Grantee Research Project Results
Final Report: A Numerical Study of the Effects of Large Eddies on Trace Gas Measurements and Photochemistry in the Convective Boundary Layer
EPA Grant Number: R825262Title: A Numerical Study of the Effects of Large Eddies on Trace Gas Measurements and Photochemistry in the Convective Boundary Layer
Investigators: McNider, R. T. , Song, Aaron , Herwehe, Jerold A. , Norris, W. B. , Biazar, Arastoo
Institution: The University of Alabama in Huntsville
EPA Project Officer: Hahn, Intaek
Project Period: February 3, 1997 through February 2, 2000
Project Amount: $275,828
RFA: Air Quality (1996) RFA Text | Recipients Lists
Research Category: Air Quality and Air Toxics , Air
Objective:
The objectives of this research project were to: (1) investigate the appropriate averaging strategies for reactive gases in the convective boundary layer; (2) determine whether large concentration fluctuations affect the chemical outcome; and (3) assess whether current modeling paradigms (first order turbulent closures) used in regional and urban photochemistry models can be justified in their neglect of the large concentration fluctuations characteristic of the convective boundary layer.
Summary/Accomplishments (Outputs/Outcomes):
Traditional views of the convective boundary layer (sometimes described as the well-mixed layer) would imply that all trace gases and other scalar quantities are vertically well-mixed. However, this is not the case on short time scales. Recent intensive field programs in the southeastern United States under the Southern Oxidant Study (SOS) have made state-of-the-art measurements of vertical profiles of volatile organic compounds (VOCs). These observations made within the convective boundary layer (CBL) have shown that significant structure exists in profiles of isoprene. Although this has caused some to question our understanding of the average behavior of the CBL and its modeling (e.g., Ordnance, et al., 1994), Norris, et al. (1998) showed that these structures are likely due to incomplete averaging relative to the long turbulent time-scales in the CBL. The CBL turbulence contains large coherent eddies that can lead to large concentration fluctuations, which are difficult to remove by averaging over the time and space scales available for observation. Thus, this project funded by the U.S. Environmental Protection Agency (EPA) is directed at addressing three fundamental questions:
· How do we determine appropriate averaging strategies for reactive gases in the CBL?
· Do these large concentration fluctuations have any impact on the ultimate chemical solution, especially through segregation effects and nonlinearities in the chemical system?
· Are the first-order turbulent closure paradigms currently used in regional and urban photochemistry models justified in neglecting these large concentrations fluctuations?
Several different tasks were completed as part of the present study. A model named LESchem was built by directly coupling a 3-dimensional large-eddy simulation (LES) and photochemistry model. This was accomplished by integrating the SMVGEAR II chemistry solver into the RAMS mesoscale model. A condensed isoprene chemical mechanism, with 45 trace gases, 77 kinetic and 15 photolytic reactions, was incorporated into LESchem to the provide a moderately realistic boundary layer photochemical reaction set. Additionally, the Carbon Bond IV mechanism has been incorporated into UAHCTM, a 1-dimensional photochemical model of the University of Alabama in Huntsville. UAHCTM was used to generate trace gas:mixing ratio profiles for initializing the LESchem simulations.
A realistic midday CBL was simulated with LES, using the LESchem model. The dynamics of the LES-generated midday CBL were statistically analyzed to the second moments and successfully compared with observations to assess the realism of the modeled CBL. These same LES CBL dynamics were regenerated during each of the coupled LES-photochemical simulations conducted, using the full chemistry.
The LESchem model was tested using two horizontally homogeneous, idealized photochemical scenarios. The first scenario specified that NO and isoprene are both uniformly co-emitting from the surface into trace gas conditions representative as of August 4, 1991, for Giles County. This first scenario represents a large homogeneously forested area on a summer afternoon emitting isoprene from the forest canopy and nitric oxide produced by microbial activity in the soil of the forest floor. The second photochemical scenario is identical to the first, except that isoprene alone was emitted from the surface into an initial NO-rich CBL, and the NO surface emission was turned off. The second scenario represents a fresh urban plume rich in nitric oxide that has advected over an adjoining forested area. For each of the two hypothetical photochemical scenarios, coupled LES and first-order closure mesoscale simulations were completed. These simulations also were performed with and without dry deposition.
The volume-averaged trace gas:mixing ratio differences found between the LES and mesoscale versions of the photochemical scenarios are most likely due to the more active dynamics and photochemistry occurring in the entrainment zone around the top of the CBL. The buoyant thermals of the LES impact the inversion base at the CBL top. Their vertical momentum causes them to overshoot, which then causes intermittent mixing of available reactants, thus initiating active chemistry in the entrainment zone. Small countergradient species fluxes also can contribute to the LES entrainment zone photochemistry. The first-order closure mesoscale version of the same scenario has small eddy diffusivity (K) values at the top of the CBL, so little mixing of the reactants takes place there, thus limiting the overall reactivity in this region. In addition, countergradient fluxes cannot be simulated using K-theory.
Horizontally averaged profiles of the intensity of segregation Is calculated for 52 reactant pairs of the LES versions of the photochemical scenarios revealed that the segregation can work not only to hinder the mean reaction rate, but also to enhance the reactivity, depending on the reactants involved. For some reactant pairs that react on a time scale shorter than the turbulent eddy time scale, a negative Is indicates that the turbulent eddies hinder the overall reaction rate by not mixing the reactants quickly enough. On the other hand, some reactant pairs have a positive intensity of segregation in the CBL, which means that their reactivity is enhanced by the turbulent premixing of the reactants toward higher-than-normal concentrations (i.e., positive mixing-ratio deviations) inside buoyant thermals where they can react in close quarters at a rate faster than the mean reaction rate in the boundary layer. These reactant segregation processes, which can only be simulated in the coupled LES, will most likely become more important in simulations with patchy or otherwise heterogeneous surface emissions and characteristics.
For the simple hypothetical photochemical scenarios modeled in the present study, it was found that first-order turbulence closure (K-theory) performs reasonably well at modeling the large scale ensemble-averaged vertical diffusion of trace gases when simulating the photochemistry of the midday CBL. If the time and space scales of interest are regional and generally homogeneous, then the K-theory approach to air quality predictions is probably adequate for the task.
The concentration fluctuations simulated in the coupled LES versions of the two photochemical scenarios demonstrated important implications regarding the appropriate averaging times needed to gather representative trace gas measurements in the CBL. To avoid misleading biases in the trace gas observations, longer spatial or temporal averages may be required to take representative measurements of the highly reactive species for the scales of interest.
We also have provided a very practical example of the use of an LES-coupled chemistry system. In Houston, TX, there have been several continual examples of extreme but short-lived ozone events. Because these spike events have not replicated in standard regulatory models, there was some concern that such models were not applicable for air quality planning. Also, there was concern that some important process might be missing in either the emission inventory or in the physical and chemical modeling system. On the other hand, it could be argued that these spikes might not be found in the current air quality simply because of the coarse resolution of the physical model.
Our LESCHEM model is most likely the only tool of its kind that can operate in an LES mode with a regulatory-type chemical mechanism. Thus, we decided to try to apply our model to a few emission scenarios in the petrochemical complexes in the Houston area to determine whether we could provide some useful information to this important and practical air pollution problem. There are two emission scenarios that may be applicable. The first is a co-release of NO and ethene or propene, as if from a flare stack with incomplete combustion. In this case, there may be positive co-variances of reactants leading to positive segregation coefficients, speeding up the effective kinetic reaction rates. The second case is for NO and ethene or propene to be emitted separately so that the emissions would be separated in different parts of the large eddy structure. This would lead to a slower effective rate than in a coarse grid air quality model.
References:
Andronache C, Chameides WL, Rodgers MO, Martinez J, Zimmerman P, Greenberg J.
Vertical distribution of isoprene in the lower boundary layer of the rural and
urban southern United States. Journal of Geophysical Research 1994;99(D8):16,989-17,000.
Norris WB, McNider RT, Song A, Herwehe JA. The role of averaging time in interpreting observations made in the convective boundary layer (CBL). In: Proceedings of the Measurement of Toxic and Related Air Pollutants Symposium, Air and Waste Management Association and the U.S. Environmental Protection Agency, Cary, NC, September 1-3, 1998.
Journal Articles:
No journal articles submitted with this report: View all 4 publications for this projectSupplemental Keywords:
ambient air, atmosphere, troposphere., RFA, Scientific Discipline, Air, Mathematics, tropospheric ozone, Atmospheric Sciences, Ecology and Ecosystems, environmental monitoring, phototchemical modeling, ambient ozone data, isoprene emission algorithm, air quality data, air sampling, air pollution models, photochemistry, atmospheric chemical cycles, chemical kinetics, atmospheric monitoring, turbulent chemical interactions, ambient aerosol particles, convective boundary layer, trace gas measurementProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.