL on an hourly basis. ETL will analyze these data in a separate in-house effort to assess the value of profiler data ingest into the FSL LAPS. However, the results of our work, if successful, will provide an independent evaluation of the diagnostic wind field analyses to be prepared by STI. It should be noted that real-time graphical output is also available at the above FSL Web address for experiment guidance and interpretation.

In addition to the data provided by ETL, DRI (Winter Proposal, Section 3.9.3.4) will collect data from other agencies such as the National Weather Service, NOAA's MESONET (through September 1996), and the Colorado Department of Transportation for the NFRAQS database at DRI.

2.6 Data Management Issues

The data sets for which NOAA/ETL will be responsible include:

It is anticipated that other data sets such as satellite images that are directly available via internet FTP access will be directly downloaded by DRI or STI. Profiler wind, temperature, and meteorological station data were transferred into our FTP-anonymous account on an hourly basis for access by STI and others during the 1996 summer field phase. Doppler sodar and BAO tower data files will be added for the major winter field program. Profiler moment data, sodar reflectivity profiles, and ceilometer files will be collected from the field sites every two-to-four weeks and after archival at ETL, will be made available via anonymous FTP. This procedure will allow the earliest access possible to the data for the NFRAQS research team. Standard data formats for ETL data depend on its source. Formats include NETCDF, ASCII, and Spyglass Special ASCII (used for visualization of time-height data displays). In the past we have converted profiler wind and temperature data to Xbase compatible formats although we do not consider this the most efficient format for analysis purposes.

2.6.1 Profiler Edited Data

Wind profilers and Doppler sodars typically produce hourly wind speed and direction profiles with data at discrete heights. The primary outputs include:

The post processed data represents level 1b editing. These data will be made available via anonymous FTP within 2 months following the retrieval of the moment data from field sites.

2.6.2 Other Data Sets

Ceilometers and monostatic sodars produce a high density of data that is best viewed in time-height cross sections of light scattering and acoustic reflectivity, respectively. The ceilometer time-height cross sections can be used to deduce cloud base as well as aerosol layer thickness. The monostatic sodar produces data that can be interpreted to produce several boundary layer characterizations:

Data from the Welby site will be analyzed for comparison with the 1987-1988 DBCS and extend the results of past ETL analyses. Sodar data from other sites will be used primarily for case study analyses.

2.6.3 Quality Assurance

Meteorological Tower Data:

ETL's meteorological tower instruments undergo routine calibration either in factory facilities or at our BAO facility between experimental periods. During field operations and in the course of data analysis these data are routinely compared with other data sets for consistency.

Profiler Wind and Temperature Data

Profiler data are quality assured through several processes. One source of error is in the siting of the instruments and input of parameters such as antenna tilt angles and magnetic declination into the computer software. These errors are identified usually as part of our standard operating procedures or through data audits in the field. A second source of error is the effect of malfunctions during operations: these usually result in sudden changes in profiler performance such as range coverage, wind direction, or failure to reach consensus within each hourly average. These errors are detected through routine real-time monitoring of data on a daily basis. After data collection, raw data from the profiler is retrieved and reprocessed at the moment level and compared with the original consensus data acquired in real time. Data are then assessed for consistency with the original data set and with supporting data sets such as meteorological towers or adjacent Doppler sodars. Because profilers often have ranges of several kilometers above ground level, their data can be compared with other profilers aloft where greater homogeneity is expected to prevail as shown by Neff (1994). Finally we expect to obtain an independent data set from a profiler located near Platteville and operated at a joint NOAA-University of Colorado field site which will be held aside for comparison with both LAPS data and diagnostic wind fields produced by STI.

Sodar Wind Data

Sodar Doppler wind data are treated similarly to that for the profilers insofar as both systems operate on the same principle. Differences lie primarily in sample volume resolution as described by Neff (1994).

Sodar and Profiler Reflectivity Data

Sodar and profiler reflectivity data provide a time-height picture of the atmospheric boundary layer and the mixing processes within it. The ability to interpret these data is limited primarily by the sensitivity of the instrument, background noise level, and for sodars, the degree of isolation from noise sources and scattering volume resolution. Quality of the data is gained from inspection of records in the field and adjustment of radar and sodar parameters as necessary. Techniques for interpreting the data are well established, provided the initial data quality is assured (Neff, 1990,1994).

3. Deliverables and Time Schedule

We have been working through 1996 to develop the observational network critical for the winter NFRAQS field experiment period. Data from sites already in place were available in near-real time for the summer study. These include the Granby, Elitch's, Brighton Ridge, Front Range Airport, and Fort Morgan wind profilers. Winter data from the Brighton Ridge and Front Range Airport are now available in ASCII daily files via anonymous FTP. Because of the limited direct funding available from the NFRAQS, the deliverables outlined below address only the winter 1996-1997 NFRAQS field program. We will continue to cooperate in the analysis of winter 1996 and summer 1996 cases as we have in the past. The NOAA/ETL effort involves the following tasks:

3.1 Network Operation

The primary operation of the network will occur from 25 November through 15 February although many sites will be operating prior to and following the official experimental period. Following the winter field program a Meteorological Operations Report will be prepared describing the data sets acquired, operations summaries, and quality assurance activities. The draft report will be completed by 1 June 1997 with a final report completed by 1 October 1997.

3.2 ETL Data Set Delivery

3.3 Analyses

NOAA/ETL's analysis effort will be carried out in close coordination with DRI/STI as was done in the 1987-1988 DBCS effort with DRI. Our analysis effort will fall into two parts. The first will focus on the broad goals of Section 1 and the hypotheses posed in Section 2. The second effort will address the air quality meteorology of the South Platte River basin outside of the immediate Denver metropolitan area where no combined meteorology-air quality measurements have been made in the past.

Task 1. Assessment of the current conceptual model for the

Denver metropolitan area

This task will examine results from the 1996-1997 field program with measurements made in 1987-1988, January-February 1996, and in the context of work published by NOAA/ETL. The analysis and classification procedures, based on Welby data will be repeated for the 1996-1997 data and compared with the summaries provided for 1987-1988. Both single day and multiday episodes will be examined in the context of the DBCS Final Report and subsequent published analyses. The initial effort will use the following data sets:

1. Meteorological time series and vertical profile data from Adams City, Welby, and Auraria/Etitch's,
2. Continuous light-scattering, light-absorption, and gas data from Adams City, Welby, and Auraria,
3. Total fine-particle mass data from fixed-time-interval filter data at Adams City and Auraria, and
4. Photographic documentation,
5. Fine-particle mass, visibility, and meteorological data bases from the 1987-1988 DBCS.

Because these data sets will be available quickly after they are collected, an initial assessment should be available by late March if the above data sets are available by 15 February. A draft report will be prepared by 1 June 1997 with an assessment of the current conceptual model and recommendations for modifications, if any. A final report will be delivered by 1 November 1997.

Task 2. Examine the applicability of the current conceptual model to

other sites outside of the Denver Metropolitan area.

This task follows naturally from Task 1 but remains a significant research task because of the lack of previous simultaneous, analyzed observations outside of Denver. This task will depend on

1. The same data as in Task 1 but for the remaining sites, particularly combined nephelometer/aethalometer data at a minimum of three additional sites,
2. Results from ETL's in-kind analysis of LAPS data at the surface, 110 m and 500 m, and
3. Surface, 700 hPa, and 500 hPa weather maps.

An initial draft report from this effort will be available by 20 July 1997 with the draft final report by 15 November 1997. Initial guidance for this effort will come from published or soon-to-be-published work based on the 1987-1988 DBCS and the 1991 Winter Validation Study at Rocky Flats.

Task 3. Assess the role of mesoscale circulations in the South Platte River basin

on local air quality and develop a basin-wide conceptual model.

This task will evolve from the analyses in Tasks 1 and 2 and be carried out in close coordination with DRI and STI. However, the absence of past integrated meteorological and chemical data sets outside of the Denver area make this a challenging task with some level of risk associated with it. We anticipate that fine particle mass analyses will be available by the end of June 1997. To this end we need to identify the major circulation regimes affecting air quality, as measured by PM2.5, and aethelometer and nephelometer measurements, as quickly as possible and then develop candidate classification parameters that are amenable to data base inclusion by the end of June. We propose the following approach:

• A series of candidate circulation regimes will be identified during the experimental period from daily inspection of weather maps, satellite imagery, and LAPS graphical output together with discussion among study PIs and the CDPHE forecaster via e-mail.
• Following the field program these candidate regimes will be evaluated using the preliminary data gathered during the study including the photographic documentation of episodes.
• These regimes will then be formulated for inclusion in a data base prior to receipt of analyzed fine-particulate data from DRI.
• We will then examine the degree to which high concentrations of total and speciated fine-particle mass, light-scattering and light-absorption data stratify by regime, if at all, as a test of the regime classification scheme.

Progress in this task will depend also on progress made by DRI in the filter analyses as well as by STI in their diagnostic wind field and chemistry analyses. However, efforts in this area will be reported in our formal quarterly progress reports to CSU as well as by intermittent e-mail correspondence copied to the project manager at CSU. Independent of efforts by others, we will provide a draft final report on NOAA/ETL analyses by 15 December 1997.

Task 3 will also benefit from the additional analyses available from the work of Dr. Pirim Kaufmann who is a post-doctoral fellow visiting ETL under sponsorship of the Swiss National Fund through 1997. Dr. Kaufmann's work (Weber and Kaufmann, 1995; Kaufmann, 1996; Kaufmann and Weber, 1996) addresses the classification of wind systems in complex terrain and will focus in particular on classification of wind systems in the South Platte River basin using the profiler and surface network deployed in the NFRAQS area by ETL.

3.4 Reports

In addition to the above task reporting, summary quarterly reports will also be submitted to CSU.

4. Program Management

This effort proposed will be supported by the staff of the Meteorological Applications and Assessment Division of NOAA's Environmental Technology Laboratory. Overall project management and the analysis effort will be led by Dr. William Neff, Chief of the Division. The field program will be managed by Mr. Clark King while the data management portion of the program will reside under Mr. Brian Templeman. The ETL team has a demonstrated competence in this area of research as authors or co-authors on 70% of the papers referenced in the Bidder's Bibliography provided by NFRAQS and with experience in major field programs in Colorado, California, the southeastern U.S., at sea, and in both north and south polar regions of the earth that has involved deployment at well over 100 wind profiler sites and produced extensive analysis of the resulting data sets.