A51G-01
The Role of Galveston Bay Meteorology in Ozone Concentrations in Houston, Texas
Galveston Bay is a body of water approximately 50 km by 30 km that opens into the Gulf of Mexico. The head of Galveston Bay is adjacent to the Houston Ship Channel, where large industrial emitters of ozone precursors are located. We consider the role of coastal circulations and mixing suppression over Galveston Bay in the location and magnitude of peak 8-h ozone values at monitors within the Houston-Galveston-Brazoria nonattainment area. Galveston Bay is found to have a salutary effect on ozone concentrations under most circumstances. Aside from boats and ships, Galveston Bay is free of ozone precursor emissions, and the onset of the sea breeze along Galveston Bay usually causes an immediate drop in ozone levels. The situation changes dramatically when the morning wind direction is from the northwest, allowing ozone precursors emitted during the night or early morning that are distributed in a shallow atmospheric layer to be carried over Galveston Bay. In morning and early afternoon, mixing is suppressed over Galveston Bay, and precursor concentrations remain high in the absence of ventilation. If winds are strong, high ozone concentrations are generally observed along the far shore of Galveston Bay. If winds are weak, a sea breeze will develop along Galveston Bay, and high levels of ozone will be carried back onshore in the vicinity of the Ship Channel. The coastal oscillation (a regular rotation of the wind vector driving by coastal heating contrasts near 30 N) can lead to wind reversals and high values of ozone by itself. We investigate peak ozone concentrations under a variety of wind conditions to infer whether the ozone concentrations are, on balance, higher due to the net effect of meteorological processes associated with Galveston Bay.
A51G-02
Mixing Heights and Three-Dimensional Ozone Structure Observed by Airborne Lidar During the 2006 Texas Air Quality Study
A new all solid state ozone lidar was deployed on a NOAA Twin Otter to study boundary layer ozone and aerosol, mostly around Houston, during the 2006 Texas Air Quality Study. The new instrument transmits high pulse-rate, low pulse-energy light at 3 wavelengths in the ultraviolet to obtain ozone profiles with 500 m horizontal resolution and 90 m vertical resolution. During the Texas field study, 20 research flights resulted in nearly 70 hours of ozone measurements during the period from August 1 to September 15. Science objectives included characterization of background ozone levels over rural areas near Houston and Dallas and variability and structure of the boundary layer over different surface types, including urban, wooded, and agricultural land surface areas as well as over Galveston Bay and the Gulf of Mexico. A histogram of all boundary layer ozone concentration measurements showed a bimodal distribution with modes at 45 ppb and 70 ppb. The lower mode correlated with southerly flow, when relatively clean air was transported onshore into the Houston area. Segmenting the observations during southerly flow by region, including the Gulf of Mexico, land within about 55 km from the coast, and further inland indicated that background levels increased by about 10 ppb as air was transported onshore. During the latter part of the experiment, as more pollution was imported into the Houston region, background levels rose to nearly 80 ppb in regions N of Houston. Two flights aimed at observing import of ozone into Texas from the east showed that ozone concentrations increased and boundary layer depths deepened upwind of Houston between September 4 and September 8. Background levels rose by more than 10 ppb over this period. In addition to ozone measurements, we also estimated boundary layer height based on maximum gradient in observed backscatter. The technique worked well when the layer topped by the strongest gradient extends down to the surface. Investigation of the correlation between ozone levels and mixing layer heights both within and external to the Houston urban plume showed a variety of relationships, depending on, e.g., wind direction and occurrence of a bay/gulf breeze. On a day-to-day basis, higher ozone levels were weakly correlated with deeper mixing levels – this was likely due to advection of the urban heat island downwind with the high-ozone urban plume.
A51G-03
Mixing heights and surface fluxes over Galveston Bay and the Gulf of Mexico: Implications for modeling of pollution episodes
During the 2006 Texas Air Quality Study, boundary layer measurements were made by in-situ instruments, lidars, and rawinsondes on the NOAA RV Ronald H. Brown as well as by radar wind profilers on land. Brown also carried instruments to measure surface heat and momentum fluxes. This presentation will emphasize measurements made in Galveston Bay and in the Gulf of Mexico near the Houston area. Details of boundary layer depth and turbulence intensity over these waters have not been well known previously, but are quite important to the understanding of high ozone episodes in Houston. One somewhat surprising result is that the boundary layer over water was almost always slightly unstable, with positive surface heat flux. Mixing depths were moderate, although mixing was generally weak compared to that over land. Boundary layer heights over the water were substantially shallower than daytime heights over land. Experiments in modeling ozone episodes with WRF at 1.5-km grid spacing will be shown and compared with the measurements.
A51G-04
Basin by Night and Plateau by Day: Air Pollution Accumulation and Ventilation in the Kathmandu Valley, Nepal
The Kathmandu Valley, Nepal, is a broad bowl-shaped basin in the Himalayan foothills, with a population of more than two million people. Its growing air pollution problem is strongly affected by wind systems generated by the heating and cooling of the surrounding topography. We carried out a field measurement campaign of air pollutants and meteorology in the Kathmandu Valley during the dry season 2004-2005, followed by MM5 simulations nested down to 1 km resolution. The model results have allowed us to interpret the observed distinct diurnal cycle and determine the major processes responsible for the observations. We found that, at night, the Kathmandu Valley behaves like a large basin, with down-slope flows on surrounding mountains contributing to an accumulating cold air pool that grows to the altitude of surrounding mountain passes. The arriving cold air pushes underneath the air mass polluted in the evening, lifting up layers of polluted air, while providing relatively clean conditions to the surface during pre-dawn hours. After sunrise the elevated layers re-circulate to the surface, leading to a sharp rise in pollution. During the morning a mixed layer begins to grow over the valley, but it peaks at mid-day. Further growth is stunted by the arrival through mountain passes of cooler air that originated over lower regions outside of the Kathmandu Valley. This creates a renewed stratification as cooler air spreads across valley bottom; such behavior has been observed elsewhere on a number of elevated plateaus. During the afternoon most of the valley's ventilation takes place at low elevations, with strong winds bringing background air through the western passes, while transporting polluted air out the eastern passes. Our study sheds light onto some of the complex patterns of air pollution transport that take places in polluted mountain areas, while helping guide future research, including a proposed study of the ventilation of Ganges Valley air through the Himalaya.
A51G-05
Investigation of Orographic Venting of Atmospheric Boundary Layer Air Using Observations and the WRF-Chem Model
Three-dimensional air-mass transport under fair weather conditions in California's Central Valley between the atmospheric boundary layer (ABL) to the free troposphere (FT) is investigated with a case study using the observations from the Central California Ozone Study (CCOS) 2000 field experiment and the WRF-Chem model. First, the comparisons of the observations and the model simulation are carried out to show that the WRF-Chem model simulates the meteorological processes of this case well. Then, the movement of the tracers simulated by the WRF-Chem model is analyzed to reveal the characteristics of three-dimensional transport. Our results indicate that the three-dimensional air-mass transport is very complex under fair weather conditions in the Central Valley. During the day, the development of the convective ABL acts as an efficient "air pump" that transports pollutants upward as the ABL grows. However, the slope wind system around the Central Valley plays an important role in redistributing pollutants vertically as well as horizontally. The tracer movement indicts that once lifted to the FT, the pollutants that originate in the ABL are vertically recirculated downward toward the ABL and horizontally transported by the synoptic flow. During the night, as the two major horizontal mesoscale low- level eddies (i.e., the Schultz and Fresno Eddies) are formed, the remnants of the pollution in the ABL converge horizontally and are recirculated along the foothills, where they remain at low-levels until sunrise and are then vertically redistributed as the daytime ABL develops. Under typical fair weather conditions in summer, there are two major lower tropospheric pathways for the ABL air masses to move out the Central Valley: eastward over the Sierra Nevada and southward through the Tehachapi Pass.
A51G-06
The Daytime Mixed Layer Observed by Radiosonde, Profiler, and Lidar during MILAGRO
During the 2006 MILAGRO field campaign centered in the Mexico City area, scientists from Pacific Northwest National Laboratory (PNNL), Argonne National Laboratory (ANL), and the University of Alabama in Huntsville (UAH) operated a variety of atmospheric profiling systems. The systems were located at the Mexican Petroleum Institute (IMP) in central Mexico City, at Tecámac University on the northeastern edge of the Mexico City urban area, and at Rancho la Bisnaga, a privately owned ranch. These sites are referred to as T0, T1, and T2 to reflect the idea of sequential arrival times of air parcels at each site under common transport conditions. T2 was approximately 50 km north of T1. Similar 915 MHz wind profiling radars were operated at all three sites. ANL additionally operated a sodar, a 523-nm micropulse lidar, and a radiosonde system at T1. At T2, PNNL additionally operated a radiosonde system. On aircraft sampling days, five radiosondes were launched at T1 during the daytime, and three were launched at T2. The collocation of profiling systems afforded an opportunity to compare the three primary methods of measuring the depth of deep mixing layers in the Mexico City environment. In this paper we will describe the methods of extracting mixing layer depth from each system as well as the results of the comparison. The results highlight the general agreement among the various methods, but also the ambiguity that results from multiple inversions for radiosondes and from elevated layers with significant particulate matter for lidars. We conclude that the wind profiler is the system of choice for identifying mixing layer depth from the MILAGRO campaign. In addition to the instrument comparison, we also have completed a description of the temporal variability of the mixing layer during the campaign as well as spatial variations among the three sites. We have interpreted our results in the contexts of the widely reported Norte events, of the days with and without occurrences of deep convection, and of wind profiler measurements of mixing layer depth from the IMADA-AVER campaign in 1997. Our findings include the results that turbulent mixing began at the surface 67.5 ± 15 min following sunrise, that daily mixed layer growth was nearly identical in the mean and in its variability between the IMADA-AVER and the MILAGRO campaigns, that conditions that produced deep convection did not systematically change morning boundary layer growth, and that Norte events did not seem to strongly perturb mixed layer behavior on the Central Mexican Plateau. Because of its relevance for aerosol particle size distributions, we will also describe the vertical structure of humidity relative to mixing layer depth.
A51G-07
Aerosol Transport Questions Arising From Micro Pulse Lidar Measurements During MILAGRO
A Micro Pulse Lidar (MPL) was operated by Argonne National Laboratory at the Universidad de Tecámac site (T1) during the MILAGRO field campaign in March 2006 in the environs of Mexico City. Located approximately 40 km north of the urban center of Mexico City, the T1 site was expected to observe the transport and evolution of aerosols as they moved out of the urban area on predominantly south winds. Because of the collocation of numerous other remote and in-situ sensors of aerosols, winds, temperatures and moisture. The MPL, operating at 0.527 microns, provides estimates of scattering in 15 m range (height) intervals, averaged over 10 sec between 200 m and (characteristically) 15 km, from which extinction profiles through and above the mixed layer can be calculated, mixed layer heights and evolution obtained and stratification and modification of aerosols observed. During the MILAGRO field campaign the growth of the mixed layer during daytime and the stratification of the atmosphere and subsequent aerosol layering were typical of previous campaigns and agreed well with structure observed by radar wind profiler, sodar, and radiosonde profiles. However, on many evenings very near sunset (1800 LT), a marked decrease in apparent aerosol concentration was observed through much of the depth of the mixed layer that persisted for 1-2 hours. Aerosol concentrations calculated with the CMAQ model coupled with MM5 wind fields will be evaluated with measured vertical profiles. Estimates of local radiative forcing for this site will be generated using MPL data set and model results.
A51G-08
Atmospheric Transport of Smoke and Dust Particulates and their interaction with the PBL as observed by multiwavelength Lidar, GOES and supporting instrumentation
In this paper, we present results showing the usefulness of multiwavelength lidar measurements to study the interaction of aerosols in PBL with long range advected aerosol plumes. In particular, our measurements are used to determine the plume angstrom exponents, which allows us to differentiate smoke events from dust events as well as partitioning the total aerosol optical depth obtained from a CIMEL sky radiometer between the PBL and the high altitude plumes. Furthermore, we show that only if the optical depth from the upper level plumes is taken into account, the correlation between the lidar derived PBL aerosol optical depth and surface PM2.5 is high. In addition, we also observe the dynamic interaction of high altitude plumes interacting with the PBL resulting in a dramatic rise in surface PM10 concentrations without a corresponding dramatic rise in PM2.5 concentrations. These observations strongly suggest the deposition of large particulates into the PBL which is consistent with both lidar angstrom coefficient measurements and backtrajectory analysis. Finally, these measurements are coordinated with GOES AOD measurement data for the case of strong fire plumes. In particular, we show that GOES AOD data is very consistent with the radiometer data and clearly shows the onset of plumes over the NYC area. Furthermore, the cloud mask of GOES is shown to be to liberal and fooled by broken clouds. Modification of the mask is shown to improve performanxce of the GOES observations