The compatibility between aircraft and ground-based air quality measurements

. Trace gas concentrations and atmospheric state parameters were measured aboard the NOAA King Air research aircraft during flights on August 16, 1988, along the Appalachian Mountains from central Pennsylvania to northern Georgia. Stepwise profiles were flown over five surface sites where measurements of certain atmospheric parameters were being made. A stationary cold front lying across southern Virginia effectively divided the area into two weather regimes; to the north of the frontal zone the air was slightly cooler and much drier than that to the south. convective activity developed from early to midafternoon along and south of the front. The comparison between the aircraft and ground sites included measurements of the primary pollutants SO 2 and several hydrocarbons, NOy, the secondary pollutants H202 and 03, and meteorological parameters. Continuity between the aircraft and surface meteorological and trace gas measurements was consistent at the northernmost site, which is situated in a relatively level valley. The agreement was poorer at the other four ground sites, which are located on or near mountaintops. Most of the meteorological and trace gas measurements, other than those made at Scotia, were found to differ by substantial margins, often by more than 10 times the resolution of the instruments. However, within a few hours after the flights, the surface measurements, particularly those of HgOg and 03, achieved values comparable to those measured with the aircraft, thus suggesting that air sampled at the canopy level did not mix readily with the bulk of the boundary layer.


Introduction
Since ground-level and aircraft measurements of meteorological parameters and atmospheric trace gas concentrations are often used concurrently to understand the variability and distribution of the trace gases, it is essential to determine the comparability of these measurements.
Although ground-level measuring sites can produce long-term continuous records of meteorological parameters and local concentrations of trace gases, the micrometeorology in the vicinity of a site influences the ability of measurements at the site to faithfully represent the wider atmospheric zone.
On the other hand, instrumented aircraft can provide measurements over a wide area and thus provide three dimensional pictures of the distributions of the trace gases, and of the complementing meteorological parameters as well, but the length in the other group of cases, when the times of the aircraft overflight are illustrated atmosphere was poorly mixed, or when the on a map of the eastern United States (Figure 1). measurement location was on the edge of an urban Some geographic features in the vicinity of each plume, the difference was large (averaging 49 of the surface sites, along with the flight ppb), with the higher O 3 concentration being tracks at the lowest altitudes above each of the almost always measured aloft.
sites, and at altitudes below the elevations of In an attempt to compare aircraft and ground the WHT and MTM sites, are shown in Figure 2. air quality measurements, four stepwise profile The instruments aboard the aircraft were flights were flown with the NOAA King Air described by Boatman et al. [1988,1989] Table 4, and the concentrations of above SCO, SHN, WHT, MTM, and BRB, respectively. some hydrocarbon trace gases are listed in Table   Flight and ground sampling information from the 5. Only data collected during the time of lowest step above the site, and in two cases at constant altitude flight were used for the an additional step below the elevation of the aircraft-ground site comparisons. For the site, are shown in Table 1   In the vicinity of the front minutes for the ground sites was deemed necessary a transition zone was identified, where the because the sampling scale on the ground is much surface flows (below the frontal boundary) were longer as a result of the differences between wind speed (2-5 m s -1) and aircraft velocity (60 m s -1) . Thus, the surface equivalent averaging time should be an order of magnitude longer than for the aircraft. These trajectories (shown in Figure 3) indicated that the airflows at sites Two-dimensional plots of potential south of the front were from the southeast, while temperature (e) and water mixing ratios (WMR) are those north of the front were primarily from the shown in Figures 4 and 5. Similar plots for 03,      Thereafter, later in the day, the highest the frontal region were significantly greater concentrations were recorded at the lowest than those to the north, especially at low altitudes, which is consistent with local altitudes, as expected based on the higher WMR photochemical production.

Meteorology
At This is a region of atmospheric stability, as indicated by the closer spacing of the 8 isopleths (Figure 4). This H20a maximum is consistent with the report of Ray et al. [1992] showing that HaOa accumulates near the top of the boundary layer.
However, it should be pointed out that 03 levels are not elevated in this region.
Ha02 concentrations (Figure 7)  In an NO-lean atmosphere there will be a buildup (>1.0 ppbv) of SO 2 and NOy were measured even at of HO 2 radicals that can promote the recombination the highest flight altitudes (Figures 8 and 9).
It is likely that these levels of SO 2 and NOy were HO2 + HO2 ---> H202 + 02. to some extent related to vertical transport through orographic lifting and convective The latter reaction proceeds faster in the activity, but the observed concentrations at presence of water [Atkinson and Lloyd, 1984]; cruise altitude were more uniformly distributed this reaction becomes dominant for removal of HO 2 through the southern sector than were O 3 and H202 in the NO-lean atmosphere. This above scheme concentrations. This suggests that part of the oversimplifies the atmospheric process. NO under SO 2 and NOy present at this altitude was some circumstances (lack of hydrocarbons and/or transported from a more distant source region.

solar radiation) scavenges O 3 by NO + 0 3 ---> NO 2 + 0 2 .
Thus, there is a wide range of circumstances under which either one or both 0 3 and 1-1202 can be present in high, or low, concentrations.
One process by which H202 is removed from the The air mass back trajectories ending at WHT and BRB, indicated in Figure 3, suggest the Midwest industrialized region as a probable source. It is noteworthy that the contours of SO 2 and NOy are similar but not identical, which suggests that the major regional pollution sources are stationary and emit both substances, rather than automotive, which is nearly exclusively a source atmosphere is solution in cloud water' therefore, of NOy. At the SCO can only be guessed, such as increased dry site, which is situated in relatively level deposition under the influence of sunlight and a terrain (Figure 2), the agreement between the degree of decoupling of the planetary boundary aircraft and surface gas analyses was reasonably layer from the air actually in contact with good (Table 3 (Figure 2), the meteorological and gas analysis data from the lower altitude traverses were closer to those from the higher altitude traverses than to the ground site data (Tables 2  and 3).
The fact that only after 1-3 hours were the data at WHT more consistent with the aircraft data suggests that the stirring from convective activity did not involve near-surface air for some time.
Although the true microstate of the weeks); therefore, transport from distant regions is an important factor in their atmospheric concentrations. As can be seen from the air mass back-trajectories of Figure 3 and the data in Table 5, there is a consistency between atmospheric transport path and hydrocarbon concentrations.
The air masses sampled at some sites passed over highly industrialized areas (SHN and WHT) and some were from rural areas (MTM and BRB) or where the air mass being sampled passed quickly over industrial zones (SCO). Two of the hydrocarbons, C3H s and n-C4H10, are the primary constituents of liquid petroleum gas (LPG) that is widely used in rural areas; in some cases leakage from LPG tanks is significant. LPG atmosphere is unknown, these observations seem to from different vendors can have widely varying indicate that the atmosphere on a very small ratios of C3H s and n-C4H•0, so little can be vertical scale, i.e., on the canopy scale, was expected regarding the apparent lack of separate from the well-mixed boundary layer. correlation between these two substances in the This is also consistent with the observation that aircraft versus the ground samples, where it took a few hours for the very near-surface air significant local influence was probable. The to come into equilibrium with the upper layers.
other two hydrocarbons, C2H 2 and C2H 6, are less As atmospheric mixing progressed in the afternoon likely to have major local point sources, hours these differences seemed to diminish. therefore, they represent regional pollution An example of midday atmospheric variability levels. That is probably why the variation is provided by the various ground stations within between ground and aircraft samples, and to a a small area at SHN. At 1300 EST the hourly certain extent between the sites, is smaller. average 03 concentrations recorded at these various sites (Table 4) differed by as much as 19 6. Conclusions ppbv, but all exhibited similar patterns, which involved strong increases in 03 concentrations The results of the comparison between the from early morning until about 1300 EST, followed aircraft and ground-based atmospheric by sharp decreases until about 1600 EST. The measurements reveal that under the atmospheric difference between the highest and lowest conditions prevailing on the day of the flights concentrations of 03 in the 3-hour period (1000-from HAR to BRB, the disagreement exceeded the 1300 EST) in the SHN area, 38 ppbv, is far limits of experimental error. The conclusion was greater than the 23-ppbv difference found between based on measurements of state parameters for the aircraft and the SH1 site. To the extent which the uncertainty is small, and is supported that the aircraft-measured 03 concentrations can by the measurements of the atmospheric pollutant be compared with the hourly averaged data from species. This suggests that measurements at the Big Meadows site, which is a predominantly these rugged surface sites may not represent the open, grassy area, the agreement is excellent (94 planetary boundary layer for certain periods of ñ 2 ppbv aircraft at 1106-1110 EST versus 90 ppbv time, such as the time of the present study. average from 1000 to 1100 EST and 99 ppbv from 1100 to 1200 EST' see Table 4