3.3 CLIMATOLOGY of Ozone Synoptic scale Transport in the EASTERN US
Bret Schichtel and Rudolf Husar
Washington University St. Louis Missouri
Richard Poirot and Paul Wishinski
Vermont Department of Environmental Conservation
Presented at the American Meteorological Society's 78th annual meeting, Jan. 11-16 1998, Phoenix AZ
OTAG (Ozone Transport Assessment Group) was a 2-year project to develop consensus solutions to the tropospheric ozone problems in the Eastern US. A central issue of OTAG was assessing the extent of regional scale ozone transport and its contribution to non attainment areas. To aid in this endeavor forward and backward trajectory analyses were conducted over 5-7 summers between 1989 - 1995 in the Eastern US (Poirot and Wishinski, 1995; Poirot and Wishinski, 1996a,b,c; Schichtel and Husar, 1996a; Schichtel and Husar, 1997). These analyses were used to identify regional scale transport directions, speeds, and potential pollutant transport pathways during high and low ozone conditions, as well as the OTAG modeling episodes. While these techniques and meteorological data drivers are able to characterize regional scale transport, they are poorly suited for, and not intended for identification of local flows or near-by source influences.
2.0 Meteorological Data
The analyses used trajectories calculated from the National Meteorological Centers Nested Grid Model (NGM) (Rolph, 1992). The NGM database, contains three dimension wind vectors with a horizontal resolution of ~160 km and ten vertical layers up to seven kilometers. The data are at a two hour time step from 1989 – 1995.
3.0 Trajectory analysis methods
3.1 Transport Wind Vectors and Source Regions of Influence
Using the CAPITA Monte Carlo Model (Schichtel and Husar, 1996b) five summers, 1991- 1995, of forward plumes were calculated for each NGM grid cell over the Eastern US. A forward plume identifies the downwind location of mass that was continuously released from a source at an instance in time. It can be though of as a snapshot of a plume from a smoke stack. The plumes were sorted based upon locally high and low ozone days, the 90th and 10th percentile of the average daily maximum ozone for each source region, i.e. each NGM grid cell, respectively. Each plume subset was aggregated together assuming a pollutant lifetime of 1 or 2 days and a constant emission rate of 1 g/s to create transfer matrices, average relative source contributions to each NGM grid cell.
Transport wind vectors and source regions of Influence (SRI) were calculated from the transfer matrices (Schichtel and Husar, 1996; Schichtel and Husar, 1997). The source region of influence is the smallest region around the source encompassing 63% of the ambient mass. The transport wind vectors is a vector pointing from the source to the centroid of the SRI. It is a measure of the resultant mass transport direction and speed from the source. For example, assuming constant speeds, a source with transport predominately in one direction will have a longer transport vector than a source with transport almost equally likely in all directions.
3.2 Residence Time Analysis
Seven summers, June – August, 1989-95 of backward trajectories for 23 ozone monitoring sites across the OTAG region were calculated using the HY-SPLIT model (Draxler, 1992) and NGM winds. The trajectory results were aggregated over 1440 80*80 km squares using an ozone concentration based residence time analysis (Poirot and Wishinski, 1995; Poirot and Wishinski, 1996a,b,c). For each receptor location, an "Every Day Probability Field" was calculated using all summer trajectories. The trajectory "residence time hours" were summed for each of the 1440 grid squares, and then expressed as a probability by dividing the square's hours by the total hours in all squares. The trajectories were then sorted according to whether the resulting ozone was "low" (below median) or "high" (above median). Additional trajectory subsets were developed for "very high" ozone levels - as defined as the upper 90th and 95th percentile of ozone concentrations at the receptor. For each of these trajectory subsets, the "everyday probability field" was subtracted from the "ozone-sorted probability field". The resulting "incremental probability" showed locations more likely than normal to be upwind prior to the specified ozone level.
4.1 High and Low Ozone Transport Characteristics
The Transport wind vectors and SRIs during the locally high ozone days are presented in Figure 1a. The high ozone in the South and the Midwest were associated with short and/or meandering transport wind vectors, and smaller SRIs. However, in the northern and western part of the domain there was stronger more persistent transport from the southwesterly and southerly directions respectively. These results suggest that the high ozone concentration in the South and Midwest result from near field transport, while in the north and westerly section the transport conditions were conducive to regional scale transport. Similar results were found from the incremental probability analysis, where the incremental probability areas increased from the south to the north during high ozone periods (Figure 2). This indicates that in the north airmasses traveled longer distances in route to the receptor in a given period of time than in the south.
It was also found that the incremental probabilities for many individual receptors had unique transport pathways associated with elevated ozone concentrations, >50th percentile (Figure 3). For example elevated ozone at Mark Twain, MO, was associated with transport from the south and east; from the west and north at Lithia Springs, GA; and from the south and southwest at Bennington VT. However, when the incremental probabilities from multiple receptors were grouped together (Figure 4a) the Central Eastern US became a common region airmass traversed in route to receptors throughout the Eastern US. This is also evident in the transport wind vectors during locally high ozone days (Figure 1a), where the locations surrounding the central industrial region have transport wind vectors pointing outward from it. In contrast, both the transport wind vectors (Figure 1b) and incremental probability plots (Figure 4b) show that low ozone concentrations are associated with swift transport from outside the (i.e., Canada, Atlantic, and Gulf of Mexico) into the Eastern US.
4.1 OTAG Modeling Episode Transport Characteristics
Using the transport wind vectors and source regions of influences, the transport conditions during the 1991, 1993, and 1995 OTAG episodes were characterized. It was found that a broad region of stagnation existed for each episode. The south had stagnant conditions during all three episodes, while it extended further north into the Midwest for the ’91 and ’95 episodes. The average OTAG episode transport, with the exception of the Southeast, east of Texas, was characteristic of region-wide high ozone concentrations, with slow meandering transport over Kentucky, Tennessee, and West Virginia and strong clockwise transport around this region of stagnation (Figure 5a). Where region-wide high ozone concentrations were the 90th percentile of the daily maximum ozone averaged over the entire OTAG domain. However, the OTAG episode transport differed from that during locally high ozone concentrations. The transport speeds were higher during the OTAG modeling episodes and the resultant direction of transport over the Illinois-Pennsylvania corridor was from the west, rather than the typical southwesterly flow (Figure 5b). In the Southeast, the average transport was characteristic of the slow meandering transport during the locally high ozone days (Figure 5b).
These long-term, trajectory-based results suggest that the central Midwest is an important source region for regional-scale transport of moderately high ozone levels to surrounding receptor areas in many directions. This regional transport results in elevated background concentrations which, combined with local-scale influences (not investigated here), can contribute to exceedances of ozone standards. This feature of a central source region affecting receptors in many directions is characteristic of locally high ozone periods. It is not likely to be observed from the OTAG modeling episodes which had transport characteristic of regionally high ozone periods.
Draxler R.R. 1992. Hybrid single-particle Lagrangian integrated trajectories (HY-SPLIT): Version 3.0 - user's guide and model description. NOAA Technical Memorandum ERL ARL-195. Air Resources Laboratory, Silver Spring Maryland.
Poirot, R. and P. Wishinski. 1995, Air trajectory residence time analysis investigation of ozone transport pathways: 1989-95 (http://capita.wustl.edu/otag/ Reports/VTTRAJ5/StatusReport1.html)
Poirot, R. and P. Wishinski. 1996a, VT DEC Air Trajectory Analysis of Long-Term Ozone Climatology, Status Report to OTAG Air Quality Analysis Workgroup: 8/15/96 (http://capita.wustl.edu/otag/Reports/VTTRAJ5/ StatusReport1.html)
Poirot, R. and P. Wishinski. 1996b, VT DEC air trajectory analysis of long-term ozone climatology, Status Report to OTAG Air Quality Analysis Workgroup: 11/7/96 (http://capita.wustl.edu/otag/reports/ Status_Dec96/Status_Dec96.html)
Poirot, R. and P. Wishinski. 1996c, VT DEC Air Trajectory Analysis of Long-Term Ozone Climatology, Status Report to OTAG Air Quality Analysis Workgroup: 12/3/96 (http://capita.wustl.edu/otag/reports/ Status_Dec96/Status_Dec96.html)
Rolph G.D. 1992. NGM Archive. NCDC Report TD-6140, July, National Climatic Data Center.
Schichtel, B. and Husar, R. 1996a. Source regions of influence for high and low ozone conditions in the Eastern US. (http://capita.wustl.edu/otag/reports/sri/ sri_hlo3.htm)
Schichtel, B.A. and Husar, R.B. 1996b. Regional simulation of atmospheric pollutants with the CAPITA Monte Carlo model. J. Air & Waste Manage. Assoc. 47, 331-343.
Schichtel, B.A. and Husar, R.B. 1997. Update on the Characterization of Transport over the Eastern US.http://capita.wustl.edu/Otag/Reports/Sricont/Sricont.html
Figure 1. Transport wind vectors and source regions of influences during high (90th percentile) local ozone days (a) and low (10th percentile) local ozone days (b).
Figure 2. Incremental probability plots for Gregg Co., Texas, Boone County, Kentucky, and Port Clyde, Maine, from left to right, for the 90th percentile of ozone concentrations.
Figure 3. Incremental probability plots for Mark Twain, Missouri, Lithia Springs Georgia, Bennington Vermont, from left to right, for the upper 50th percentile of ozone concentrations.
Figure 4 a) Frequency of back trajectories for 22 receptor sites for low ozone days (lower 50th percentiles) (a) and high ozone days (upper 50th percentiles) (b).
Figure 5. Comparison of transport winds during the 1991, '93, and '95 OTAG domain episodes and the regionally high (a) and locally high (b) ozone conditions.