Click
Here
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 113, D17315, doi:10.1029/2007JD009745, 2008
for
Full
Article
An analysis of the vertical structure of the atmosphere and the upperlevel meteorology and their impact on surface ozone levels in
Houston, Texas
Bernhard Rappenglück,1 Ryan Perna,1,2 Shiyuan Zhong,3 and Gary A. Morris4
Received 20 December 2007; revised 14 April 2008; accepted 26 June 2008; published 13 September 2008.
[1] Despite emission reductions, Houston continues to be designated as a nonattainment
area for ozone (O3) by the Environmental Protection Agency. Upper-level synoptic
maps and information about the vertical structure of the lower troposphere obtained by in
situ measurements were analyzed to characterize ozone exceedances in which peak
8-h average concentration exceeded 85 ppb during the Texas Air Quality Study-II in
August–September 2006. Cluster analysis of meteorological conditions showed that the
highest background surface O3 concentrations occurred under northerly or easterly
flow regimes at 850 hPa, coinciding with the advection of dry continental air. Exceedance
days in September 2006 occurred almost exclusively in postfrontal environments.
These frontal passages are associated with shifts in wind direction and may lead to
increases in background O3 from 30 ppbv (marine) to 60–70 ppbv (continental)
throughout the lower troposphere. Several factors are identified to be important for 8-h
average ozone peaks in Houston under well-developed land-sea-bay breeze conditions,
including (1) the presence of easterly winds advecting industrial emissions from the
Ship Channel, and (2) the presence of persistent large-scale northerly flows aloft advecting
elevated continental background ozone levels that are eventually entrained into lower
layers through the growth of the convective planetary boundary layer.
Citation: Rappenglück, B., R. Perna, S. Zhong, and G. A. Morris (2008), An analysis of the vertical structure of the atmosphere and
the upper-level meteorology and their impact on surface ozone levels in Houston, Texas, J. Geophys. Res., 113, D17315, doi:10.1029/
2007JD009745.
1. Introduction
[2] The Houston-Galveston-Brazoria (HGB) region, located
close to the Gulf of Mexico, is one of the largest metropolitan areas in the United States located in a subtropical region
with extended hot and humid periods and intense solar
radiation. The HGB is well-known for enhanced ozone
levels under proper meteorological or upset emissions
conditions [Kleinman et al., 2002; Reiss, 2006]. An abundance of highly reactive volatile organic compounds
(VOCs) and nitrogen oxides (NOx) are produced in the
HGB, leading to rapid ozone production [e.g., Kleinman et
al., 2002; Ryerson et al., 2003; Daum et al., 2004].
[3] The Texas Commission on Environmental Quality
(TCEQ) produces a daily air quality forecasts to warn the
general public, and especially those who are sensitive to
1
Department of Geosciences, University of Houston, Houston, Texas,
USA.
2
Now at Regional Office, Texas Commission on Environmental
Quality, Houston, Texas, USA.
3
Department of Geography, Michigan State University, East Lansing,
Michigan, USA.
4
Department of Physics and Astronomy, Valparaiso University,
Valparaiso, Indiana, USA.
Copyright 2008 by the American Geophysical Union.
0148-0227/08/2007JD009745$09.00
such ozone pollution, when an ozone exceedance is
expected. In 2006 the 8-h ozone standard as specified by
the United States Environmental Protection Agency (EPA)
was effectively 84 ppbv (http://www.epa.gov/air/ozonepollution/pdfs/2008_03_factsheet.pdf). In this paper an ‘‘ozone
exceedance’’ day is defined as by TCEQ (www.tceq.state.
tx.us/compliance/monitoring/air/monops/sigevents06.html),
i.e., when the measured 8-h average ozone mixing ratio at
any given station in the HGB was greater than or equal to
85 ppbv.
[4] Many of Houston’s ozone exceedances result from
emissions originating from the petrochemical industrial
region of Houston in and around the Ship Channel to the
east of the downtown area, including intense sporadic upset
emissions. However, these emissions are not the sole
determining factor of Houston’s ozone exceedances. Meteorological conditions also play an important role. Although
Houston is known to experience ozone exceedances nearly
year-round, most of these exceedances occur within the
months of August and September [Nielsen-Gammon et al.,
2005b]. With high temperatures, abundant sunshine, weak
winds, and a higher frequency of winds from the north in
September [Nielsen-Gammon et al., 2005a], background
ozone in Houston tends to increase this time of year. In
addition, the frequent dominance of an anticyclonic regime
helps to inhibit cloud formation and limit shower develop-
D17315
1 of 11
D17315
RAPPENGLÜCK ET AL.: ATMOSPHERIC STRUCTURE AND O3 IN HOUSTON
D17315
(Texas Air Quality Study-II) field campaign during the
months of August and September 2006.
2. Methods and Data Used
Figure 1. Location of rawinsonde and ozonesonde site on
University of Houston main campus and Continuous
Ambient Monitoring Site stations used to determine 8-h
background ozone: (1) Conroe, (2) northwest Harris
County, (3) Westhollow, (4) Croquet, (5) University of
Houston main campus, (6) Galveston, and (7) Crosby. In
addition the location of the Ship Channel is indicated. Color
coding reflect some major land use types: water bodies
(blue), forests (green), urbanized area (light red), and
densely urbanized area (red). No color indicates rural areas
with shrub-like vegetation.
ment; the former leading to increased radiation for ozone
production and the latter decreasing the ability of the
atmosphere to wash out the pollution. Both factors increase
the probability of an ozone exceedance.
[5] Another meteorological factor that may be important
in air quality forecasting is the height of the planetary
boundary layer (PBL). Berman et al. [1999] and Angevine
et al. [2002] found that a lower PBL height in the morning
hours can be responsible for concentrating ozone precursors
closer to the surface, increasing the probability that an
exceedance will be reached. In addition, Alonso et al.
[2000] suggested that a subsidence inversion may have an
indirect effect on raising ozone levels: ozone can become
trapped underneath this elevated inversion and can mix
downward into the boundary layer. Subsidence can also
result in a downward mixing of ozone from the residual
layer left over from the previous day [Zhang and Rao,
1999].
[6] The influence of the Gulf of Mexico on Houston’s
ozone exceedances also must be considered. Southerly
winds typically bring in marine air with lower precursor
and ozone background levels [Banta et al., 2005]. However,
there are situations in which the sea breeze circulation can
recirculate pollutants into the HGB after having been
advected out to sea by the land breeze [Banta et al., 2005].
[7] The purpose of this study is to better characterize the
meteorological conditions that favor high ozone levels in
the HGB through analyses of data from the TexAQS-II
[8] Synoptic weather patterns play a significant role in
determining surface ozone concentrations [Yu and Pielke,
1986; Banta et al., 2005; Nielsen-Gammon et al., 2005a]. In
order to address the common patterns for ozone exceedances in August and September 2006, composite maps were
created for all exceedance days and all nonexceedance days
using the Web interface (http://www.cdc.noaa.gov/Composites/
Day/) hosted by the Climate Analysis Branch of the
Physical Sciences Division of NOAA’s Earth System Research Laboratory. Such maps may allow us to classify
synoptic weather conditions for past ozone exceedances and
therefore better forecast future ones. Several pressure levels
were explored, but the 850 hPa pressure level turned out to
be the most relevant in our study. It is sufficiently far aloft
from surface friction impacts and still low enough to
provide spatially resolved synoptic patterns for the study
region. The 850 hPa maps were placed in ‘‘clusters’’ based
upon the classifications of Ngan and Byun [2007]. These
clusters were then related to background ozone levels, i.e.,
8-h peak ozone levels as determined using surface ozone
data from the Continuous Ambient Monitoring Site
(CAMS) network. Previous cluster analysis studies in the
HGB either related the synoptic conditions at the 850 hPa
pressure level with 1-h maximum ozone values [Davis et
al., 1998] or examined the impact of surface winds on
maximum hourly ozone levels in Houston and addressed
diurnal cycles of surface wind clusters [Darby, 2005].
August 2006 saw no frontal passages through the HGB
before 29 August. However, September 2006 was a month
of regular frontal passages. These frontal passages will be
investigated in more detail in order to identify the conditions that may raise Houston’s background ozone levels.
[9] The second part of the study addresses the local
meteorological conditions associated with Houston’s ozone
exceedances based on frequent rawinsonde soundings from
the University of Houston main campus (29.74°N;
95.34°W; 11 m AGL) approximately 5 km to the southeast
of downtown Houston (location shown in Figure 1). The
atmospheric sounding profiles obtained during August and
September 2006 provide insights into the PBL height and its
evolution as well as the effects of the surface-based radiation inversion and the elevated subsidence inversions on
Houston’s ozone concentration.
[10] Remote sensing instruments such as lidars and radar
wind profilers which permit continuous observations
throughout the day have been used in previous studies to
measure the characteristics of the boundary layer, both
within and outside of the HGB [Angevine et al., 1998;
Banta et al., 2005; Eresmaa et al., 2006; Nielsen-Gammon
et al., 2008]. Under meteorological conditions favorable for
high ozone events, the wind profiler and lidar data compare
well with rawinsonde data [Marsik et al., 1995; NielsenGammon et al., 2008]. Many boundary layer studies that
use remote sensing techniques to determine the PBL height
use rawinsonde data as a standard for comparison [Marsik et
al., 1995; Grimsdell and Angevine, 1998; Berman et al.,
1999; Eresmaa et al., 2006; Nielsen-Gammon et al., 2008].
2 of 11
RAPPENGLÜCK ET AL.: ATMOSPHERIC STRUCTURE AND O3 IN HOUSTON
D17315
D17315
Table 1. Cluster Frequency During Ozone Exceedances and Nonexceedance Days for All 61 Days in August
and September 2006a
Exceedance days
Nonexceedance days
Average background
ozone (ppb)
Cluster I
Cluster II
Cluster III
Cluster IV
Cluster V
Cluster VI
5
10
52.4
6
6
41.6
0
9
31.0
1
12
20.6
0
12
28.0
0
0
n/a
a
Last row indicates the background ozone mixing ratios associated with the different clusters. For definitions of
Cluster I – VI, see text.
[11] Rawinsondes were launched twice per day at 0700
and 1900 CDT (central daylight time) from 01 August to
30 September 2006, corresponding to the standard launches
completed across the country by the National Weather
Service (1200 GMT and 2400 GMT, respectively). Intensive
observational periods (IOPs) were conducted on 10 days
with high ozone forecasts (8-h averages greater than
85 ppbv) during the 2-month field campaign. For these
IOPs, additional rawinsondes were launched at 0500, 1000,
1300, 1600, and 2200 CDT in order to capture the detailed
evolution of the PBL growth. The soundings used RS-92
GPS sondes (available at http://www.vaisala.com) and 100
g balloons with an average ascent rate of 5 m s 1 and
burst altitudes of 14 – 18 km. Data used in this study
comprised the original data set at 1 Hz sampling rate.
[12] Wiegner et al. [2006] showed that temperature and
water vapor mixing ratio profiles can be used to help
identify the top of the boundary layer. For the morning
soundings, the height of the inversion was used to determine
the boundary layer height [Eresmaa et al., 2006]. The PBL
height in the mornings was defined (through visual inspection) as the first level where temperature began to decrease
with height. For the remaining soundings, the PBL height
was determined in a subjective manner by finding the point
at which a sharp increase in potential temperature [Senff et
al., 2002; Yi et al., 2001], and a sharp decrease in water
vapor mixing ratio is observed [Nielsen-Gammon et al.,
2008].
[13] On 31 days in August and September 2006, ozonesondes were launched at 1300 CDT in order to capture the
vertical distribution of ozone when the boundary layer is fully
developed. Ozonesondes were also launched at 0700 CDT
on 8 days with high ozone forecasts to examine the vertical
ozone profile before mixing took place and to capture the
preexisting residual layer(s).
[14] Ozonesondes consisted of GPS enabled En-Sci Model
2Z electrochemical concentration cell (ECC) [EN-SCI
Corporation, 1997, 2003a, 2003b] coupled with an RS
80– 15N rawinsonde (available at http://www.vaisala.com).
Ozonesondes typically reached 20 – 30 km before burst,
ascending at a rate of 5 m s 1. The ozonesonde program
in Houston is described by Morris et al. [2006].
[15] The ECC-type ozonesondes have been characterized
thoroughly over the last decade [Russell et al., 1998; Smit et
al., 2007] and intercomparison studies, among them the
Jülich Ozone Sonde Intercomparison Experiment (JOSIE),
showed that ECC-type ozonesondes had better accuracy and
precision than the Brewer Mast and KC79 sondes [Smit and
Kley, 1996]. Intercomparisons with other ozone measuring
instruments have demonstrated the ECC sonde precision to
be ±6% near the ground and – 7% to +17% in the upper
troposphere [Kerr et al., 1994; Komhyr et al., 1995, Reid et
al., 1996].
3. Results and Discussion
3.1. Classification of Synoptic Patterns for Ozone
Exceedances
[16] To better define the weather conditions that create
elevated ozone levels in the HGB, 850 hPa maps were
visually inspected, analyzed, and placed in clusters based
upon a classification scheme developed by Ngan and Byun
[2007] using data from the May – August periods of 2005
and 2006. The classifications are as follows: Cluster I is
characterized by northerly flow with a high-pressure system
to the west of Houston (the ‘‘postfrontal’’ category); Cluster
II is characterized by easterly flow with a high-pressure
system to the north; Cluster III is characterized by southeasterly flow with a high-pressure system to the northeast;
Cluster IV is characterized by southerly flow from the Gulf
with a high-pressure system to the east of Houston; Cluster
V is characterized by southwesterly flow with a highpressure system in the Gulf; and Cluster VI occurs when
Houston is under the direct influence of a tropical storm
(which did not happen in 2006).
[17] Table 1 illustrates the frequency of the clusters for
ozone exceedance and nonexceedance days. The data in
Table 1 demonstrate that Clusters I and II are the most
common scenarios for ozone exceedances in August and
September 2006. The majority of Cluster I days are postfrontal and often precede an ozone exceedance. In fact,
many ozone exceedances in September 2006 occurred in a
postfrontal environment with a light northerly flow that
opposes the inland propagation of clean marine air with the
sea breeze and/or brings with it elevated continental
background levels. The nonexceedance days in this cluster
were related to cold fronts with strong northerly winds
(about 8 m s 1) that suppress ozone concentrations by
blowing pollutants away from the HGB. The precipitation
that frequently accompanies such frontal passages cleans
out pollutants, further decreasing ozone levels.
[18] While an equal number of exceedance and nonexceedance days can be found in Cluster II, 50% of all
exceedance days are found in Cluster II and only 12% of
all nonexceedance days are in Cluster II.
[19] In September, Cluster II was common after a frontal
passage as high pressure developed to the north of the HGB.
Cluster II conditions can result in elevated ozone levels due
to winds carrying precursors from the Ship Channel region
mixed with elevated background ozone (see below) from
upwind of the Ship Channel region.
3 of 11
D17315
RAPPENGLÜCK ET AL.: ATMOSPHERIC STRUCTURE AND O3 IN HOUSTON
Table 2. Dates of Cold Frontal Passages in Relation to Ozone
Exceedances in 2006
Frontal Passage
Exceedance
29 Aug
5 Sep
12 Sep
18 Sep
24 Sep
28 Sep
31 Aug, 1 Sep, 3 Sep
7 Sep
13 Sep, 14 Sep
20 Sep
26 Sep, 27 Sep
none
[20] The cluster analysis is applied to classify synoptic
patterns associated with high background ozone concentrations as determined using surface ozone data from the
CAMS network. Several CAMS stations on the perimeter
of the network far from power plants or chemical refineries
and with limited effects of NO titration were chosen as
representative sites following Nielsen-Gammon et al.
[2005a] (shown in Figure 1). One additional station was
added on the northeast side of Houston (CAMS site
‘‘Crosby’’). The background ozone was defined as the
minimum of the daily maximum 8-h mean ozone values
among all background stations following Nielsen-Gammon
et al. [2005a]. This value usually occurs in the CAMS
station that is upwind of the HGB [Nielsen-Gammon et al.,
2005a]. The same six cluster classifications previously
discussed were also applied in the postfrontal environment
to determine the effect of the background concentration on
Houston.
[21] As indicated in Table 1, Cluster I has the highest
background values and is associated with a high-pressure
system to the west of Houston, bringing in northerly flow
from the continental United States. This cluster is very
common in the wake of a cold front. Cluster II also has high
background values typical of easterly flow coming into the
HGB. Being associated with anticyclonic flows this cluster
often exhibited a northerly flow component at an earlier
stage that would have brought continental air to Houston
and thus increased the background ozone level. Back
trajectory analyses for the ozone exceedances of September
2006 indicated that all trajectories on Cluster I and II days
came from either northern Texas or northeast Louisiana. On
Cluster I and II days, therefore, Houston was under the
influence of a continental air mass.
3.2. Role of Frontal Passages
[22] In September 2006 nearly every ozone exceedance
occurred in the wake of a frontal passage. This is quite
different from the preceding year 2005. Because of a very
active tropical season, the first cold front in 2005 did not
pass through Houston until late September. The tropical
activity in 2006 was far less than that of the 2005 season
[Lau and Kim, 2007], with the first major frontal passage on
29 August 2006, 2 days before the days with the poorest air
quality index of the year (www.tceq.state.tx.us/compliance/
monitoring/air/monops/sigevents06.html). The dates of each
cold frontal passage in 2006 are shown in Table 2, along
with the dates of the ozone exceedances.
[23] It is well known that high pressure favors the buildup
of ozone in the PBL over a timescale of a couple of days
due to stagnant air that is repeatedly exposed to solar
D17315
radiation under clear sky conditions. However, the Houston
case is unique since frontal passages that precede these
high-pressure systems are associated with a significant
change in background ozone levels (as seen in Table 1),
providing elevated ozone levels which local and regional
photochemistry further enhance. Elevated background
ozone results from the transport of continental air and/or
subsidence occurring behind the front. A detailed example
will be provided in the discussion of the 28 August to
1 September case study below.
[24] The one exception found in Table 2 is for the frontal
passage of 28 September, which was not followed by an
ozone exceedance. On the day following this frontal passage, there was a quick return to southeasterly flow due to
the relatively high speed with which this high pressure
system advanced, unlike the previous frontal passages that
did lead to exceedances.
3.3. Effect of the PBL Height
[25] Twice daily rawinsonde launches were performed for
the purposes of determining the relation between the PBL
heights and the ozone peak. The strength of the morning
inversion was captured well from the 0700 CDT soundings.
The daytime development of the PBL height was captured
well on IOP days when five additional soundings were
launched (see above).
[26] Berman et al. [1999] hypothesized that the morning
inversion height and strength has an effect on ozone
precursor concentrations and may affect hourly ozone peaks
during the day. This process occurs during a typical morning rush hour, when mixing layer heights are low and ozone
precursors are being emitted rapidly. Stronger early morning
inversions can slow the subsequent development of the
mixed layer in the midmorning hours and lead to higher
observed ozone levels in the afternoon [Berman et al.,
1999]. Higher ozone concentrations can also occur by a
downward mixing of the air from the residual layer in the
midmorning hours as the morning inversion breaks up and
the mixed layer grows into the residual layer [Zhang and
Rao, 1999]. However, statistical analysis of the TexAQS II
rawinsonde data set reveals very little correlation between
morning inversion strength (i.e., the temperature difference
between the top and bottom of the inversion level).
[27] Figure 2 shows the average rate of PBL development
on exceedance versus nonexceedance days. The results
closely match those found in the work of Berman et al.
[1999]. The development of the PBL in the morning may be
slower on ozone exceedance days due to the fact that most
of the exceedance days occurred in a postfrontal environment with relatively cool morning temperatures that delayed
boundary layer development. However, this explanation
remains speculative, since differences in PBL height between exceedance and nonexceedance days are not significant, and the temporal resolutions of the rawinsonde
measurements are usually too coarse to capture the mixed
layer growth. Other meteorological factors also influence
ozone concentrations, including the degree of stagnation,
the intensity of solar radiation, the wind direction, the
presence/absence of precipitation, the air temperature, and
the photochemical history, and the rate of freshly entrained
precursor emissions. Variations of the PBL height may only
play a minor role.
4 of 11
D17315
RAPPENGLÜCK ET AL.: ATMOSPHERIC STRUCTURE AND O3 IN HOUSTON
D17315
Figure 2. The average growth of the boundary layer on ozone exceedance days and nonexceedance
days. Times given in central daylight time (CDT). Nonexceedance day data points are offset 1/2 h for
clarity of presentation.
3.4. Role of Elevated Inversions
[28] In a postfrontal environment, the soundings on ozone
exceedance days had several features in common. One of
these features was a strong elevated inversion, significantly
limiting the growth of the mixed layer. Since an elevated
inversion was present during most of the high ozone days
(especially in September), an attempt was made to analyze
the different characteristics of the air masses associated with
these inversions.
[29] Elevated inversions for each sounding at 0700 CDT
and 1900 CDT were categorized by the strength of inversion
(i.e., the amount of temperature increase), moisture, and
height to find the possible role that elevated inversions play
on Houston’s 8 h ozone peak. Inversions less than 0.3°C
were ignored. The results of the 0700 CDT analysis are
shown in Figure 3. The strongest elevated inversions are not
correlated with the highest ozone values. However, the dry
air masses aloft (low relative humidity) are associated with
frontal passages that occurred in the days preceding ozone
exceedances. Figure 3 indicates that the highest ozone
averages occurred when the relative humidity of the air
mass was low and the inversion itself may not have been
very strong. Although the inversion still exists, it is getting
weaker as instability gradually increases in the postfrontal
environment. Figure 3 also shows that moist inversions
typically do not occur on exceedance days. Moist inversions
are usually associated with cloud layers that inhibit the
photochemical production of ozone.
3.5. Case Study of 28 August to 1 September 2006
[30] This section describes a case study of a 5-day period,
28 August to 1 September 2006, that includes the days with
the poorest air quality index (31 August to 1 September) in
2006 (www.tceq.state.tx.us/compliance/monitoring/air/
monops/sigevents06.html), using rawinsonde and ozonesonde observations. On 31 August, ozone levels measured
‘‘Very Unhealthy,’’ based upon the EPA’s Air Quality Index
(AQI) scale, with a peak 8-h average of 126 ppbv measured
at the Westhollow site. On 1 September ozone levels were
also ‘‘Very Unhealthy’’ with a peak 8-h average of 129 ppbv
at the LaPorte site.
[31] Figure 4 depicts a sequence of rawinsonde and
ozonesonde profiles at about 1300 CDT for the 5-day
period. This sequence provides a picture of the free troposphere and the boundary layer on the days prior to the ozone
exceedance (28 and 29 August). Figures 5 and 6 show the
two ozone exceedance days, 31 August and 1 September, in
greater detail using highly temporally resolved rawinsondes.
The days 28 and 29 August were prefrontal days. The
frontal passage occurred on the evening of 29 August.
[32] On 28 August, low ozone values (30 ppbv, typical
of marine background levels) are found from the surface up
Figure 3. Eight-hour average ozone peak compared with
elevated inversion strength (°C) and relative humidity (RH, %)
at the top of the inversion.
5 of 11
D17315
RAPPENGLÜCK ET AL.: ATMOSPHERIC STRUCTURE AND O3 IN HOUSTON
D17315
Figure 4. Boundary layer development from 28 August to 1 September 2006 (for description, see text).
to 3000 m AGL. Between 3000 and 4500 m AGL ozone
increases steadily to 50 ppbv. It remains 50 ppbv up to
5200 m AGL, above which ozone steadily increases to
70 ppbv (not shown here). On 29 August before the frontal
passage, the ozone profile remains almost unchanged in the
layers above 3500 m AGL. However, below 3500 m AGL
background levels rise to 35– 45 ppbv. Below 1000 m AGL
additional photochemical ozone production in the PBL
leads to 45 ppbv of ozone. Back trajectory analyses suggest
that the peak ozone values of around 45 ppbv passed
through southwestern and western Texas.
[33] After the frontal passage, the 30 August data show a
strong increase in background ozone of 30 ppbv up to
3500 m AGL and of 20 ppbv between 3500 and 5000 m
AGL. This increase is also present in the surface ozone data.
According to the TCEQ Web site (www.tceq.state.tx.us/
compliance/monitoring/air/monops/sigevents06.html), the
approximate peak regional background levels were about
68 ppbv on 31 August and about 66 ppbv on 1 September,
and thus close to the ozone values observed on top of the
PBL around 1300 CDT (Figure 4). Without the increase of
the background ozone levels, ozone exceedances would not
have occurred. Back trajectory analyses suggest increasing
inflow of continental air masses from the north.
[34] Note that the vertical ozone profile changed drastically from 28 to 30 August. On 28 August, ozone values
6 of 11
D17315
RAPPENGLÜCK ET AL.: ATMOSPHERIC STRUCTURE AND O3 IN HOUSTON
D17315
Figure 5. Boundary layer development on 31 August 2006 based on rawinsonde launches at 0500,
0700, 1000, 13000, 1600, 1900, and 2200 CDT (if not otherwise noted). (top left) Time series of
boundary layer height and (top right) ozone profiles obtained at 0700 and 1329 CDT. (middle left)
Profiles of potential temperature and (middle right) wind direction. (bottom left) Wind speed and (bottom
right) relative humidity.
were constant with altitude up to 3000 m AGL. Above 3000 m
AGL ozone increased steadily from around 25 ppbv up to
45 ppbv at 5000 m AGL. On 29 August ozone values were
between 35 and 45 ppbv throughout the troposphere up to
5000 m AGL, and thus already 10 –20 ppbv higher up to
3000 m AGL than on the preceding day. On 30 August
ozone values decreased with altitude, a negative gradient
which persisted on 31 August and 1 September as well.
[35] Furthermore, the PBL height increased each day
from 29 August until 1 September (Figure 4). With increasing convective processes, more of the air aloft with enhanced continental background ozone levels was likely
7 of 11
D17315
RAPPENGLÜCK ET AL.: ATMOSPHERIC STRUCTURE AND O3 IN HOUSTON
D17315
Figure 6. Same as Figure 5, but for 1 September 2006.
mixed down into the photochemically active PBL over the
HGB. As shown in Figure 5, 31 August was characterized
by (1) a high PBL height of almost 2000 m AGL, (2) a
residual layer with background ozone levels of 60 ppbv,
and (3) an additional local photochemical production of
ozone of 40 ppbv.
[36] Figure 4 reveals a significant increase in ozone in the
boundary layer and within a portion of the lower free
troposphere. The large increase of ozone values above the
boundary layer from 35 ppb on 29 August to 55– 60 ppb
on 30 August can be attributed to air masses with different
origins, as indicated by changes in the wind directions (see
Figure 4). While Houston was under the influence of a
maritime air mass from the Gulf of Mexico on 28 August,
the ozone levels in the troposphere up to 2500 m AGL were
rather low. After the frontal passage on 29 August, the
background levels increased substantially.
[37] Figure 4 also suggests significant local ozone production in the PBL. Apart from high solar radiation and
stagnant conditions, the first two postfrontal days of 30 and
8 of 11
D17315
RAPPENGLÜCK ET AL.: ATMOSPHERIC STRUCTURE AND O3 IN HOUSTON
31 August both were characterized by dry conditions aloft
and strong elevated inversions, indicating an extremely
stable atmosphere (see Figure 4), thereby reinforcing the
previous discussion of postfrontal environments associated
with the analysis of Figure 3. On 31 August, an anticyclone
located to the west of Houston at 850 hPa, resulted in north/
northeasterly upper-level winds over the HGB throughout
the day. On 1 September the high at 850 hPa had intensified
and moved closer to southeast Texas. The days 31 August
and 1 September were characterized by a combination of
constant inflow of continental ozone and the development
of the local land-sea-bay breeze system leading to the ozone
exceedances within the PBL.
[38] The morning inversion height on 31 August at
0700 CDT was 150 m AGL, and the mean wind in this
inversion layer was from 82° at 4.9° m s 1 (Figure 5). The
ozone sounding at this same time revealed an increase in
ozone mixing ratio with altitude from the surface to the top
of the inversion layer (Figure 5). In the residual layer above
the inversion, ozone concentrations rapidly increased to
63 ppbv at 800 m AGL, before rapidly decreasing to 31 ppbv
at an altitude of 1000 m AGL. As ozone decreased near
1000 m, the relative humidity rapidly increased from 46%
to 67%, in association with a cloud layer. Such features
have been observed on three of the 12 morning ozone
soundings and appear to be chemical in origin. A detailed
discussion of such features is beyond the scope of this paper
that is focused instead upon meteorological influences on
ozone. However, one possible explanation is provided
below.
[39] SO2 can be a major ingredient of industrial plumes in
Houston. SO2 in liquid phase may deplete ozone (H. G. J.
Smit, personal communication, 2006). While Wang and
Sassen [2000] report an ozone depletion event in clouds
depending on the liquid water content, they do not address
an SO2 mechanism. Enhanced values of SO2 were observed
on 31 August. In particular, it is striking that the increase of
the boundary layer height around 1000 – 1100 CDT, when
the boundary layer height reached levels above 1000 m
AGL, was accompanied by a strong increase in ozone and
peroxyacetyl nitrate by almost 50 ppbv/h and 1.7 ppbv/h,
respectively, as measured by instruments at the Moody
Tower on the University of Houston campus. The CAMS
site C81 (TCEQ Houston Regional Office) near the University of Houston also reported an increase in SO2 from
2 ppbv to almost 70 ppbv at the same time. It is likely that
the increasing boundary layer height led to downward
mixing of pollutants from the residual layer possibly including lamina of SO2. One possible explanation of the
ozonesonde data, therefore, is that preexisting SO2 and
clouds in the residual layer may have led to O3 depletion.
[40] In the same morning sounding, a strong temperature
inversion was observed at 3500 m AGL, with dry continental air coming from the north and a dew point depression
of about 42°C (2– 3% relative humidity). This inversion
also corresponded with a sharp increase in ozone mixing
ratio, which is seen in the 0700 CDT ozone profile.
[41] The 1329 CDT vertical ozone profile (Figure 5)
shows a well-mixed layer, with ozone concentrations up
to 104 ppbv between the ground and an inversion at approximately 1600 m AGL. This same inversion can also be seen
clearly on the 1300 CDT rawinsonde profile (Figure 5).
D17315
Ozone concentrations decreased above the inversion due
to its ‘‘capping’’ effect, and the wind speeds increased and
shifted to northerly. At 3500 m AGL, an increase in ozone is
observed underneath the higher inversion. In the midafternoon, the atmosphere was well-mixed with easterly winds
from the surface to 150 m AGL. The sea breeze from
distinct SE-S wind directions was evident in the 1900 CDT
rawinsonde sounding, with a height of approximately 700 m
AGL. The 2200 CDT sounding revealed stronger southerly
winds up to 1000 m AGL. The strong inversion aloft was
still present but had fallen to 2800 m AGL.
[42] The morning inversion on 1 September began with a
height of 125 m AGL at 0100 CDT and increased to 140 m
AGL by 0500 CDT (see Figure 6). While the surface
temperatures of the two soundings were similar (26.5°C
and 26.7°C, respectively), the wind patterns revealed some
differences. The mean wind in the morning inversion layer
at 0100 CDT was from 203° at a speed of 3.4 m s 1, while
at 0500 CDT it was from 239° at a speed of 3.0 m s 1.
While no ozone data were collected for the 0700 CDT
launch, the afternoon ozone sounding at 1354 CDT showed
a well-mixed layer up to approximately 1800 m AGL, the
height at which an inversion is also seen in the rawinsonde
profile (Figure 6). High ozone concentrations of 100 –
117 ppbv were observed in this layer. Above this layer,
ozone levels decreased significantly, possibly due to a cloud
layer (evident on the rawinsonde profile where the relative
humidity is close to 100%; see Figure 6), and the presence
of the strong elevated inversion that limited the upward
mixing of high ozone levels from below. The cloud layer
was approximately 200 m thick at 1800 – 2000 m AGL. The
remainder of the ozone decrease seems to occur at a
‘‘normal’’/slower pace above 2000 m AGL. At 3800 –
4000 m AGL, a sharp increase in ozone to 84 ppbv occurred
in the vicinity of the strong inversion that remained in place
from 31 August. A weak sea breeze (2 m s 1) from the
southeast (136°) became evident in the lowest 700 m
AGL in the 1600 CDT sounding.
[43] In a layer between 2700 m and 3500 m AGL
background ozone levels 60 ppbv persisted for the entire
3 day period (30 August to 01 September) (see Figure 4).
Above this layer the ozone profile showed more variability,
possibly due to long range transport events that did not
penetrate into the lower troposphere. The layer between
2700 AGL and 3500 m AGL altitude was always associated
with northerly wind directions.
[44] As can be seen from Figures 5 and 6, land-sea breeze
conditions clearly occurred on 31 August and 1 September
and were confined to the PBL. These 2 days are both
characterized by strong wind shear at the top of the PBL.
Usually the land-sea breeze system can be identified best by
examining the wind pattern at the coastal Galveston site that
allows for the characterization of the marine background
conditions under onshore wind conditions. On 31 August
the Galveston site reported 60 ppbv of ozone during
daytime under onshore wind conditions, an ozone level that
coincides with the tropospheric background conditions
between 2700 and 3500 m AGL altitude. The corresponding
values on 1 September are 20 ppbv higher, indicating the
impact of recirculating air masses which under the land-sea
breeze condition likely will lead to higher regional background ozone levels. These higher background levels are
9 of 11
D17315
RAPPENGLÜCK ET AL.: ATMOSPHERIC STRUCTURE AND O3 IN HOUSTON
also reflected in the overall increase of the afternoon ozone
levels that rise from 100 ppbv in the PBL up to 1500 m
AGL on 31 August to 120 ppbv on 1 September (as
indicated by the ozonesonde data). Strong increases in
ozone levels at individual sites in the HGB are usually
associated with winds that shift through easterly directions,
coinciding with bay breezes at this time [University of
Houston, 2007]. The strong increases in ozone levels
around 1000 CDT (as indicated by CAMS surface data,
not shown) on 31 August are closely related to these
conditions, whereas on 1 September, corresponding
increases in ozone levels occur slightly later, consistent
with the later wind shifts on that day.
4. Conclusions
[45] This study investigated meteorological conditions
leading to ozone exceedances in the HGB during TexAQS-II
in August and September of 2006. This study included a
classification of synoptic patterns and a characterization of
the planetary boundary layer and the free troposphere above
up to about 5000 m AGL based on rawinsonde and
ozonesonde launchings from the University of Houston’s
main campus approximately 5 km to the southeast of
downtown Houston.
[46] Cluster analysis of synoptic patterns and surface
background ozone levels showed that Houston experienced
high background surface O3 values under northerly or
easterly flow regimes at 850 hPa, coinciding with the
advection of dry continental air. Observations showed that
many exceedance days were accompanied by the presence
of strong elevated inversions, which are typical of postfrontal soundings. In fact all O3 exceedances in the month
of September occurred in a postfrontal environment. Higher
afternoon ozone peaks occurred when there was an extremely dry air mass aloft, although they were not always
associated with the strongest observed inversion. Because
these latter conditions were usually found right after a
frontal passage, the northerly flow was still strong enough
to eliminate any chance for stagnation. However, in all
postfrontal environments, background ozone was elevated,
and analysis of ozonesonde data revealed that regional-scale
transport may have an impact on local O3 levels. Depending
on the air mass origin, background ozone levels may range
from 30 ppbv (marine) to nearly 60– 70 ppbv (continental)
throughout the lower troposphere. These different background ozone patterns are also present in the surface O3
data, i.e., increased background ozone levels may help to
raise ozone values to the 8-h ozone exceedance level. These
changes often occurred after frontal passages that result in
shifts in wind direction from marine to continental advection.
[47] A previous hypothesis that a stronger morning inversion could lead to higher daytime ozone peaks was not
supported by the August – September 2006 Houston data.
Rather, the important factors for the case study 28 August to
1 September 2006 seem to be that (1) under well-developed
land-sea-bay breeze conditions easterly wind conditions
occur which bring Ship Channel emissions into the HGB
and that (2) these conditions occur during late morning
hours at relatively low PBL heights at a time when fast
photochemistry occurs. Low PBL heights at this time will
also likely produce a volume effect leading to higher
D17315
concentrations that also can enhance the overall probability
for an 8-h ozone exceedance but not necessarily lead to
higher peak ozone values in the afternoon.
[48] The overall results for the HGB indicate that largescale northerly flows, often initiated by frontal passages,
will lead to a rapid and significant change in background
ozone levels due to the shift in wind direction and associated change of source areas for air masses (continental
versus marine). Postfrontal subsidence will most likely lead
to an additional O3 increase. As the front departs the area
and high pressure settles in, local and regional photochemistry will produce ozone on top of the already enhanced
background levels, which may lead to an ozone exceedance.
Increased convective processes will also lead to increased
PBL heights, and the enhanced continental ozone background levels aloft are likely mixed into the photochemically active PBL over the HGB. While the upper layers
frequently show winds from the north and continue to
advect continental air masses into the region, winds in the
HGB are dominated by the land-sea-bay breeze system. As
this land-sea-bay breeze system frequently includes recirculation processes, air masses confined to this system will
repeatedly entrain emissions that when exposed to solar
radiation lead to high ozone concentrations.
[49] Acknowledgments. The authors would like to express their
gratitude to the Texas Commission on Environmental Quality for supporting and funding this research under grants 582-5-64594 and 582-5-708070016 and for providing us with the CAMS data sets. Also, thanks to the
many graduate and undergraduate students at the University of Houston for
making the rawinsonde campaign possible. In particular, thanks to Bridget
McEvoy-Day and Rennee Boudreaux for helping on the ozonesonde
launches and Ashley Mefferd for valuable discussions. Special thanks to
Fong Ngan for helping compile the synoptic maps and providing the
category classifications. Additional support for the ozonesonde program
came from NASA’s Office of Earth Science and from the Rice University
Shell Center for Sustainability.
References
Alonso, L., G. Gangoiti, and M. Navazo (2000), Transport of tropospheric
ozone over the Bay of Biscay and the eastern Cantabrian coast of Spain,
J. Appl. Meteorol., 39, 475 – 486, doi:10.1175/1520-0450(2000)039<0475:
TOTOOT>2.0.CO;2.
Angevine, W. M., A. W. Grimsdell, L. M. Hartten, and A. C. Delany
(1998), The flatland boundary layer experiments, Bull. Am. Meteorol.
Soc., 79, 419 – 431, doi:10.1175/1520-0477(1998)079<0419:
TFBLE>2.0.CO;2.
Angevine, W. M., A. B. White, K. Knupp, R. Coulter, T. Martine, J. C.
Doran, and D. White (2002), Mixed layer depths variability in the Houston area during TEXAQS2000, paper presented at 4th Conference on
Atmospheric Chemistry, Am. Meteorol. Soc., Orlando, Fla.
Banta, R. M., C. J. Senff, J. Nielsen-Gammon, L. S. Darby, T. B. Ryerson,
R. J. Alvarez, S. P. Sandberg, E. J. Williams, and M. Trainer (2005), A
bad air day in Houston, Bull. Am. Meteorol. Soc., 86, 657 – 669,
doi:10.1175/BAMS-86-5-657.
Berman, S., K. Jia-Yeong, and S. Trivikrama Rao (1999), Spatial and
temporal variation in the mixing depth over the northeastern United
States during the summer of 1995, J. Appl. Meteorol., 38, 1661 – 1673,
doi:10.1175/1520-0450(1999)038<1661:SATVIT>2.0.CO;2.
Darby, L. (2005), Cluster analysis of surface winds in Houston, Texas, and
the impact of wind patterns on ozone, J. Appl. Meteorol., 44, 1788 – 1806,
doi:10.1175/JAM2320.1.
Daum, P. H., L. I. Kleinman, S. R. Springston, L. J. Nunnermacker, Y.-N.
Lee, J. Weinstein-Lloyd, J. Zheng, and C. Berkowitz (2004), Origin and
properties of plumes of high ozone observed during Texas 2000 Air
Quality Study (TEXAQS 2000), J. Geophys. Res., 109, D17306,
doi:10.1029/2003JD004311.
Davis, J. M., B. K. Eder, D. Nychka, and Q. Yang (1998), Modeling the
effects of meteorology on ozone in Houston using cluster analysis and
generalized additivce models, Atmos. Environ., 32, 2505 – 2520,
doi:10.1016/S1352-2310(98)00008-9.
10 of 11
D17315
RAPPENGLÜCK ET AL.: ATMOSPHERIC STRUCTURE AND O3 IN HOUSTON
EN-SCI Corporation (1997), Instruction manual for EN-SCI Corporation
model KZ-ECC atmospheric ozone sounding system, Boulder, Colo.
EN-SCI Corporation (2003a), VO3 Reference Manual: Digital ozone and
rawinsonde data acquisition and processing software (for use with EN-SCI
Corporation model 2Z-ECC ozonesonde, incorporating the Innovative
Systems Design V2 interface with optional GPS, version 5), Boulder,
Colo.
EN-SCI Corporation (2003b), Information sheet for EN-SCI Corporation
model 2Z-ECC atmospheric ozone sounding system, Boulder, Colo.
Eresmaa, N., A. Karppinen, S. M. Joffre, J. Räsänen, and H. Talvitie
(2006), Mixing height determination by ceilometer, J. Atmos. Chem.
Phys., 6, 1485 – 1493.
Grimsdell, A., and W. Angevine (1998), Convective boundary layer height
measurement with wind profilers and comparison to cloud base, J. Atmos.
Oceanic Technol., 15, 1331–1338, doi:10.1175/1520-0426(1998)015<1331:
CBLHMW>2.0.CO;2.
Kerr, J. B., et al. (1994), The 1991 WMO International Ozonesonde Intercomparison at Vanscoy, Canada, Atmos. Ocean, 32(4), 685 – 716.
Kleinman, L. I., P. H. Daum, D. Imre, Y.-N. Lee, L. J. Nunnemacker, and
S. R. Springston (2002), Ozone production rate and hydrocarbon reactivity in 5 urban areas: A cause of high ozone concentration in Houston,
Geophys. Res. Lett., 29(10), 1467, doi:10.1029/2001GL014569.
Komhyr, W. D., R. A. Barnes, G. B. Brothers, J. A. Lathrop, and D. P.
Opperman (1995), Electrochemical concentration cell performance evaluation during STOIC, J. Geophys. Res., 100, 9231 – 9244, doi:10.1029/
94JD02175.
Lau, W. K. M., and K. M. Kim (2007), How nature foiled the 2006 hurricane forecasts, Eos Trans. AGU, 88(9), doi:10.1029/2007EO090002.
Marsik, F., K. Fischer, T. Mcdonald, and P. Samson (1995), Comparison of
methods for estimating mixing height used during the 1992 Atlanta field
intensive, J. Appl. Meteorol., 34, 1802 – 1814, doi:10.1175/15200450(1995)034<1802:COMFEM>2.0.CO;2.
Morris, G. A., et al. (2006), Alaskan and Canadian forest fires exacerbate
ozone pollution over Houston, Texas, on 19 and 20 July 2004, J. Geophys. Res., 111, D24S03, doi:10.1029/2006JD007090.
Ngan, F., and D. Byun (2007), Trajectory analysis for source-receptor
relation of high O3 events over Houston/Galveston area, paper presented
at 9th Conference on Atmospheric Chemistry, Am. Meteorol. Soc., San
Antonio, Tex.
Nielsen-Gammon, J., J. Tobin, A. McNeel, and G. Li (2005a), A conceptual
model for eight-hour ozone exceedances in Houston, Texas Part I: Background ozone levels in eastern Texas, 52 pp., technical report, Houston
Adv. Res. Cent., Houston, Tex.
Nielsen-Gammon, J., J. Tobin, and A. McNeel (2005b), A conceptual model
for eight-hour ozone exceedances in Houston, Texas Part II: Eight hour
ozone exceedances in the Houston-Galveston metropolitan area, 79 pp.,
technical report, Houston Adv. Res. Cent., Houston, Tex.
Nielsen-Gammon, J., C. L. Powell, M. J. Mahoney, W. M. Angevine,
C. Senff, A. White, C. Berkowitz, C. Doran, and K. Knupp (2008),
Multisensor estimation of mixing heights over a coastal city, J. Appl.
Meteorol. Climatol., 47, 27 – 43, doi:10.1175/2007JAMC1503.1.
Reid, S. J., G. Vaughan, A. R. W. Marsh, and H. G. J. Smit (1996),
Accuracy of ozonesonde measurements in the troposphere, J. Atmos.
Chem., 25, 215 – 226, doi:10.1007/BF00053792.
D17315
Reiss, R. (2006), Temporal trends and weekend-weekday differences for
benzene and 1,3-butadiene in Houston, Texas, Atmos. Environ., 40,
4711 – 4724, doi:10.1016/j.atmosenv.2006.04.023.
Russell, J. M., III, H. G. J. Smit, D. Hofmann, and C. Rodgers (1998),
Stratospheric processes and their role in climate (SPARC): Assessment of
trends in the vertical distribution of ozone, WMO Ozone Res. And Monitor. Proj. Rep. 43, World Clim. Res. Prog., Geneva.
Ryerson, T. B., et al. (2003), Effect of petrochemical industrial emissions of
reactive alkenes and NOx on tropospheric ozone formation in Houston,
Texas, J. Geophys. Res., 108(D8), 4249, doi:10.1029/2002JD003070.
Senff, C., R. Banta, L. Darby, W. Angevine, A. White, C. Berkowitz, and
C. Doran (2002), Spatial and temporal variations in mixing height in
Houston, 58 pp., technical report, Texas Comm. on Environ. Qual.,
Houston, Tex.
Smit, H. G. J., and D. Kley (1996), Jülich ozone sonde intercomparison
experiment (JOSIE), World Meteorol. Org. Global Atmos. Watch, Geneva.
Smit, H. G. J., et al. (2007), Assessment of the performance of ECCozonesondes under quasi-flight conditions in the environmental simulation chamber: Insights from the Juelich Ozone Sonde Intercomparison
Experiment (JOSIE), J. Geophys. Res., 112, D19306, doi:10.1029/
2006JD007308.
University of Houston (2007), Rawinsonde launches, tethered-balloon
ozone soundings, and micrometeorological flux measurements and analyses, final report, 68 pp., Texas Comm. on Environ. Qual., Houston.
Wang, Z., and K. Sassen (2000), Ozone destruction in continental stratus
clouds: An aircraft case study, J. Appl. Meteorol., 39, 875 – 886,
doi:10.1175/1520-0450(2000)039<0875:ODICSC>2.0.CO;2.
Wiegner, M., S. Emeis, F. Volker, B. Heese, W. Junkermann, C. Munkel,
K. Schäfer, M. Seefeldner, and S. Vogt (2006), Mixing layer height over
Munich, Germany: Variability and comparisons of different methodologies, J. Geophys. Res., 111, D13201, doi:10.1029/2005JD006593.
Yi, C., K. Davis, and B. Berger (2001), Long-term observations of the
dynamics of the continental planetary boundary layer, J. Atmos. Sci.,
58, 1288 – 1299, doi:10.1175/1520-0469(2001)058<1288:LTOOTD>
2.0.CO;2.
Yu, C.-H., and R. Pielke (1986), Mesoscale air quality under stagnant
synoptic cold season conditions in the Lake Powell area, Atmos. Environ.,
20, 1751 – 1762, doi:10.1016/0004-6981(86)90124-1.
Zhang, J., and S. T. Rao (1999), The role of vertical mixing in the temporal
evolution of ground-level ozone concentrations, J. Appl. Meteorol., 38,
1674 – 1691, doi:10.1175/1520-0450(1999)038<1674:TROVMI>2.0.CO;2.
G. A. Morris, Department of Physics and Astronomy, Valparaiso
University, Valparaiso, IN 46383, USA.
R. Perna, Regional Office, Texas Commission on Environmental Quality,
5425 Polk Street, Houston, TX 77023, USA.
B. Rappenglück, Department of Geosciences, University of Houston,
4800 Calhoun Road, Houston, TX 77204, USA. (brappenglueck@uh.edu)
S. Zhong, Department of Geography, Michigan State University, 116
Geography Building, East Lansing, MI 48824, USA.
11 of 11