109
Chapter 4
Direct comparison of airborne and ground based air quality
data
Factors influencing the mixing in the troposphere, which in turn
influences the agreement between airborne and surface monitored
air quality data will be considered in this chapter. Direct
comparison of airborne and ground based air quality data will be
made.
Challenges of comparing airborne and surface monitored air quality data
The comparability of airborne air quality data to surface measurements is dependent on
the extent of mixing of air in the troposphere. The mixing is in turn influenced by
meteorological conditions, atmospheric lifetime of air pollutants, sources distribution and
the height at which the pollutants are released into the troposphere
(Annegarn et al., 1996a; Luke et al., 1998).
Influence of the diurnal evolution of the mixing layer on air dispersion
The diurnal evolution of the mixing layer plays an important part in the agreement
between airborne and ground based measured air quality data (Luke et al., 1998). In the
morning the surface emissions are trapped below a shallow mixing layer caused by the
nocturnal near ground level inversion, and emissions from tall industrial stacks are
prevented from mixing with the air below this inversion. The consequence of this
nocturnal near ground level inversion is a decoupled lower and upper troposphere, which
in turn lead to a vertical gradient or discontinuity in air pollutants concentration
distribution. About mid-day the surface inversion is mixed out and the mixing layer is
deep enough to allow air aloft to mix with surface air. This leads to a deep mixing layer
with air homogeneously mixed in vertical up to high altitudes (Turner, 1996; Luke et al.,
1998), a favourable condition to compare the two data sets (Luke et al., 1998).
110
Table 4.1 shows the data that were collected in the morning over Secunda during the
Highveld autumn campaign. The data were collected on 18/03/2005 approximately at
167 magl, 333 magl and 667 magl flight levels, from 10:35:00 to 12:37:45 (SAST). Data
from industrial plume penetration incidences are included in this data set. On 18/03/2005
Secunda was under the influence of surface trough as is shown on Figure 3.1(c). In the
morning the surface was capped by a nocturnal ground level inversion at 92 magl (Figure
3.2(a)), which was mixed out in the afternoon resulting a deep mixing layer which was
capped by an inversion at 2775 magl (Figure 3.2(b)).
The uneven vertical distribution of air pollutants levels in the morning is clearly evident
in Table 4.1. The air pollutants were not well mixed within the column that was
monitored. The SO2 and NOX average concentrations were relatively lower at 167 magl,
and relatively higher at 333 magl and 667 magl and their maximum values indicate
penetration of industrial plumes from the Secunda tall industrial stacks (Freiman and
Picketh, 2002). O3 was only slightly higher at 167 magl than at the two higher flight
levels because of destruction by NO (Kley et al., 1994; Poulida et al., 1994; Hobbs et
al., 2003, Taubman et al., 2004) from tall industrial stacks. This uneven concentration
distribution in the vertical can lead to disagreements between airborne and surface air
quality measurements.
111
Table 4.1: Air pollutants levels over Secunda at different heights in the morning
Pollutant
Min Conc
(ppb)
Max Conc
(ppb)
Avg Conc
(ppb)
StdDev %
Height (magl)
O3 16.37 86.201 34.506 39.196 167
NOX 0 8.262 2.518 72.681 167
SO2 2.605 65.775 19.352 77.65 167
O3 8.679 77.037 31.409 49.284 333
NOX 0.186 24.443 10.188 54.871 333
SO2 3.895 157.724 63.733 64.232 333
O3 13.151 64.588 32.415 32.083 667
NOX 0 16.811 8.644 49.179 667
SO2 3.555 131.753 55.449 64.87 667
Table 4.2 shows the data that were collected in the afternoon over the Vaal Triangle
during the Highveld autumn campaign. The data were collected on 17/03/2005
approximately at 167 magl and 333 magl flight levels, from 15:17:00 to 16:53:30
(SAST). Data from industrial plume penetration incidences are included in this data set.
Like in the Secunda case study the Vaal Triangle was also under the influence of the
surface trough (Figure 3.1(b)). In the morning the surface was also capped by a nocturnal
inversion at 276 magl (figure 3.2(a)), which was mixed out in the afternoon resulting a
deep unstable mixing layer (Figure 3.2(b)).
The uniform mixing in vertical in air pollutants levels in the afternoon can be clearly seen
in Table 4.2. The air pollutants average concentrations were comparable at both
monitoring flight levels, though their maximum values especially of SO2 indicates
penetration of industrial emissions from the Vaal Triangle tall industrial stacks. This
homogeneity in the vertical in the afternoon favours the comparison of the airborne and
surface air quality measurements (Luke et al., 1998).
112
Table 4.2: Air pollutants levels over the Vaal Triangle at different heights in the
afternoon
Pollutant
Min Conc
(ppb)
Max Conc
(ppb)
Avg Conc
(ppb)
StdDev %
Height (magl)
O3 37.866 47.928 42.505 5.442 167
NOX 0 0 0 0 167
SO2 4.91 14.05 6.972 23.232 167
O3 35.428 47.372 41.2113 5.916336 333
NOX 0.006 3.96 1.994271 57.51993 333
SO2 3.339 62.084 8.694359 106.3663 333
Spatial variation of air pollutants
Figures 4.1 to 4.3 show concentration frequency distribution of SO2, NOX, and O3 over
Secunda. This data is part of the data that was collected over Secunda in the morning on
18/03/2005 during the autumn campaign. The data was collected at approximately 167
magl flight level.
The uneven distribution of SO2 concentration in space can be clearly seen in
Figure 4.1. SO2 concentration ranges between 0-6 ppb, with 28% frequency of
occurrence. It was followed by the 18.1-24 ppb concentration range, with 18% frequency
of occurrence. The remaining concentration ranges had a frequency of occurrence of
+/- 10%. The concentration ranges of higher concentration had the least frequency of
occurrence. This uneven spatial distribution of SO2 could lead to disagreements between
airborne and surface air quality data (Luke et al., 1998).
113
Figure 4.1: SO2 concentration frequency distribution over Secunda at approximately
167 magl during the autumn campaign.
The uneven distribution of NOX concentration in space can also be seen in Figure 4.2.
The majority of the NOX data were lying within the concentration range 1.1-2 ppb, with
17% frequency of occurrence. It was followed by the 0-1 ppb and 8.1-9 ppb
concentration ranges, both with 9% frequency of occurrences. Then the 2.1-3 ppb and
3.1-4 ppb concentration ranges, both with 8% frequency of occurrences. The remaining
concentration ranges had a frequency of occurrence that was less than 6%. This spatial
variation of NOX concentration can also lead to poor comparison of airborne and surface
air quality data (Luke et al., 1998).
114
Figure 4.2: NOX concentration frequency distribution over Secunda at approximately
167 magl during the autumn campaign.
In comparison to SO2 and NOX in Figure 4.1 and 4.2 respectively, Figure 4.3 shows that
O3 concentration varied the least in space. O3 is a relatively long lived air pollutant, hence
it has more time to mix and become uniformly distributed in space (Luke et al., 1998;
WMO, 2006). Most of the O3 concentration was within the concentration range 32.1-40
ppb, with 37.5% frequency of occurrence. It was followed by the 24.1-32 ppb
concentration range, with 24% frequency of occurrence. Then the 16.1-24 ppb
concentration range, with 20.6% frequency of occurrence. The remaining concentration
ranges of higher concentrations had frequency of occurrences of about 2%. The relative
uniform distribution of O3 in space in comparison with SO2 and NOX is also confirmed
by the relative standard deviation of these air pollutants at 167 magl flight level
(Table 4.1). The relative uniform distribution of O3 in space is a favourable condition for
comparing airborne and surface measurements of this pollutant (Luke et al., 1998).
115
Figure 4.3: O3 concentration frequency distribution over Secunda at approximately 167
magl during the autumn campaign.
Air pollution source height levels
Figure 4.4 shows the data that was collected in the morning along the boundaries of the
Vaal Triangle during the air pollution flux provincial cross boundary campaign. The data
was collected on 31/03/2006 during vertical profile flights of up to 3 km altitude. Irene
weather observation station midday upper air data was also used to complement the
temperature vertical profile from the aircraft. The data generated from this campaign is
used to show the vertical uneven distribution of pollutants caused by the release of fresh
emissions at different heights.
The temperature vertical profiles in Figure 4.4(a) and Figure 4.4(b) from both platforms;
the aircraft and balloon-borne radiosonde, show an unstable lower column of the
troposphere. However the SO2 concentration vertical profiles in Figure 4.4(a) and Figure
4.4(b) show two bands of SO2 plumes at different altitudes. Figure 4.4(a) shows 48.4 ppb
and 26 ppb SO2 concentration peaks at 1720 masl and 2000 masl respectively.
Figure 4.4(b) shows 27.6 ppb and 5.4 ppb SO2 concentration peaks at 1864 masl and
2688 masl respectively. This vertical gradient in SO2 concentration distribution is caused
by fresh emissions of air pollutants at different heights (Luke et al., 1998). Both the
116
Vanderbijlpark and Denesyville SO2 vertical profiles have a peak close to the surface and
another at a higher altitude. This variation of SO2 concentration in the vertical leads to
poor comparison of airborne and surface air quality data for this pollutant (Luke et al.,
1998).
Figure 4.4: SO2 and temperature vertical profiles. The dotted line on both Figures 4.4(a)
and 4.4(b) are temperature profiles measured over Irene weather station and the SO2 and
the other temperature profiles are measured from the aircraft. Figure 4.4(a) is a vertical
profile over Vanderbijlpark, Figure 4.4(b) is a vertical profile over Denesyville.
117
Direct comparison of airborne against ground based air quality data
The Figures 4.5 and 4.6 show the comparisons of SO2 and O3 of ground based data
against airborne data. The ground based data are a one hour averaged data from Sasol and
Mittal Steel air pollution monitoring sites. Airborne data were collected in a vicinity of
these ground based air pollution monitoring sites. The airborne data measurement is
instantaneous and the data were collected when the aircraft was flying within a 20 km
radius from the ground based monitoring sites. Because of the unavailability of high
resolution temporal ground based data. Airborne data were compared with an hour
average and the month average of that specific hour that correspond to a time the aircraft
flew within 20 Km radius from a ground station. The variability of the ground based data
was determined by calculating the monthly standard deviations, using a specific hour that
correspond to a time the aircraft flew within 20 Km radius from a ground station. The
observations were made during the autumn and winter field campaigns. Table 4.3 shows
the times and the altitudes at which the air pollutants in Figure 4.5 and 4.6 were
monitored by both platforms.
The comparisons between airborne and surface hourly measurements of SO2 in Figure
4.5 are not as close as the comparisons of O3 measurements from the two monitoring
platforms which are almost exact (Figure 4.6). The difference between the comparisons
of airborne and ground based measurements of SO2 and O3 can be explained by a number
of factors. To begin with the two data sets are being averaged over different temporal
scales. The ground based data is averaged over an hour and the airborne data is an
instantaneous data (averaged over a second). The uneven spread of SO2 sources in space
over the study sites (Wells, 1996) cause an uneven spatial distribution of SO2
concentrations. This spatial variation of SO2 levels is established in Figure 4.1. Relative
standard deviations in Table 4.1 and Table 4.2 also show that SO2 is more variable in
space than O3 at all flight levels monitored. The emission of SO2 at different heights over
the study sites (Wells, 1996) creates a vertical concentration gradient in their vertical
distribution (Luke et al., 1998). Figure 4.4(a) and Figure 4.4(b) show this vertical
concentration gradient caused by fresh emissions from different source heights, even
118
though the lower troposphere was unstable and well mixed. The relatively short
atmospheric lifetime of SO2 (one week) in comparison with O3 (28 days) (Luke et al.,
1998; Seinfeld and Pandis, 2006) results in smaller spatial extent of higher concentrations
of SO2 (Annegarn et al., 1996a), which in turn leads to spatial variation of SO2. The
vertical and horizontal variation of SO2 caused by the above mentioned factors, reduces
the extent of agreement between airborne and surface monitored air quality data.
Figure 4.5 show that the hourly SO2 ground station data at Langverwatch and Bojesspriut
is twice the levels monitored by the aircraft. The hourly SO2 ground based data at Opsis
350 and Opsis 620 are relatively comparable to aircraft data. The monthly standard
deviations for all ground stations show that SO2 is temporarily highly variable. The
standard deviations for SO2 airborne data are small and close to zero. This is because the
data extracted within 20 Km radius from ground stations are few values measured within
short distances and times. Data measured within a short distance away from sources can
be relatively uniform.
Figure 4.5: Direct comparison of airborne and ground based measured SO2 data.
119
The comparisons between airborne and surface measured O3 in Figure 4.6 show some
relative good agreements. O3 is a secondary air pollutant with a relatively long
atmospheric lifetime (28 days). This characteristic of O3 affords it more opportunity to be
uniformly distributed in space as compared to short lived pollutants like SO2 and NOX
(Luke et al., 1998). The smaller variation in space of O3 can be seen in Figure 4.3.
Relative standard deviations of O3 in Table 4.1 and Table 4.2 derived from morning and
afternoon monitoring respectively, show that O3 is more uniformly distributed in space
and the comparable O3 average concentrations at different flight levels suggests
uniformity in vertical as well. This relative uniformity in space of O3 explains the good
comparison of the two data sets. Monthly standard deviations for Leitrim ground station
show that O3 is temporarily less variable. This explains the relatively good agreements
between the O3 hourly ground data and instantaneous airborne data. Standard deviations
for airborne data were small and zero at some instances. This is due to the same reason
already given in the case of SO2 airborne data.
Figure 4.6: Direct comparison of airborne and ground based measured O3 data.
120
Table 4.3: The times and altitudes at which SO2 and O3 were monitored by the aircraft
and ground air quality monitoring stations.
Date Time
(SAST)
Campaign Monitoring
platform
Altitude
(m)
Pollutant
17/03/2005 17:00:00 Autumn Opsis 350 (Mittal Steel) Ground SO2
17/03/2005 16:44:40 Autumn Opsis 350 (airborne) 1735.71 SO2
17/03/2005 16:00:00 Autumn Opsis 620 (Mittal Steel) Ground SO2
17/03/2005 15:58:37 Autumn Opsis 620 (airborne) 1673.86 SO2
18/03/2005 11:00:00 Autumn Langverwatch (Sasol) Ground SO2
18/03/2005 10:28:00 Autumn Langverwatch (airborne) 1604.96 SO2
18/03/2005 13:00:00 Autumn Bosjesspruit (Sasol) Ground SO2
18/03/2005 12:46:00 Autumn Bosjesspruit (airborne) 2385.28 SO2
21/07/2005 15:00:00 Winter Leitrim (Sasol) Ground O3
21/07/2005 14:41:30 Winter Leitrim (airborne) 1270.00 O3
25/07/2005 14:00:00 Winter Leitrim (Sasol) Ground O3
25/07/2005 14:17:00 Winter Leitrim (airborne) 1632.18 O3
03/08/2005 15:00:00 Winter Leitrim (Sasol) Ground O3
03/08/2005 14:42:34 Winter Leitrim (airborne) 1673.50 O3
************************************
Factors influencing mixing in the troposphere, which in turn
influences the agreement between airborne and surface monitored
air quality data, were considered in this chapter. The diurnal
evolution of the mixing layer plays an important role in the
agreement between airborne and surface air quality data. Uneven
spatial distribution of air pollutants sources, with different
temporal emission cycles can lead to disagreements between
airborne and surface air quality data. The atmospheric lifetime of
air pollutants also complicates the comparison of the two data sets.
O3 an air pollutant with a relatively long atmospheric lifetime gives
good comparison of the two data sets, and SO2 a highly variable
pollutant in space and time with a short atmospheric lifetime gives
less agreements between the two data set.