insights into houston air quality from rawinsonde and
TRANSCRIPT
Insights into Houston Air Quality from Rawinsonde and Ozonesonde DataRyan Perna1, Sharon Zhong2, Gary Morris3, Bernhard Rappenglueck1 , Barry Lefer1, Renee Boudreaux1, Craig Clements1, Bridget Day1, Ricky Fuller1, Scott Hersey4, Brad Morris5, Monica Patel1, and Leonardo
Pedemonte1, Michael Leuchner1, James Flynn,1 Barry Shaulis1, Paulina Guerrero1, Beryl Onakpoya1, Meongdo Jang1, Daegyun Lee1, Susan Street1, and Wenqing. Yao2
1Dept. of Geosciences, University of Houston, Houston, TX 772042Department of Geography, Michigan State University
3Dept. of Physics & Astronomy, Valparaiso University, Valparaiso, IN 463834Dept. of Chemical Engineering, California Institute of Technology, Pasadena, CA 91125
5Brown College, Rice University, Houston, TX 77005
Although the Gulf Coast is simple in topography, the boundary layer structure and wind circulation in theGulf Coast region of southeastern Texas can be highly variable due to the frequent presence of sea-land
breeze circulations, strong nocturnal low-level jets, and larger-scale diurnal wind oscillations. To better
understand these boundary layer phenomena, we carried out a suite of boundary layer measurementsduring the TEXAQS-II (Texas Air Quality Study – II) field campaign. These measurements included
Rawinsonde and Ozonesonde soundings to obtain information of the atmospheric thermodynamic structureand wind over urban areas of Houston, along with information about the vertical distribution of ozone.
The rawinsonde data can be used to determine the height of the Planetary Boundary Layer (PBL), thestrength and depth of the land/sea-breeze circulation, and the structure of the low-level jet and strength of
nocturnal inversions. The ozonesonde data can be used to complement other routine ground-based ozonemonitoring, validate special aircraft observations, validate and improve air quality model simulations of
ozone episodes. Together, they help achieve a better understanding of air pollution problem in Houston andthe region.
Introduction
Figure 2. Ozonesonde package being launched by Ryan Perna
(left) and Bridget McEvoy-Day (right). Photo by Craig Clements.
For this study, the August-thru-September time frame was chosen since these months usually exhibit the highestfrequency of “high ozone” days based on long-term ground-based ozone observations. A “high ozone” day was
defined using the criteria specified by the U.S. Environmental Protection Agency (EPA) where the measured 8-hour average ozone mixing ratio in any given station is greater than or equal to 85 ppbv. During 2005,
rawinsondes were launched twice per day, one at 700 CDT, corresponding to the standard morning rawinsondelaunch time (12Z), and another in early afternoon between 1300 and 1500 CDT. Beginning in 2006, the afternoon
rawinsonde launch was moved to 1900 CDT to also coincide with the standard launch time (00 Z). On IOPs(Intensive Operation Periods) that were forecast to be high ozone days, rawinsondes were launched at 500, 1000,
1300, 1600, and 2200 CDT in addition to the standard twice daily launches in order to capture the temporal andvertical development of the PBL. The rawinsondes used were Vaisala RS-92 GPS sondes that measured
temperature, pressure, relative humidity, and wind speed/direction.
The balloon-borne ozone measurements utilized aqueous solution Model 2Z electrochemical concentration cell(ECC) ozonesondes manufactured by EN-SCI Corporation of Boulder, Colorado, and provided by Valparaiso
University. The ozonesondes also measured ambient air pressure, air temperature, relative humidity, and GPSposition data for wind speed/direction. Ozonesondes were launched at 700 CDT during the forecasted high-
ozone IOPs to illustrate the complex nature of the vertical ozone profile before mixing took place, and to capturethe pre-existing residual layer(s). On IOPs and other selected days during the TexAQS-II period, ozonesondes
were also launched at 1300 CDT in order to capture the vertical distribution of ozone concentrations at a time ofthe day when the mixed boundary layer height is fully developed.
Site Location
The program began in August 2005 when rawinsondes were launched from the University of Houston(UH) coastal center in La Marque, TX (approx. 65 km south of Houston). The launch site was moved to
the UH main campus in September of 2005 once approval from the Federal Aviation Administration wasobtained. This was done in order to gain further understanding of the urban boundary layer development.
The main UH campus site (29.7421 N and 95.3395 W, 11 m above sea level) was chosen due to its
central location approximately 5 km to the southeast of downtown Houston and thus it is the mostrepresentative site to characterize the structure and dynamics of the urban boundary layer in the Houston
area.
Methods
Downtown
Houston
University of
Houston
Results
Figure 1. Site Location with map of Texas (top left), map of Houston (top right), and
satellite image of the Houston area (bottom).
Figure 3. Ozonesonde package after launch.
Figure 7. September 2, 2006 morning profile with
ozone mixing ratio (ppb), relative humidity (%), and
potential temperature (K)
Figures 7 and 8 show a phenomena that was observed on September 2, 2006 which seems to be a decrease in ozone
mixing ratio within a cloud layer. Though ECC ozonesondes are known to exhibit a negative interference with SO2 itcannot be excluded that liquid phase SO2 chemistry may also have an impact on O3 chemistry. The winds in this cloud
deck appear to be from the East which means they may include some increased SO2 coming from the ship channel
industrial region of Houston. This phenomenon was not observed in any other soundings as of yet, since ozonesondesare not typically launched into a cloud layer. More research is underway to determine the nature of this O3 decrease.
Figure 8. Skew-T of lowest 2000 meters showing easterly
winds at 1000 m level, in region of ozone drop (5 AM
image)
Figure 10. Afternoon ozone profile of 9-20-06.Figure 9. Morning ozone profile of 9-20-06
Figure 9 shows a morning ozone profile revealing a significant residual layer above the morning inversion at about 1000 m
a.g.l.. Ozone levels are quite low underneath the morning inversion likely due to deposition to the surface and titration by NO.
It is likely that downward mixing of approximately 60 ppbv of this residual layer had a significant effect on the increasing O3
peak values in the afternoon (see Figure 10. The difference between the maximum O3 values of about 90 ppbv in the afternoon
boundary layer and the 60 ppbv originating from the input of the residual layer is most likely due to in-situ O3 formation.September 20, 2006 was declared a “high-ozone” day by the TCEQ.
September 2, 2006 5:00 AM LDT
August 31, 2006 PBL Development ("High Ozone Day")
Potential Temperature(K)
295 300 305 310 315 320 325
Altit
d(
)
0
1000
2000
3000
4000
5000
5 AM
7 AM
10 AM
1 PM
4 PM
7 PM
10 PM
September 2, 2006 "Moderate" Ozone Day
Potential Temperature(K)
295 300 305 310 315 320 325
Altit
d(
)
0
1000
2000
3000
4000
5000
5 AM
7 AM
10 AM
1 PM
4 PM
7 PM
10 PM
Figure 5. PBL Development ona “Moderate” ozone day as
defined by the TCEQ.
Figure 6. PBL developmenton a “High” ozone day as
defined by the TCEQ.
Figures 5 and 6 illustrate the difference between a “Moderate” and a “High” ozone day in the nature of the PBL
development. September 2 was rather well-mixed from early in the morning which helped to keep O3 levels down.August 31 was a high O3 day, and the lower PBL height along with the delay in development may have been a
factor in increasing Houston’s O3 levels.
Conclusions.
The profiles shown on this poster represent select examples of vertical ozone and rawinsondedata. These profiles offer an insight into the vertical differences of ozone vs. non-ozone days.
It was found that the downward mixing of the early morning residual layer may have a
significant effect on afternoon ozone levels. It was also found that cloud layers may have an
impact on ozone concentration. Further analysis will be completed to determine the cause ofthis effect. Ozone concentration can also be affected by the vertical wind distribution, with
easterly winds being especially important in bringing ozone precursors into the urban
environment from the industrial region of Houston. Easterly wind directions can also be
accompanied by higher continental background ozone levels.
Contact info:
Ryan Perna [email protected]
Bernhard Rappenglueck [email protected]
Gary Morris [email protected]
Sharon Zhong [email protected]
Figure 4a. Potential temperature profiles for ozone
episode days in August 2005.
Figure 4b Potential temperature profiles for non-ozone
episode days in August 2005.
Figures 4a, b compare ozone day rawinsonde profiles launched from LaMarque. A slightly lower PBL height and
more stable stratification as shown in figure 4 may favor higher ozone concentrations later in the afternoon. It canalso be noted that the winds for the non-episode days tended to be from the south, bringing in cleaner air from the
Gulf and additional moisture for clouds and precipitation. Figure 4b illustrates this as the profile is much better-mixed
when compared to figure 4a, indicated by the vertical potential temperature profile in the boundary layer.
Easterly Winds at
1000 meters
Decrease in O3
at 1000 m
University of Houston
Acknowledgements
This project was funded by the Texas Commission on Environmental Quality (TCEQ).
Houston, TX
Early morning (12Z) soundings of potential temperature for non-
ozone episode days in August 2005
Slightly Higher PBL
Height on non-
ozone episode days
mb
A11A-0824
September 20, 2006 7:00 AM LDT
Ozone(ppb) Potential Temperature(K)
0 50 300
Altit
d(
)
0
1000
2000
3000
4000
Ozone(ppb)
Potential Temperature(K)
September 20, 2006 1:00 PM LDT
Ozone(ppb) Potential Temperature(K)
40 60 80 300 310 320
Altit
d(
)
0
1000
2000
3000
4000
5000
Ozone(ppb)
Potential Temperature(K)