W. R. Barchet
The DOE Research Aircraft Facility at the Pacific Northwest National Laboratory (PNNL) is dedicated to the fulfillment of DOE and national goals related to understanding atmospheric processes as they relate to the DOE's environmental missions and the global environment. Central to this facility is an advanced sampling platform, the PNL Grumman Gulfstream I (G-1), and its flight crew, technical and engineering support staff, and state-of-the-art instrumentation.
The G-1 is a large twin turboprop with performance characteristics of contemporary "in
production" aircraft. It is capable of measurements to altitudes approaching 30,000 feet over ranges of 1500 nautical miles, and can be operated at speeds that enable both relatively slow sampling and rapid deployment to field sites throughout the world. The aircraft is configured for versatile research applications. It accommodates a variety of external probes for aerosol and turbulence measurements and internal sampling devices for measuring atmospheric composition and state variables. The G-1 has sufficient cabin volume, electrical power and payload capabilities, and flight characteristics to accommodate a large variety of instrument systems and experimental equipment configurations. Internal instrumentation is mounted in removable racks that enables rapid reconfiguration. Data acquisition is performed by an onboard computer system that can accommodate a wide range of analog and digital inputs. Acquired data are processed post-flight to incorporate calibration data and user-specified formatting. Output can be provided on PC-compatible disks or magnetic tape in a format acceptable to the user.
Priority for G-1 use is given to users from the DOE Atmospheric Chemistry Program. Proposals stating research objectives and requesting aircraft time must be submitted to the DOE for review by a DOE-empowered Research Aviation Facility Advisory Panel as far in advance as practical. Technical proposals to DOE should include a request for aircraft support. The panel reviews the technical feasibility of the proposed research flights and makes recommendations to DOE on the award of flight hours. After the DOE awards flight hours, the research aircraft facility coordinates the approved requests and schedules flight time.
Although DOE/OHER provides flight hours without cost, projects using the facility must estimate the following costs: (1) landing fees, (2) hangar rental, (3) ground support facilities, and (4) crew time and expenses for a PNNL flight crew of two pilots and one engineer. This information will be used in the review of the proposed request for flight hours.
The research aircraft facility does not cover the cost of engineering studies and airframe modifications needed for custom installation of project-specific instrumentation. Such costs must be budgeted for in the proposal to DOE. The staff of the research aircraft facility will assist the user in estimating these costs upon request. DOE covers all base operating costs (hangar rental and utilities, airport fees, calendar-based maintenance, flight crew training and proficiency maintenance, and flight crew and engineering staff hours) and all operating, maintenance, and logistical costs for up to 250 flight hours annually, including the off-site costs for the flight crew.
Requests for information and assistance can be directed to Dr. W. Richard Barchet, Pacific Northwest National Laboratory, P.O. Box 999, Richland, WA 99352, (509) 372-6158, e-mail wr_barchet@pnl.gov. Additional information on the G-1 can be found on the Internet at http://w3.pnl.gov:2080/atmos_sciences/
C. M. Berkowitz, E. G. Chapman, J. M. Hubbe
B. J. Finlayson-Pitts
Recent non-specific measurements of gaseous inorganic chlorine in the marine boundary layer suggest the presence of reactive chlorine, and measurements of ozone destruction during Arctic springtime have suggested the presence of reactive bromine. Modeling studies also suggest the present of halogens in reactive gaseous form. A number of investigators have proposed mechanisms for reactive halogen formation in the gas phase from sea salt. Reactions of chlorine and bromine could have a very significant influence on the levels and lifetimes of hydrocarbons, ozone, dimethyl sulfide, and cloud condensation nuclei in marine air. For example, the reaction of chlorine atoms with alkanes is 100 to 1000 times faster than the reaction of hydroxyl radicals with alkanes. Because of the potential importance of gaseous halogen reactions to tropospheric ozone and CCN, we initiated a study to develop and apply atmospheric sampling tandem mass spectrometry to investigate molecular halogens in the marine boundary layer. The findings of this work indicate that atmospheric pressure ionization tandem mass spectrometry can monitor molecular halogens in air with high specificity and sensitivity. The detection limits are 15-20 ppt for Cl2 and 5-10 ppt for Br2. Field studies indicated that the nighttime near-surface concentrations of Cl2 in marine air ranged from less than 50 to about 150 ppt; during the daytime, the concentration of Cl2 was less than 10 ppt. The field data suggest the presence of gaseous bromine, but at levels below the quantitative detection limit. For the future, an unexplained Cl2 artifact during airborne operation must be resolved, and some improvement in the sensitivity for Br2 is desirable.
*Battelle, Columbus Ohio
We have identified a reaction product from the isoprene and chlorine atom reaction that is unique to chlorine atom chemistry. This product, 1-chloro-3-methyl-3-butene-2-one (CMBO) should serve as an important marker for chlorine atom chemistry in the marine boundary layer. Other products, such as methyl vinyl ketone have also been confirmed as reaction products. In addition, a molecular formula consistent with 2-methylene-3-butenal was detected, however, further studies are needed to confirm this product.
J. B. Weinstein-Lloyd
J.H. Lee
Peroxides were measured from 6/29/95 to 7/20/95 near Nashville, TN, during the summer intensive of the Southern Oxidants Study. We deployed a three-channel continuous flow peroxide analyzer for the first time aboard the DOE G-1 aircraft. A modified version of two
channel analyzer designed by Lazrus and Kok, this instrument employs Fe(II)-benzoic acid and p-hydroxyphenylacetic acid (pOHPAA) reagents to discriminate between H2O2, methyl hydroperoxide (MHP) and hydroxymethyl hydro-peroxide (HMHP). Mean concentrations of the three peroxides for the entire measurement period were 2.59, 1.81 and 1.20 ppbv, respectively. H2O2 was correlated with the product of ozone and water vapor concentration above the boundary layer, as expected from the production mechanism. No such correlation was observed for organic peroxides. The evolution of peroxide concentrations immediately after sunrise was followed on two days. Both days showed fairly uniform H2O2 profiles on the earliest flight legs, with concentrations decreasing with time. This is contrary to the expectation that mixing from aloft would result in an increase in the concentration of a species subject to nocturnal deposition. Concentrations of the organic peroxides, however, showed an increase with time after sunrise, suggesting different mechanisms for surface deposition of these species. Vertical peroxide profiles exhibited different patterns. The most common one, occurring on 10 out of 14 flights, was characterized by [H2O2] > [MHP] and [MHP] similar to or slightly higher than [HMHP] and increasing slightly with altitude. The remaining flights exhibited [MHP] greater than or similar to [H2O2], with much lower [HMHP].
The G-1 and NOAA P-3 flew intercomparison runs on 6/29/95, 7/6/95 and 7/13/95. The two instruments employ different inlet systems, geometries and flow rates, and rely on fundamentally different chemistry to assess the partitioning of total peroxide into H2O2 and ROOH. Comparison of the pOHPAA channels, after correction for differences in response time, shows remarkably close agreement for most of the flight data. Agreement is poor for H2O2 and ROOH concentrations, with the BNL system showing generally higher concentrations of ROOH and lower lower concentrations of H2O2 than the NOAA instrument. Analysis indicates that the observed differences lie outside experimental error, but the results agree well enough that mathematical modeling is unlikely to identify one or the other measurement as correct. These results highlight the need for a formal intercomparison of techniques for the analysis of H2O2 and individual organic peroxides.
S.R. Springston and L.J. Nunnermacker
The available suite of instrumentation for the continuous measurement of ambient gases includes techniques for six individual species: O3, NO, NO2, NOy, CO, and SO2. To meet the current and anticipated requirements of the Atmospheric Chemistry Program, the capabilities of these techniques have been enhanced. Extensive tests were conducted to confirm the validity of measurements under a variety of flight conditions. Finally, ambient measurement data have been processed and are available for dissemination from two recent ACP programs and a major interagency collaboration. Instrument refinements have been necessitated by demands for faster response and enhanced detection limits. Oxides of nitrogen are measured with a three-channel instrument based on chemiluminescence designed and built at BNL. Laboratory studies recently revealed a strong correlation between baseline drift and ambient water vapor. This drift compromised the instrument accuracy whenever the water vapor concentration changed between instrument zeros. Adding water-vapor to the reagent stream reduced the effect and a small arithmetic correction for changes in ambient water-vapor levels improved the stability of the baseline to ~±10 pptv for a 10-s sampling period. The electronics and the flow path of a commercial SO2 analyzer were redesigned to improve the response time by 3X to ~15 s for 95% response. The existing commercial CO analyzer was only marginally suitable for aircraft operations because of low sensitivity and poor stability. The stability was improved by operating the sample cell at constant pressure and flow. The signal to noise ratio has also been improved.
During recent field programs, instrument performance was characterized on the ground and in flight during actual sampling missions. All four instrument systems were challenged in a performance audit on the ground. Challenges were conducted in both zero air and by the method of standard addition in ambient air. Preliminary results of the independent audit indicate no significant measurement errors. Extensive performance tests were also conducted in flight by the instrument operators. These tests were designed to reveal systematic problems due to pressure, ambient water vapor or other interferences as well as to confirm the results of standard ground calibrations. Within experimental error, instrument response in flight was confirmed to be identical with that measured during ground calibrations. Importantly both the NO2 and NOy converters showed less than ~1% variation in efficiency under all flight conditions.
P. V. Doskey and W. Gao
The objectives of this investigation are to analyze (1) the effect of energy-related trace chemicals on the photochemistry of the atmosphere and (2) the long-range transport of these substances and their transformation products over global scales. The approach involves a combination of instrument development, measurements of nonmethane organic compound (NMOC) profiles above terrestrial and marine surfaces from an aircraft, modeling efforts, and analysis of satellite data. Measurements of NMOCs are made in situ by a gas chromatograph designed for use aboard an aircraft and also by analysis of whole-air samples collected in stainless steel canisters. Chemical measurements and satellite data will be used in large-scale numerical models to investigate air
surface exchange rates of the NMOCs.
A cryogenic preconcentration high-resolution gas-chromatographic system was designed, assembled, and tested aboard the Battelle G-1 aircraft for measurement of C2-C6 nonmethane hydrocarbons (NMHCs). Features of the system include (1) simultaneous sampling into a canister and the gas chromatograph, (2) ozone removal with ascarite prior to NMHC preconcentration on fused silica beads at -185°C, (3) removal of the oxygenated hydrocarbons and high molecular weight NMHCs by a polar capillary precolumn, (4) separation of the C2-C6 NMHCs by high resolution gas chromatography with a porous polymer capillary column, (5) NMHC detection with a flame ionization detector, and (6) data acquisition with a laptop computer. All gases to the gas chromatograph are controlled with mass flow controllers to eliminate pressure effects. Air samples can be collected and analyzed on a 14-min cycle with a detection limit of 10 pptC.
NMOC measurements using stainless steel canisters were made along a transect and in vertical profiles above a forested region in Massachusetts during July of 1996 as part of the Atmospheric Chemistry Program's participation in the North American Research Strategy for Tropospheric Ozone (NARSTO)-Northeast 1996 Summer Ozone Study. Isoprene concentrations at an altitude of 500 m varied from 0.4-1.5 ppbv along the transect and attained maximum levels above the Harvard Forest. Concentrations of methanol covaried with isoprene and were 12-25 ppbv. These NMOCs are emitted from woody vegetation at similar rates. Isoprene and methanol concentrations above the Harvard Forest were 0.7 and 22 ppbv, respectively at an altitude of 500 m and decreased to 0.2 and 14 ppbv, respectively at an altitude of 2000 m. In comparison, the concentrations of ethane, a long-lived NMHC of anthropogenic origin, were about 1 ppbv and invariant within the boundary layer. Biogenic emission algorithms, landuse maps, and satellite data will be used in large-scale numerical models to derive vertical profiles of these NMOCs for the Harvard Forest location. These profiles will be compared with the chemical measurements to improve the predictive capabilities of the models.
J. S. Gaffney and N. A. Marley
Nitrogen oxides (NOx) play important roles in the chemistry of the troposphere on urban, regional , and global scales. Nitrogen dioxide and peroxyacyl nitrates (PANs) are of particular interest. NO2 is a key species in the formation of tropospheric ozone and in the production of nitric acid which can play important roles in aqueous aerosol chemistries. PANs are in equilibrium with nitrogen dioxide and the peroxyacyl radicals formed from organic hydrocarbon oxidations. This equilibrium is strongly temperature-dependent with PANs being quite stable at colder temperatures. PANs can thus play an important role in transporting NOx over long distances at colder temperatures. Rapid response, high sensitivity instrumentation that can be used for G-1 aircraft measurement of these species is needed for future ACP field studies.
Measurement of PANs and NO2 using luminol chemiluminescent detection has been demonstrated. This has been accomplished by coupling a short (3-m) capillary column, DB-1 to the detection cell via a inert six-port gas sampling valve and using helium as a carrier gas to perform gas chromatographic separation and luminol detection for the analysis. Nitrogen dioxide, peroxyacetyl nitrate (PAN), peroxypropionyl nitrate (PPN), and peroxybutryl nitrate (PBN) can be separated and analyzed in under one minute. Sensitivity studies in the laboratory have demonstrated <30 ppt detection sensitivities for NO2 and PANs. The instrument uses a Hewlett-Packard integrator to collect the data which is interfaced to a laptop computer using Peak 96 (HP) software. By making use of the thermal equilibrium of PANs with nitrogen dioxide, the use of a prereactor for PAN calibration as nitrogen dioxide was demonstrated. The sensitivity of nitrogen dioxide using helium as a carrier gas was found to be 1.5 times greater that PANs. The instrument range can be adjusted by photomultiplier operating voltage, the sample loop volume, or combination of both to allow measurement sensitivities appropriate for urban, regional, or global scenarios. The instrument can also be used for laboratory studies of NO2 and PANs heterogeneous loss on surfaces. Preliminary studies of NO2 and PAN loss on soils and carbonaceous soot indicate that the water content of the surfaces can play an important role in the loss mechanisms.
Future plans are to rack mount this instrument for testing on the G-1 after testing of multicapillary columns (1-m) are completed. We expect to perform direct comparison with the GC/ECD PANalyzer and a NOx chemiluminescent monitor on a series of test flights during late summer of 1997.
P. H. McMurry
A three-pronged effort (instrument development, atmosphereric field measurements, laboratory research) aimed at better understanding the nucleation and growth of sub-10 nm atmospheric particles is described. This cumulative research effort has been funded by DOE ACP and by NASA. The DOE funding has been used to support instrument development, the Mauna Loa measurements, and some laboratory research.
Key conclusions of our atmospheric and laboratory work are:
€Particles are formed orders of magnitude faster than is predicted by the binarytheory of nucleation for water and sulfuric acid. This binary theory is the most commonly used theory for atmospheric aerosol models.
R. Leifer, L. Kromidas, B. Albert, R. Knuth and P. Roiz
A. Muyshondt
The Environmental Measurements Laboratory (EML) has built an aerosol sampling package containing a newly designed impactor to characterize the aerosol composition and size distribution in and above the marine boundary layer during flights of the G-1 aircraft. In addition, we evaluated a newly designed shrouded aircraft aerosol probe to improve aerosol collection from the ambient air at sampling speeds of 100 m sec-1. The resultant data will be used for characterizing chemical changes in the marine aerosol. Specifically, losses of chlorine and increases in the sulfur of sea salt particles collected for different size ranges will be looked at and compared to changes in gaseous chlorine measurements independently measured by other researchers on the G-1 aircraft.
The DOE Research Aircraft Facility
Pacific Northwest National Laboratory
wr_barchet@pnl.gov
Development and Application of a New Method to Measure Molecular Halogens in Marine Air
C. W. Spicer, R. A. Plastridge
Battelle, Columbus Ohio
spicerc@battelle.org
Pacific Northwest National Laboratory
University of California, Irvine
Unique Chlorine Containing Organics from the Isoprene and Chlorine Atom Reaction: Potential Markers for Chlorine Atom Chemistry in the Marine Boundary Layer
T. Nordmeyer, W. Wang, M. Ragains, C.W. Spicer*, and B.J. Finlayson-Pitts
University of California, Irvine
bjfinlay@uci.edu
Concentrations of H2O2, HMHP and MHP Determined aboard the G-1 Aircraft During the SOS Nashville '95 Campaign
SUNY/Old Westbury
jlloyd@bnl.gov
Brookhaven National Laboratory
Instrumentation Improvements for Real-Time Measurements Aboard the G-1
Brookhaven National Laboratory
srs@bnl.gov
An Aircraft Gas Chromatograph for Hydrocarbon Measurements
Argonne National Laboratory
paul_doskey@qmgate.anl.gov
Development of a PANalyzer for the G-1 Based on Luminol Chemiluminescence
Argonne National Laboratory
jeff_gaffney@qmgate.anl.gov
Nucleation and Growth of Atmospheric Nanoparticles
University of Minnesota, Minneapolis
McMurry@me.umn.edu
€Observed particle formation rates depend on sulfuric acid squared, suggesting collision-controlled nucleation. The binary theory predicts a much stronger dependence of particle formation rates on sulfuric acid (roughly ~sulfuric acid to the tenth power).
€Measured nanoparticle growth rates are about ten times higher than can be explained by sulfuric acid condensation, suggesting that other phenomena contribute to growth.
€Nucleation was observed downwind of Macquarie Island. It is our hypothesis that ammonia emissions from penguin rookeries lead to high nucleation rates by ternary nucleation involving ammonia, sulfuric acid, and water.
€Nucleation over the Southern Ocean was routinely observed in the outflow regions of convective clouds during daytime.
€The sulfuric acid equilibrium vapor pressures above aqueous solutions containing mixtures of sulfuric acid and ammonium sulfate drop markedly (~3 orders of magnitude) when the ammonium to sulfate ratio exceeds about 0.8.
€The mass accommodation coefficient of sulfuric acid vapor on particle surfaces varies between 0.65 and 1.0 on sodium chloride and ammonium sulfate and between 0.19 and 0.31 on stearic acid.
Marine Aerosol Characterization using EML's G-1 Aircraft Aerosol Sampling System and Scanning Electron Microscope
Environmental Measurements Laboratory, U. S. Department of Energy
Rleifer@eml.doe.gov
University of Arkansas, Fayetsville, AR 72701
Among the major concerns when sampling aerosol on a moving platform, such as the G-1 aircraft, are whether a sampled aerosol is representative of the true ambient aerosol and is the aerosol modified as it passes through the aircraft sampling probe. To evaluate the performance of our shrouded probe, we installed a size distribution instrument (Met One, Optical Particle Counter (OPC) ) in the aircraft to isokinetically sample the airstream from the probe. A second size distribution instrument (Particle Measurements Systems, PCASP), operated by the Pacific Northwest Laboratory is mounted in the free airstream to provide a reference for comparison.
Case Studies During Field Experimental Campaigns
H. N. Lee and R. J. Larsen
Environmental Measurements Laboratory, U. S. Department of Energy
hnlee@eml.doe.gov
Vertical profiles of Radon-222 concentrations measured from September 3-6, 1995 in the northeastern U.S., using a new radon instrument designed for aircraft measurements, are presented. A vertical diffusion model was employed to simulate the distributions of Radon-222 concentrations by using time-dependent profiles of vertical eddy diffusivity in the lower atmosphere. To determine these profiles, O'Brien's simple formulation was applied using the height of the boundary layer and the turbulence velocity scales. The model produced good agreement between the calculated profiles of Radon-222 concentrations and the measurements. Model simulations were also calculated using different values of the vertical eddy diffusivity above the boundary layer. Using the value of the vertical eddy diffusivity above the boundary layer equal to 10% of its maximum boundary layer value in the model produced better vertical profiles of the calculated Radon-222 compared with the measurements within the boundary layer. The results demonstrate the feasibility of using the naturally occurring radionuclide, Radon-222, as a tracer in the study of vertical mixing processes in the lower atmosphere. These data are useful for model validations.
An airborne atmospheric pressure ionization mass spectrometer was successfully used to measure formic and acetic acids over the western North Atlantic Ocean. The sensitive, specific and real
time measurements provided by the mass spectrometer yielded the first highly resolved tropospheric profiles of these gases. Formic and acetic acid mixing ratios showed substantial vertical variation, varying by factors of up to 13 and 6, respectively, within profiles extending from the boundary layer into the free troposphere. Substantial horizontal variation was also observed at constant altitude transects within both the boundary layer and the free troposphere.
Airborne Measurements of Carboxylic Acids Over the Western North Atlantic
E. G. Chapman, D. V. Kenny*, K. M. Busness, J. M. Thorp, C. M. Berkowitz, and C. W. Spicer*
Pacific Northwest National Laboratory
eg_chapman@pnl.gov
*Battelle, Columbus, OH
 |
| Figure 1. Vertical distribution of FA observed on the WP3, with histograms shown in the insets for above and below 2000 m. |
 |
| Figure 2. Composite diurnal dependence of FA (triangles) and the calculated fraction of radicals produced from FA (dots). The curve represents a bi-weight average of the data. |
 |
| Figure 3. Correlations among concentrations of FA, GA, GL, and MG observed on the G1. |
 |
| Figure 4. Composite diurnal dependence of the distribution of nitrogen oxides components. Box plots show the fraction concentrations of individual components |
Analysis of SOS data has focused on three questions:
1. What is the relative importance of natural vs. anthropogenic hydrocarbons for ozone production?
2. What is the O3 production efficiency in the Nashville urban plume in the context of NOx emissions?
3. Is ozone production in the Nashville urban plume and in the surrounding area limited by the availability of NOx or hydrocarbons?
Although this analysis is not yet completed, we are able to make the following statements with regard to the above questions: We find that average hydrocarbon reactivity (as measured by the product of the OH reaction rates times the hydrocarbon concentration) is highest near the surface, and decreases monotonically with altitude. Isoprene, which is a naturally emitted hydrocarbon, dominates the reactivity near the surface, becomes less imprtant with altitude and is unimportant in the free troposphere. CO is the second most important hydrocarbon at all altitudes . Non-isoprene hydrocarbons are relatively unimportant at all altitudes. Although on average, isoprene dominates hydrocarbon reactivity near the surface, we also find that isoprene concentrations exhibit a high degree of spatial variabilty due no doubt to an uneven distribution of isoprene sources. The dominance isoprene reactivity in the boundary layer has important implications with respect to ozone control strategies.
The efficiency of ozone production in the context of O3 produced per NOx molecule consumed was examined in the Nashville urban plume by comparing upwind downwind differences in ozone and NOz concentrations viz., e = (delta)O3/(delta)NOz, where (delta)O3 is the upwind-downwind difference in the ozone concentration and (delta)NOz is the corresponding difference in the NOx oxidation product concentration. In crosswind transects of the Nashville urban plume it was generally found that the value of e was highest at the edges of the urban plume (8-10), and lowest at the center of the plume (2-4) where concentrations are highest. This finding is consistent with our understanding of ozone formation chemistry wherein at high NOx concentrations (i.e., in the center of the plume), NO2 efficiently competes with hydrocarbons for the radicals that propagate the chemical processes producing ozone, shortening chain lengths, and decreasing O3 production efficiency.
NOx/hydrocarbon limitations to the rate of ozone production in the Nashville urban plume and surrounding atmosphere were examined using the indicator variable (total-peroxide/NOz), where total-peroxide is the sum of the hydrogen and organic peroxide concentrations, and NOz is the NOx oxidation product concentration. In the Nashville urban plume, it was generally observed that O3 formation rates were limited by the availability of reactive hydrocarbons at the center of the plume where NOx concentrations are high, and to tend towards NOx limted ozone production rates at the edges of the plume. A gradual transition between hydrocarbon and NOx limited regimes was observed as the plume aged and mixed with the surrounding atmopsphere. Ozone production in the background atmosphere was always NOx limited. Hydrocarbon limited behavior in the center of the plume suggests an ozone control strategy of lowering hydrocarbon emissions. However, since hydrocarbon reactivity is dominated by natural hydrocarbons (see above), it may be concluded that this strategy would not be effective.
R. Alvarez
Formation and distribution of O3 and related species in the Nashville urban plume were characterized by flights of the DOE G-1, and the NOAA Casa 212 on July 3, 1995. Flights were made during mid-day both upwind and downwind of the city; winds were from the southwest transporting the urban plume to the northeast. The boundary layer was well-defined at about 1300 m (upwind) and 1700 m (downwind, as indicated by temperature, dewpoint and O3 profiles). Background concentrations observed during the upwind transect were relatively uniform with [CO] ‰ 280 ppbv, [O3] ‰ 72 ppbv , and [NOy] ‰ 5 ppbv; isoprene concentrations were high and variable in the boundary layer and very low aloft. On the downwind transects a well defined plume was sampled at approximately 40 and 75 km northeast of the urban center. The maximum observed O3 concentration was about 120 ppbv, which was associated with maximum concentrations for CO of 490 ppb and for NOy of 15 ppbv. A high degree of correlation between CO and O3 (r2 = 0.87) and between NOy and O3 (r2 = 0.92) was observed, consistent with an urban source for this plume. Boundary layer concentrations of isoprene were depleted to negligable values. Formation of O3 in the urban plume will be examined in the context of NOx and hydrocarbon concentrations and also the concentrations of photochemical product species such as peroxides and formaldehyde. Measurements of the horizontal and vertical distribution of O3 concentrations of Nashville region by the CASA 212 will be used to provide a context for interpretion of the in-situ measurements by the G-1.
W. J. Shaw and C. M Berkowitz
Measurements of chemical and meteorological variables were made at several altitudes over a surface chemistry site near Nashville, TN, during the 1995 Southern Oxidants Study (SOS). The measurements were designed to reveal the effects of turbulent mixing on atmospheric chemistry. They were made under conditions of clear skies and light winds during the morning transition from nocturnal stable stratification to the afternoon convective mixed layer. Early morning ozone mixing ratios measured by the aircraft were approximately 70 ppb, while those measured by surface instrumentation were about 25 ppb. Corresponding to growth of the morning turbulent layer, surface ozone values steadily increased with time until they matched the 70 ppb values aloft by mid-morning. The mixing ratios of isoprene at altitudes above the surface increased by several orders of magnitude with the onset of turbulence at each measusurement altitude. The slope of O3 as a function of NOy-NOx for each of the flight legs was also sensitive to the presence of turbulence. Measurements from non-turbulent flight legs yielded slopes that were considerably steeper than those from measurements made in turbulence. This study showed that the concentration of ozone precursors aloft is clearly dependent on the presence of turbulence, and turbulent mixing could explain the evolution of ozone concentrations at the surface. In general, conclusions regarding pollutant concentrations must account for both chemical and local dynamic processes.
A Case Study of Urban Plume Evolution for July 3, 1995, Nashville, TN.
L. J. Nunnermacker, P. H. Daum, J. H. Lee, J. Weinstein-Lloyd, Y-N Lee, S. R. Springston and L. Newman
Brookhaven National Laboratory
phdaum@bnl.gov
NOAA/Environmental Technology Laboratory
Vertical Mixing and Ozone Concentrations from SOS '95
Pacific Northwest National Laboratory
wj_shaw@pnl.gov
Modeling Associated with Major Field Experiments
Three pollution events during the North Atlantic Regional Experiment (NARE) field campaign in the summer of 1993 are examined by using aircraft measurements and a Lagrangian particle dispersion model that is coupled to an atmospheric mesoscale model. The mesoscale model employs a four-dimensional data assimilation technique to limit forecast error growth so that a high degree of confidence can be placed in the meteorological fields. By tracking hundreds of thousands of passive particles over a 15-day period, the meteorological processes responsible for the observed layer of high ozone mixing ratios can be identified. The timing and location of the simulated peak particle concentrations correspond quite well to the observed maximum ozone mixing ratios, indicating that the mesocale model reproduces many of the meteorological mechanisms associated with pollutant transport and diffusion during the 15-day period.
During the 1992 field study of Northern Atlantic Regional Experiment, relatively localized O3 maximum were observed by aircraft flight along a horizontal sampling path between Nantucket and Halifax on several occasions. While sampling at a constant altitude between 1500 and 2000 m, the aircraft encountered a plume having O3 mixing ratio twice as much as the values that had been measured prior to entering the plume. Concurrent with this increase in O3 mixing ratio was a pronounced drop in the dew point. This feature occurred over a horizontal distance of approximately 200 km. Prior to descending to a lower elevation, a second O3 maximum/dew point minimum appeared to be encountered. Several hypotheses concerning the origin of this localized O3 maximum along a horizontal path are rejected, and the periodic nature of the phenomena suggests that they could be explained by propagating internal gravity wave at the top of the boundary layer. Observational support of this hypothesis are soundings from surrounding stations that had temperature profiles resemble those that are typical for gravity wave occurrence. A simple numerical model is used to simulate the variations of ozone concentration, temperature, and dew point caused by perturbation horizontal and vertical wind fields associated with traveling internal gravity waves and to illustrate how the localized O3 maximum are formed from downward displacement of air parcels from the free troposphere by these waves. The observed ozone oscillation corresponds reasonable well in amplitude and phase with the simple model prediction.
*SUNY Old Westbury
Ozone production is conventionally represented by models that simulate the chemical interactions of perhaps 100 species undergoing multiple hundreds of reactions. In our calculations and field work over the past decade we have focused on the broad organizing features that underlie these hundreds of individual processes. Our presentation at the 1996 ACP meeting illustrates several ways of looking at oxidant production that follow from a conservation principal that the production rate, Q, of free radicals (i.e., odd hydrogen, OH, HO2, and RO2) must be equal to their destruction rate:
Q = L(H2O2) + L(ROOH) + L(R+R') + L(R+NOx)
where L stands for loss rate. The first 2 terms in (1) represent loss of radicals due to formation of H2O2 and organic peroxide. L(R+R') is a small term due to Radical-Radical reactions such as OH+HO2. L(R+NOx) represents all Radical+NOx reactions that remove radicals such as OH+NO2-->HNO3. Under low NOx conditions the major radical loss processes are due to peroxide formation. Equation (1) can then be solved for total peroxy radical concentration in terms of Q, which in turn can be estimated from measured concentrations of O3 and HCHO, the principal precursors for radicals. This line of reasoning led to the prediction that [H2O2] should be correlated with Q or factors that Q depends such as [O3][H2O] or even [H2O]. This prediction has been verified by many groups. Ozone production, P(O3), is due to the reaction of peroxy radicals with NO which leads us to a formula valid for low NOx conditions that gives P(O3) in terms of quantities that can be measured.
P(O3) = k/(2keff)1/2 Q1/2 [NO]
where k is the rate constant for HO2 or RO2 + NO and keff is an effective rate constant for peroxide formation. Eq (2) was applied to a month of data taken in Metter, GA. P(O3), so obtained, agreed to within measurement uncertainty with P(O3) obtained from the photostationary state relations. Under high NOx conditions Q in Eq.(2) becomes "Q-L(R+NOx)". In recent work we have applied the high NOx form of Eq (2) to data collected during the SOS Nashville field program. By differentiating this equation with respect to either [NO] or [Hydrocarbon] we get a simple analytic equation that describes whether O3 production is NOx or hydrocarbon limited and by how much. The sensitivity of P(O3) to [NO] and [Hydrocarbons] depends on a single variable, the fraction of radicals that are removed by reactions of the type Radical+NOx. Our results provide theoretical justification for using the "indicator species" ratio [H2O2]/[HNO3] as a means of diagnosing NOx and hydrocarbon limitations to O3 production.
Meteorological Factors Affecting Ozone Profiles over the Western North Atlantic
Doran, J. C., Zhong, S., and Berkowitz, C. M.
Elevated Localized Ozone Maximum Associated Gravity Waves
Xindi Bian and Carl M. Berkowitz, and Sharon S. Zhong
Pacific Northwest National Laboratory
x_bian@pnl.gov
O3 Production Rates: Observation-Based Analysis
L. Kleinman, P.Daum, J. Lee, Y.-N. Lee, L. Nunnermacker, S. Springston, J. Weinstein
Lloyd*, and L. Newman
Brookhaven National Laboratory
kleinman@bnl.gov
Stratospheric Chemistry Modeling
*Chinese Institute for Climate Change, Beijing, China
Air enters the stratosphere through the tropical tropopause, is transported upward and poleward by the Brewer-Dobson circulation and planetary wave mixing, and descends back into the tropopause at middle and high latitudes. Recent observations of long-lived trace species (e.g. Trepte and Hitchman, 1992; Randall et al, 1993) indicate that mixing is limited between the tropics and the midlatitudes of the stratosphere. They also show that transport occurs during Rossby wave breaking events that lead to filaments being drawn from the tropical region into midlatitudes. In this paper, we address the question of how well the mass exchange between tropics and midlatitudes can be represented in a dynamically consistent 2-D model and to what extent parameters can be constrained by the requirement to reproduce available data of long lived trace species. We use a coupled 2-D model to describe the dynamics, chemistry and radiation of the stratosphere. The effects of Rossby wave mixing are included using externally specified coefficients, Kyy, applied consistently in the zonal mean momentum equation and the diffusion term of the tracer transport equation. Rossby wave mixing is reduced in the tropics which means that transport between the tropics and midlatitudes is controlled by advection by the residual circulation, predicted by the model. Meridional transport in the model is analysed and results of the model are compared with observations of long lived tracers made by instruments aboard the Upper Atmospheric Research Satellite (UARS) and with data from the Atmospheric Trace Molecule Spectroscopy experiment (ATMOS). To test the accuracy of the model simulation of transport in the lower stratosphere, model generated correlations of N2O and CO2 are compared with measurements obtained during the Stratospheric Photochemistry, Aerosols and Dynamics expedition (SPADE) and the Airborne Southern Hemisphere Ozone Experiment and Measurement for Assessing the Effects of Stratospheric Aircraft (ASHOE/MAESA). It is shown that the seasonally reversing advective flow calculated by the model, together with rapid diffusive transport in the winter hemispheres yields distributions of long-lived tracers in good agreement with observations in the tropics and the midlatitudes. The seasonal characteristics of the measured correlations of CO2 and N2O are well captured by the model. The CO2 data require that eddy mixing coefficients in the Southern Hemisphere must exceed a value of 0.5X105 m2 sec-1. The simulation of these data is sensitive to the parameterizations used in the model troposphere and to the dissipative processes assumed in the tropical stratosphere. However, it is insensitive to pattern changes in the circulation induced by extratropical forcings in the stratosphere, such as changing Kyy by a factor of 2 (as long as the threshold value is exceeded), which implies variations in the overturning rate of the order of 30%.
Temporal variations in the abundance of stratospheric ozone reflect a suite of natural dynamical influences in addition to chemical influences attributable to human activity. Isolating the anthropogenic components require a careful analysis of the natural variability. The quasi-biennial oscillation (QBO), associated with interannual changes in column ozone in excess of 5%, is particularly important in this context. We have reproduceda QBO in our two-dimensional model using the equatorial wave forcing parameterization of Gray and Pyle [1989]. The oscillation has a period of 28 months with maximum easterly and westerly zonal winds of -25 m s-1 and +13 m s-1, respectively. The oscillation in the zonal wind in the tropics induces a QBO in ozone in the subtropics which is synchronized with the seasonal cycle. Since the effects of planetary wave mixing are specified as fixed eddy diffusion coefficients, Kyy, the model is unable to reproducethe observed high latitude QBO in ozone. We find that the buildup of ozonein the subtropical anomaly is produced by advection of ozone out of thetropics and down from higher altitudes by the QBO
induced meridional circulation. The subtropical anomaly in ozone is synchronized with the seasonal cycle because the induced circulation is stronger in the winter hemisphere, reflecting seasonal asymmetry in wave forcing across the equator. As observed by Hitchman et al. [1994], the influence of the QBO on transport out of the tropics in the model can be separated into an "upper" and a "lower" regime. When the easterly winds are a maximum in the lower stratosphere, transport out of the tropics in the lower stratosphereis enhanced, while in the middle stratosphere it is reduced. On the other hand, when the westerly wind is a maximum in the lower stratosphere transport out of the tropics in the lower stratosphere is reduced, while in the middle stratosphere it is enhanced.
Effects of the Quasi-biennial Oscillation on the Zonally Averaged Transport of Tracers
D. B. A. Jones, H. R. Schneider, and M. B. McElroy
Harvard University, Cambridge, MA
mbm@io.harvard.edu
hrs@io.harvard.edu
| Author | Title |
| Wang et al. | Atmospheric Ozone as a Climate Gas |
| Grossman et al. | Chemical and Radiative Factors Affecting the Calculations of the Global Warming Effects of Ozone |
Ivar S. A. Isaksen
University of Oslo, Norway
The objectives of the project are to improve understanding of the chemical and dynamic processes that control mid-latitude O3 in the lower stratosphere and free troposphere, and to develop improved predictions of future O3 changes in these regions and their influence on (and response to) future climate changes. Three tasks have been conducted:
1. The observed O3 climatology together with radiation and climate models were used to evaluate the importance of atmospheric O3 to radiation balance, and to study how the observed O3 changes during the decade 1980s and 1990s affect the radiation fluxes and climate as well as their comparison with increasing other greenhouse gases.
2. Climate-chemistry interactions were studied based on the NCAR GCMs (CCM1, GENESIS, CCM3) and the University of Oslo's chemical-transport models (CTMs). First, the CTMs were used to calculate the tropospheric O3 distribution using a variety of emission scenarios for NOx, HC, and CO. Next, the O3 and other gases consistently calculated from the CTMs will be used as input to radiation model to study the effect on radiative forcing and UV flux reaching the surface. These experiments were conducted for both current climate state and the global warming climate state induced by increasing concentrations of greenhouse gases (such as CO2, CH4, N2O and CFCs).
3. We continue the development of an interactive climate-chemistry global model. Progress are being made to incorporate a chemical code used in the University of Oslo's 3-D CTM into CCM3. During the initial phase, we focus on the interface of inputs and outputs as well as the transport of trace species. In addition, in parallel to the coupled model development, we examine the consistency in the physical and chemical processes important to both climate and chemistry.
The project so far has eleven (11) publications in the refereed journals and conference proceedings.
J. E. Penner and J. J. Walton
Univ. of Michigan, Ann Arbor
We have developed a global, three-dimensional model capable of simulating the cycles of many tropospheric species, including: O3, OH, HO2, H2O2, NO, NO2, PAN, HNO3, CO, CH4, isoprene, HCHO, C2H6, C3H8, C4-5 alkanes, C6-8 alkanes, C2H4, C3H6, higher alkenes, benzene, toluene, xylene, and other higher hydrocarbons. Ozone (O3) is an important tropospheric pollutant, acts as a greenhouse gas, and also damages materials. The hydroxyl radical (OH) is formed when O3 photolyzes. These two species are highly coupled and control the troposphere's oxidizing capacity.
Our model includes source emissions, chemistry, transport, and wet and dry deposition. The model physics is driven by meteorology supplied by a General Circulation Model, and is updated at least every 4 - 12 hours. Our past work has used meteorology from NCAR's CCM (Community Climate Model). Recently, we have began updating our model to accept fields from ECHAM (a climate model devloped at the Max Planck Institut fur Meteorologie). The ECHAM fields include a more thorough boundary layer treatment, a diurnal cycle, and water as a prognostic variable. Coupling with ECHAM will allow us to include heterogeneous chemistry. A preliminary simulation of the sulfur cycle which included the aqueous conversion of SO2 by H2O2 and O3 showed that gaseous H2O2 concentrations were decreased. More work will continue in this area in the future.
We have used the model to predict the impact of energy-use scenarios on chemically and radiatively important species. Below we show how the concentration of O3 decreases over the North Atlantic Ocean if fossil fuel combustion sources of NOx are decreased by 50% in North America. Concentrations of O3 decrease from North America across the North Atlantic by 5 - 15 ppbv, depending on location.
A global three-dimensional chemical transport modeled, called MOZART (Model of Ozone and Related Species in the Troposphere), is used to calculate abundances of chemical species and their seasonal evolution MOZART is a fully diurnal model that calculates the time evolution of about 45 chemical species from the surface to the upper stratosphere. It accounts for surface emissions of source gases, wet and dry deposition, photochemical transformations, and transport processes. The dynamical variables are provided the by National Center for Atmospheric Research (NCAR) Community Climate Model (CCM3) at T42 resolution (2.8°x 2.8°) and 18 levels in the vertical. Calculated concentrations of chemical species are generally in fair agreement with observations. Monthly mean distributions of ozone calculated in July are shown near the surface Owing to intense photochemical production, high concentrations of ozone (above 40-50 ppbv) are predicted during July over North America, Europe, and East Asia. The seasonal variation of ozone is also compared to observational data. It indicates that the calculated ozone range on average from 31 ppbv in winter to 40 ppbv in summer, generally close to observed values. This newly developed model will be a useful tool for the study of tropospheric chemistry.
T. Pierce
A large number of atmospheric trace gases have an important source or sink associated with terrestrial vegetation (e.g., biogenic emission of NMHCs and dry deposition of various pollutants). For a given meteorological condition and chemical concentration, surface chemical fluxes can change significantly with vegetative conditions to cause a general difficulty in estimating their values over important regional and global scales. This project is to take advantage of remote sensing data from the advanced very-high-resolution radiometers onboard NOAA polar orbiting satellites to derive vegetative parameters critical to modeling atmosphere-vegetation exchange of chemical substances. An operational algorithm, which has been developed to retrieve satellite remote sensing data, was evaluated with available ground measurements. Coupling with numerical modules for biogenic emission and dry deposition, the algorithm can be used to effectively describe spatial, seasonal, and annual changes in surface chemical fluxes that have a source or sink driven by terrestrial vegetation. The current version of the algorithm can be readily coupled with processed AVHRR satellite data at a spatial resolution of up to 1 km and a temporal resolution of up to one day. The goal of this work is to incorporate this algorithm to improve the estimation of surface biochemical fluxes in regional and global chemistry models
A Global Three-dimensional Chemical Transport Model (MOZART): Simulation of Ozone Distribution in the Troposphere
G. Brasseur, D. Hauglustaine, X. Tie, and S. Walters
National Center for Atmospheric Research, Boulder, CO
brasseur@ncar.ucar.edu
Detecting Spatial, Seasonal, and Annual Changes in Atmosphere-Biosphere Exchange of Chemical Substances by using High-Resolution Satellite Data
W. Gao, M. L. Wesely, and J. Qiu
Argonne National Laboratory
weigang_gao@qmgate.anl.gov
Atmospheric Research and Exposure Assessment Laboratory, U.S. EPA
G. R. Carmichael and F. Potra
University of Iowa, Iowa City
gcarmich@icaen.uiowa.edu
Interactions between ozone, aerosols, and ultraviolet radiation are investigated using a detailed radiation model combined with a three-dimensional atmospheric chemistry model. These models are joined with Automatic-Differentiation software to enable desired sensitivities to be calculated on-line with the radiation/chemistry computations. The combined methodology is used to investigate the following:
(a) UV-B radiation at the earth's surface as a function of changes in ozone and aerosols in the stratosphere and troposphere; and
(b) effects of changes in solar actinic flux on the photochemicaloxidant cycle of the troposphere.
The project team consists of Gregory R. Carmichael, Professor & Chair, Department of Chemical & Biochemical Engineering and Co-Director of the Center for Global & Regional Environmental Research; Florian Potra, Professor of Mathematics and Computer Science; and a group of graduate students from Chemical & Biochemical Engineering, Applied Mathematics and Computer Science. The work to date has focused on three areas: investigations of the couplings between tropospheric chemistry and radiative transfer using the STEM-II model; implementation and use of automatic differentiation; and improvements in the numerical integration of the models.
A major goal of this research is to quantify the interactions between UVR, ozone and aerosols. One method of quantification is to calculate sensitivity coefficients. A novel aspect of this work is the use of Automatic Differentiation software to calculate the sensitivities. We have demonstrate the use of ADIFOR for the first time in a multidimensional framework. Automatic Differentiation is used to calculate such quantities as: sensitivities of UV-B fluxes to changes in ozone and aerosols in the stratosphere and the troposphere; changes in ozone production/destruction rates to changes in UV-B flux; aerosol properties including loading, scattering properties (including relative humidity effects), composition (mineral dust, soot, and sulfate aerosol, etc.); and the sensitivity of UVR transfer and ozone formation/destruction to changes in photochemical parameters including cross-sections and quantum yields. The combined radiation/chemistry model offers an important test of the utility of Automatic Differentiation as a tool in atmospheric modeling.
Since the combined sensitivity analysis with detailed radiationand atmospheric chemistry modeling is computationally very complex. Inrecognition of this fact, and taking advantage of the strengths of ourinterdisciplinary research group, we have also addressed severalcomputational issues. The primary focus is on the computationalconsiderations of ADIFOR. However the accuracy of the ADIFOR sensitivitiesdepends on the numerical methods used in STEM-II and the radiation code.Therefore we have been investigating ways to improve the performance(speed and accuracy) of the key algorithms in the models. We havedeveloped new numerical techniques that can provide the most accuratesolution with the least computational efforts for the integration of thechemistry equations when used with forward mode AD. We have also developedsome general purpose processing tools to enable researchers to buildsensitivity analysis into their research activities.
P.-T. Wu and C. H. Bischof
A comprehensive sensitivity analysis of a multi-phase atmospheric chemical mechanism has been conducted under a variety of atmospheric conditions ranging from remote marine to heavily
polluted atmospheres. The ADIFOR automatic differentiation techniques were applied to evaluate the local sensitivity of concentrations of 145 chemical species in various phases with respect to 407 model parameters including gas-phase reaction rate constants, aqueous-phase reaction rate constants, Henry's law constants, mass accommodation coefficients, uptake coefficients, species initial concentrations, temperature, relative humidity, cloud water content and cloud droplet size. The sensitivities of indicators such as O3/(NOy-NOx), H2O2/HNO3, HCHO/NOy, and H2O2/HCHO with respect to various model parameters were derived. The results show that the sensitivities of these indicator species provide a good marker to evaluate O3-NOx-hydrocarbon sensitivity. Higher sensitivities for VOC reactions and lower sensitivities for NOx reactions indicate hydrocarbon
sensitive O3. The main reaction pathways in the gas, aqueous and aerosol phases and the most influential model parameters to photochemistry and heterogeneous chemical processes are identified. Our base simulation results show that the presence of clouds and aerosols reduces the total oxidizing capacity and alters the O3-precursor relations through changing species concentrations. Cloud chemistry is the dominant heterogeneous process under the remote and marine atmospheres. Aerosols are important scavenger for gaseous species and contribute to the heterogeneous perturbation to photochemistry in polluted atmospheres. The sensitivity analysis shows that cloud chemistry is most sensitive to changes in cloud water content, and the magnitude of the effect of heterogeneous aerosol processes on photochemistry is highly sensitive to changes in uptake coefficients of individual species such as O3 and HCHO.
*This project was initially funded by DOE through PNNL's Laboratory Directed Research and Development program and is currently supported by DOE's Atmospheric Chemistry Program.
H. A. Rabitz
The application of sensitivity analysis, SA, tools to atmospheric models and the development of new techniques that are applicable to mult-dimensional models were presented. SA was applied to a stratospheric ozone chemistry model and a marine boundary layer chlorine chemistry model. Our investigation of the stratospheric chemistry involved the calculation of the total model output uncertainty of the ozone model due to the uncertainty in the individual kinetic rates. This was performed for air parcels at 45 degrees north latitude at 20, 30, and 40 km. The output uncertainties were 35%, 24%, and 24%, respectively. The reactions that most significantly contributed to the uncertainty were subsequently identified and the top 5 are:
1) HO2 + O3 -> OH + O2 + O2
The use of a direct steady state method to calculate the sensitivity coefficients was also presented. This provides a fast (seconds, instead of hours) means to calculate approximate sensitivities when the model in near a pseudo steady state. These results can then be used for a quick investigation of the chemical mechanism and agreed well with the more rigorous methods. The marine boundary layer chemistry (MBL) model of PNNL was analyzed using sensitivity analysis to identify the key chlorine containing species and reactions in the model. This information may be used reduce the size of the chemical mechanism, without sacrificing the quality of the results. The current model contains 264 reactions (93 contain chlorine) and 127 species (47 contain chlorine). The smaller mechanism may be incorporated into a 3-D chemical transport model. With the 3-D model, the effects of MBL air on coastal environments may be investigated. This work is in collaboration with Elaine Chapman and Carl Berkowitz of PNNL. Here we have identified the key reactions affecting ClONO2, ClO, and Cl at sunrise using sensitivity coefficients.
ClO + NO2 -> ClONO2
The important initial conditions affecting ClONO2, ClO, and Cl were also identified using Green's functions are: O3, Cl2,isoprene, PAN, HCl, methane, and alkanes. We presented a new tool that learns the input-output responses of a system (either from a model or data) and then uses these responses 1) to make extremely fast and accurate output predictions for the system, given a new input parameter set and 2) to elucidate the system sensitivities and non-linearity's. In essence, the tool encapsulates all the behavior of the system and relates it as a Fully Equivalent Operational Model, FEOM. This new tool was applied to the NASA 2-D atmospheric chemistry model and the resulting FEOM made spatially resolved ozone concentration predictions to within 5% of what the full model would calculate. The FEOM took less than a second which compares to four hours for the full model. Additionally, utilizing the response information separately, the model output ozone response to the key reaction rates were identified. It was shown that the FEOM algorithm captured information on the radiative feedback as well as quantifing the very strong gradients in the ozone sensitivity to NOx reactions.
Sensitivity Analysis of a Multi-Phase Chemical Mechanism using Automatic Differentiation*
Y. Zhang and R. C. Easter
Pacific Northwest National Laboratory
y_zhang@pnl.gov
Argonne National Laboratory
Application and Development of Sensitivity / Uncertainty Analysis Tools in 0-D and 2-D Chemical Kinetic Models
J. A. Shorter
Mission Research Corporation
jshorter@mrcnh.com
Princeton University
2) N2O + O(1D) -> NO + NO
3) OH + HO2 -> H2O + O2
4) O(1D) + M -> O + M
5) H2O + O(1D) -> OH + OH
Cl2 -> Cl + Cl
NO2 + NO2 + sea salt -> ClNO
Direct light scattering of solar radiation and enhanced cloud albedo due to anthropogenic sulfate aerosol are hypothesized to lead to a shortwave radiative forcing of climate by about -1 W m-2 global average (that is, exerting a cooling influence), comparable but opposite to longwave forcing by anthropogenic greenhouse gases. Inclusion of this forcing in climate models requires improved estimates of the geographical distribution of this forcing, and in turn of sulfate concentrations. This can be achieved only with models. Here we report results of applying our subhemispheric scale transport and transformation model for aerosol sulfate and comparison with observations, and outline the approach to incorporating aerosol microphysical properties in the model.
The model has been described previously (Benkovitz et al., 1994). The distinguishing feature of the model is that it is driven by observationally derived meteorological data, thereby allowing comparisons with observations not just at specific locations, as is customary, but also at specific times. These comparisons indicate that a substantial fraction of the temporal variance at any given location is accounted for by the temporal variability of the model, all of which is due to temporal variability in the meteorological field driving the model. Examples of comparisons of SO2 and sulfate mixing ratios are presented illustrating the temporal variability and the ability of the model to capture this variability. The comparisons include observations from multiple observing stations within a single model grid cell (1.125 deg X 1.125 deg) illustrating substantial subgrid spatial variability. The model permits attribution of sulfate to source regions and to chemical production mechanism.
It is intended to represent aerosol microphysical properties in the model in terms of the moments of the radial size distribution. The moments accurately represent observable physical properties of the aerosol (McGraw et al., 1995). Under limited (but important) conditions the moments evolve according to a closed set of differential equations and may thus be treated as chemical species (McGraw and Saunders, 1984), but in general this set of equations is not closed. Recently we have developed an extension of this approach that permits the set of equations to be closed by an approximation involving gaussian quadratures (McGraw, 1996). Initial tests indicate that this approximation is highly accurate. This quadrature method of moments should thus be of great utility in representing aerosol dynamics in regional to hemispheric or global scale models.
Benkovitz C. M., Berkowitz C. M., Easter R. C., Nemesure S., Wagener R. and Schwartz S. E. (1994) Sulfate over the North Atlantic and adjacent continental regions: Evaluation for October and November 1986 using a three
dimensional model driven by observation-derived meteorology. J. Geophys. Res. 99, 20725-20756.
McGraw R. (1996) Description of atmospheric aerosol dynamics by the quadrature moment of methods. Aerosol Sci. Tech., submitted.
McGraw R., Huang P. I., and Schwartz S. E. (1995) Optical properties of atmospheric aerosols from moments of the particle size distribution. Geophys. Res. Lett. 22, 2929-2932.
McGraw R. and J. H. Saunders (1984) A condensation feedback mechanism for oscillatory nucleation and growth. Aerosol Sci. Technol. 3, 367-380.
S. Nemesure and R. Wagener
Satellite and surface-based aerosol optical depth data and surface concentration and deposition data for August 1994 were used to evaluate PNNL's global tropospheric aerosol model. The model treats four lognormal aerosol size modes (accumulation, Aitken, coarse dust, and coarse sea-salt), with the accumulation and Aitken modes consisting of sulfate, methanesulfonic acid, organic carbon, and black carbon. The aerosol model is coupled with PNNL's version of the NCAR Community Climate Model V2, with winds and temperature nudged by observed meteorology. AVHRR satellite data were analyzed to obtain daily aerosol optical depths using analysis procedures developed at BNL. Surface-based measurements of aerosol optical depth were identified and acquired. The number of measurement sites has increased greatly over 1992-1994, especially in the U. S. Surface concentration and wet deposition data from several networks in North America and Europe were utilized.
Model simulations were performed for August 1994 at 300 km resolution. The model simulates observed monthly-averaged spatial patterns fairly well, but has less skill at simulating day to day temporal variations. Simulated aerosol optical depths over several oceanic regions having an anthropogenic influence were higher than satellite observed values. This appears to be primarily due to the simulated aerosol water content being too high, and secondarily to the dry-aerosol concentrations themselves. Aerosol optical depths in the Saharan dust plume off the west coast of Africa were also overestimated. Simulated surface sulfate concentrations were low over the eastern U.S. and high over Europe, and several processes contribute to this, including precipitation scavenging and aqueous sulfate production in low-level non-precipitating clouds. The evaluation identified several areas for potential improvement of model performance.
S. D. Schery, B. Nemetz, F. Yarger, and P. T. Wasiolek
Dry deposition velocities for nanometer-size particles in the outdoor environment are poorly known. Published values are limited and conflicting. We have found evidence for higher than expected deposition velocities for 1 to 2 nm diameter ultrafine radon progeny in the analysis of gradients observed at several different locations, including Mauna Loa Observatory, Hawaii. In an effort to make more direct measurements of deposition velocities of ultrafine particles, we are developing a relaxed eddy accumulation system that simultaneously collects ultrafine and attached (100 - 500 nm) particles on diffusion screens and filters. We have been testing the system at a pasture site west of Socorro, NM, which has an approximate aerodynamic roughness of 30 cm. Although adequate counting statistics are a problem, and refinement of our techniques continues, preliminary measurements indicate deposition velocities for 1-nm-diameter particles that increase, as expected, with horizontal wind speed. For a 4-m reference height, deposition velocities appear to be in the range 10 to 20 cm/s for horizontal winds around 6 m/s. These deposition velocities are larger than predicted by some conventional models.
D. J. Cziczo, J. B. Nowak, J. H. Hu, and J. P. D. Abbatt
Motivated by the importance of tropospheric aerosols to both issues of climate and atmospheric chemistry, the infrared extinction spectra of model tropospheric aerosols have been measured as a function of relative humidity. Experimentally, submicron-sized aerosol particles are generated with an atomizer and spectroscopically monitored at they flow through an absorption cell. By monitoring absorption features due to either constituent ions or water molecules, we can infer both the physical phase and the chemical composition of the aerosols. It is observed that: (i) solid (NH4)2SO4 and NaCl aerosols exhibit deliquescence, respectively, at 79% and 75% relative humidity, very close to their thermodynamic values, (ii) (NH4)2SO4 and NaCl liquid aerosols exhibit crystallization at relative humidities of 33% and 43%, respectively, well below their deliquescence points, (iii) NH4HSO4 aqueous aerosols remain in the liquid state to relative humidities as low as 2%, far below the thermodynamic deliquescence humidity of 39%, (iv) artificial sea water aerosols show strong H2O absorption features at low relative humidities, most likely arising from the solid hydrates of Mg2+ salts. For understanding behavior of both laboratory and ambient particles, these observations illustrate the importance of determining the water content of the particles. Specifically, for (NH4)2SO4 and NaCl particles, the water content is expected to be low at relative humidities below the crystallization point and high at higher relative humidities. On the other hand, NH4HSO4 and artificial seawater aerosols will contain signficant quantities of water down to extremely low relative humidities, most probably as a liquid in the case of NH4HSO4 and as salt hydrates for artificial seawater.
Global Tropospheric Aerosol Model Evaluation for August 1994
R. C. Easter, S. J. Ghan, N. S. Laulainen, L. R. Leung, and Y. Zhang
Pacific Northwest National Laboratory
rc_easter@pnl.gov
Brookhaven National Laboratory
Progress on Development of a Relaxed Eddy Accumulator for Measurement of Dry Deposition Velocities of Nanometer-size Aerosol Particles
(Guest Presentation)
New Mexico Tech, Socorro
schery@nmt.edu
Infrared Spectroscopy of Model Tropospheric Aerosols as a Function of Relative Humidity: Observation of Deliquescence, Crystallization and Hydrate Formation
(Guest Presentation)
The University of Chicago
jabbatt@midway.uchicago.edu
A DRIFTS Study of the Reaction of NO2 and HNO3 with Synthetic Sea Salt and MgCl2.6H2O
S. Langer, R. S. Pemberton, D. T. Quach, and B. J. Finlayson-Pitts
University of California, Irvine
bjfinlay@uci.edu
The kinetics and mechanisms of the reactions of oxides of nitrogen with airborne sea salt particles have largely been studied using NaCl as a model for sea salt. Indeed, these heterogeneous reactions have been the focus of many studies due to their potential to act as precursors to atomic halogens in the atmosphere. It is not clear, however, how well these studies translate to the analogous reactions with the complex mixture of inorganics found in sea salt particles. To address this concern, and to determine if NaCl is an appropriate model for sea salt, we have performed comparative studies with synthetic sea salt. To this end, the reactions of NO2 and HNO3 with a synthetic sea salt mixture containing NaCl as well as hydrated salts and trace amounts of transition metals have been investigated. The reactions were studied using diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) to follow the formation with time of nitrate on the salt surface at 298 K in the presence of either He or air as the carrier gas. The infrared bands due to surface nitrate formed during the reaction of NO2 are shown to be similar to those from the reaction of MgCl2.6H2O, a major hydrate in the mixture which was subsequently used as a surrogate for all of the crystalline hydrates in the mixture. Indeed, the separate contribution to the surface nitrate infrared band envelopes from the components in a mixture of NaCl and MgCl2.6H2O could clearly be distinguished when compared to spectra for the reaction of NO2 with the individual salts. The reaction order of 1.6 for the NO2-NaCl reaction in both He and air is in agreement with previous work (Vogt, R. and Finlayson-Pitts, B.J., J. Phys. Chem., 1994, 98, 3747; 1995, 99, 13052); it is assumed that for an order >1.6 that N2O4 is the reacting species. The reaction order of 1.8 for the NO2-synthetic sea salt reaction in He is consistent with that for NaCl, however, the order of 1.2 for the reaction in air suggests that NO2 may also contribute as a reacting species. The strong influence of hydrate species on the reaction is clearly evident and suggests that NaCl may not, in fact, be an appropriate model for sea salt.
High resolution FTIR gas phase spectra have been obtained on peroxyacetyl nitrate (PAN) peroxypropionyl nitrate (PPN) and peroxybutryl nitrate (PBN) and on their peracid precursors. Liquid phase spectra have also been obtained on these compounds using the cylindrical internal reflectance accessory (CIRCLE). Use of internal reflectance crystals with these compounds has been very difficult, especially with the peracids, due to their extremely reactive natures. All crystal substrates studied showed some surface corrosion when exposed to the PAN's and peracids with the exception of sapphire. However, sapphire is not transparent across the entire wavelength range of interest. The material found to give best results was AMTIR (a form of ZnS). A comparison of FTIR spectra of the PAN's taken in gas phase with those taken in tridecane gave preliminary evidence for existence of a cyclic structure for these compounds in lipids. UV spectra has also been obtained for the peracids in acid solution which demonstrate UV cross sections similar to their acid counterparts. These preliminary studies indicate that peracid photolysis is likely to be slower than aldehydes or ketones in the gas phase. FTIR studies of the photolysis of organic aldehydes are reported which yield evidence for the formation of the peracids under low NO conditions.. The observation of peracid formation during photolysis of the aldehydes and lack of formaldehyde observed in the photolysis of acetaldehyde and propionaldehyde in the absence of NOx is consistent with the observed UV absorbance for peracetic acid.
The chemiluminescent reaction of ozone with hydrocarbons is the basis for an organic detector which has been shown to yield higher sensitivities than a conventional flame detector for a number of compounds including the natural hydrocarbons and oxygenates. The temperature dependence of this reaction has been studied and has been suggested as a basis of separating signals from different organic compound classes. As well, addition and abstraction reactions for various organics can be probed using this technique as it appears to correlate with OH reactivities.
Recently, we have taken five size-fractionated aerosol samples using an eight stage cascade impactor, and have determined the 210Pb and 210Po content. Using the disequilibria between the two 222Rn daughters, an estimate can be made for the aerosol atmospheric lifetimes. Once formed 210Pb (22.3 yr. half-live) decays to 210Bi (5 day half-live) which decays to 210Pb (138 day half
live) which in turn decays to the stable 206Pb isotope. All of these species attach themselves to fine aerosols (<2m diameter). The age of the aerosol is estimated by determining the activity ratio of 210Po/210Pb. The atmospheric residence time (Tr, days) is calculated using the following equation:
In previous ACP supported work, we used this method to estimate the apparent age of the wet deposited aerosols in rain, snow, and hail. The ages ranged from 10-40 days. The size-fractionated aerosol ages ranged from 30-170 days. The youngest ages were associated with a sample that was known to contain substantial local soot from diesel sources. The oldest age was observed in the 0.1-0.3m size range, consistent with modeling results for diffusional loss and gravitational settling. Similar measurements made in the 60's and 70's found ages ranging from 40-150 days for whole aerosol samples. The observation made at that time was explained by soil contamination or coarse aerosol sources (wind blown dust, etc.) or "bomb" derived fallout which led to elevated 210Po/210Pb ratios and the apparently long live times. These results on fine aerosols rule out a wind blown dust source and bomb-derived 210Po would have decayed since the testing. These data lead to an alternative explanation. Fine carbonaceous soot may have longer lifetimes than the more hydroscopic sulfate aerosols and may be responsible. This hypothesis has significant implications in modeling aerosols in tropospheric chemistry and in global climate change. Current models use a 9 day lifetime for lower tropospheric aerosols and a 23 day lifetime for upper tropospheric aerosols. Also, carbonaceous soot is a strong absorber in the UV, visible, and IR.
Reactions of chlorine atoms with biogenic hydrocarbons are suspected to play an important role in the formation and decay of O3 in the polluted marine boundary layer. Chlorine atoms are formed at least in part by photolysis of the reaction products of nitrogen oxides on sea salt particles, which are produced from sea spray. Typical Cl concentrations are believed to be in the range of 103-105 cm-3. Hydrocarbons in coastal include for example isoprene from both ocean phytoplankton and plants and (alpha) -pinene, (beta) -pinene from vegetation. The goal of this project is to measure the rate constants of the reactions of chlorine atoms with biogenic hydrocarbons: isoprene, limonene, (alpha) -, (beta) -pinene, myrcene, 3-carene, p-cymene, etc. by using two different techniques: relative rate and a fast flow discharge system. By using two different techniques, systematic errors can be reduced. Some of the rate constants have been measured by the relative rate technique with air as carrier gas. A comparison with measurements in N2 as carrier gas shows that the rate constants in air are slightly high, possibly due to interferences from OH chemistry. The fast flow discharge system has been optimized for the detection of Cl and will also be used to determine the absolute rate constants and their temperature dependence. The measurements will be also be extended to other biogenic hydrocarbons, including those containing oxygen. The rate constants in N2 show that the lifetime of the hydrocarbons due to reactions with chlorine atoms are similar to their lifetimes due to the reactions with OH radicals for typical atmospheric conditions. As the OH radical concentration is small in the early morning hours and the nighttime oxidant NO3 has already been photolyzed, Cl atoms may be the major oxidizing species in the marine boundary layer. These reactions of chlorine atoms with biogenics may lead to a faster production of ozone in the polluted marine troposphere than with OH chemistry alone. This can cause such effects as a change of the location of the peak ozone concentration in coastal cities like Los Angeles.
E. Swartz, O. Rattigan and P. Davidovits
The objective of the Aerodyne-Boston College ACP project is the investigation of heterogeneous kinetics related to ozone formation and destruction processes in clouds and aerosols relevant to the upper atmosphere. Specifically, we have studied the interaction of formaldehyde (CH2O) with aqueous and acid solutions in order to evaluate the possibility that dissolution and reaction of CH2O in atmospheric aerosols can perturb nitrogen oxide speciation that controls ozone chemical reactivity. There has been much speculation that heterogeneous reaction with CH2O could reduce nitric acid (HNO3) to regenerate NOx species (e.g., NO, NO2) in the upper atmosphere. These has involved laboratory measurement of CH2O heterogeneous reaction via gas-liquid interactions utilizing liquid droplets, gas bubbles, and a stirred liquid reactor. These different experiments probe heterogeneous kinetics on time scales from milliseconds to minutes, spanning a wide range of fast and slow heterogeneous processes. For example, time dependent uptake in sulfuric acid droplets (on a millisecond time scale) confirmed previous observations of enhanced CH2O uptake in concentrated acid, determining the solubility of protonated CH2OH2+. Uptake of CH2O into bubbles (on a ~1 sec time scale) confirmed reported aqueous hydrolysis kinetics, including acid and base catalytic rates. Most recently, reaction of CH2O with HNO3 in concentrated H2SO4 solution has been observed to evolve of HONO and CO2 products in the gas phase over a stirred liquid reactor. The reaction is autocatalytic in HONO concentration, with induction periods measured up to tens of minutes. While the autocatalytic behavior is not relevantto the atmosphere (HONO will evaporate from, rather than accumulate in atmospheric aerosol and cloud particles), experiments in progress will elucidatethe mechanism and kinetic models appropriate for inclusion of this heterogenous process in general atmospheric photochemical models.
J.T. Jayne, and D.R. Worsnop and C.E. Kolb
We describe results of experiments that measure the uptake of NH3 gas into liquid water and sulfuric acid solutions on millisecond time scales. Time dependent uptake measurements resolve the relative contributions of mass accommodation probability, gas and liquid diffusion, gas solubility, and aqueous reactivity. The following results are observed. In basic solution, (pH > 10) uptake is limited by physical solubility (Henry's law constant, H~100 M/atm), with enhanced uptake observed at t < 10-2 sec indicating the formation of a surface complex. At lower pH, uptake increases as NH3 reacts with H+ in solution. At low pH<2, uptake reflects the mass accommodation coefficient (~0.08 at 283K) which exhibits a negative temperature dependence consistent with a cluster nucleation model. In concentrated sulfuric acid solutions the NH3 uptake coefficient increases to near unity, independent of temperature, indicating enhanced surface reactivity of NH3 with H+. The implications of this work on gas / liquid partitioning of NH3 in the troposphere and stratosphere are discussed. Co-deposition studies in which an aqueous surface, initially at pH4, was simultaneously exposed to both gas phase NH3 and SO2 were also performed.
Fine Aerosol Residence Times Using Radon Daughters 210Pb and 210Po*
J. S. Gaffney and N. A. Marley
Argonne National Laboratory
jeff_gaffney@qmgate.anl.gov
(After Nevissi, A.E., (1991) J. Radioanal. Nucl. Chem., 148, 121-131.
*This work was funded by ANL-LDRD funds, and was a follow up to previously funded ACP work on wet deposition of 210Pb and 7Be.
Kinetics and Mechanisms of the Reactions of Atomic Chlorine with Biogenic Hydrocarbons
B. Buehler, C.J. Keoshian, J. Stutz, and B.J. Finlayson-Pitts
University of California, Irvine
bjfinlay@uci.edu
Knudsen Cell Studies of the Uptake of Nitric Acid on NaCl, Synthetic Sea Salt, and MgCl2.6H2O Surfaces at 298 K
D.O. De Haan and B.J. Finlayson-Pitts
University of California, Irvine
bjfinlay@uci.edu
Nitric acid plays a significant role in the chemistry of sea salt aerosol in the marine boundary layer by displacing chloride from particles as HCl, in competition with other NOx reactions which release chlorine in more active forms which destroy ozone. The results of Knudsen cell studies indicate that the reaction of gaseous HNO3 with both synthetic sea salt and magnesium chloride hexahydrate surfaces at 298 K are fast compared to our previously measured reaction probability with NaCl ( g = 0.014). The efficiency of nitric acid uptake depended on the existence of strongly adsorbed water at the salt surface, even at H2O vapor pressures well below the deliquescence point. It appears that NaCl is not a good experimental substitute for sea salt in this case, since the amounts of strongly adsorbed water at the surfaces are evidently much larger in our experiments on synthetic sea salt and MgCl2.6H2O surfaces. We hypothesize that the atmospheric reaction of nitric acid on sea salt may best be described using liquid-phase chemical equilibrium models, with a rate depending on humidity levels.
Heterogeneous Uptake of Formaldehyde in Aqueous and Concentrated H2SO4/HNO3 Solution: Solubility and Reactivity
D. R. Worsnop, J. T. Jayne and C. E. Kolb
Aerodyne Research, Inc.
worsnop@aerodyne.com
Boston College
Heterogeneous Accomodation and Reaction Kinetics for NH3 Uptake on Aqueous and Concentrated Sulfuric Acid Droplets
Q. Shi, E. Swartz and P. Davidovits
Boston College
paul.davidovits@bc.edu
Aerodyne Research, Inc.
 |
|
Table 1. Preliminary rate constants determined at the specified pH
| Compound | pH= 6.80 | 9.08 | 9.32 |
| KI, x109M-1s-1 | 2.34 ± 0.07 | 2.55 ± 0.07 | 2.59 ± 0.08 |
| NaNO2, x105 M-1s-1 | 6.41 ± 0.34 | 6.39 ± 0.10 | |
| Na,sub>2S + NaHS, x109 M-1s-1 | 2.94 ± 0.11 | ||
| Humic acidx105(g/L)-1s-1 | 2.98 ± 0.19 | 3.53 ± 0.13 | |
| Resorcinol, x108 M-1s-1 | 5.69 ± 0.14 | 10.8 ± 0.04 | |
| Guaiacol, x108 M-1 s-1 | 6.41 ± 0.41 | 13.1 ± 0.09 | |
| Ascorbic acid, x109 M-1s-1 | 1.64 ± 0.22 | 1.75 ± 0.22 | |
| Phenol, x109 M-1s-1 | 1.11 ± 0.07 pH =10.2 | 1.80 ± 0.15 pH=10.8 | |
L. Harrison, J. Berndt, G. Lala, P. Kiedron
Atmospheric Sciences Research Center, SUNY-Albany
LEE@solsun1.asrc.albany.edu
N. Laulainen
Pacific Northwest National Laboratory
A prototype ultraviolet radiation rotating shadowband spectroradiometer (UV-RSS) has been developed and tested to measure solar irradiance at wavelengths between approximately 295 nm and 340 nm. The device is a UV-CCD spectrograph, which provides greater spectral resolution, a discrete-continuous realization of the spectrum rather than a small number of passbands, equal or better wavelength registration accuracy, better radiometric stability, and lower cost compared either to existing grating monochromators or to an approach employing a prism spectrograph optic and a rotating low-precision shutter wheel in front of a photomultiplier tube. Outdoor tests in Albany indicate that the original objectives have been achieved and that the anticipated problems of achieving adequate sensitivity and out-of-band rejection have largely been overcome. These data are shown and compared to data taken by a double grating monochromator operated at the North American UV Spectroradiometer Intercomparison, Table Mountain CO, 1995. At the present the UV-RSS performs as well as or better than single-Brewer spectroradiometers except for the minimum detection limit. This prototype was a preliminary proof-of-concept assembly. Work is in progress to improve the instrument by substitution of UV-grade optical components, a redesign fore-optic optimized for this instrument, and an improved CCD array.
*Space Science Institute, Boulder, CO
A new inversion technique has been developed for SAGE ozone and aerosol data with particular emphasis on lower stratospheric ozone determinations. The determination of long-term („10 years) ozone changes in the lower stratosphere is an important scientific and social question. The SAGE data base, which extends back to 1979, is the longest high vertical resolution satellite ozone record in existence. The challenge in the lower stratosphere is to carefully estimate and remove the effects of aerosol extinction, a contaminant in the ozone extinction channel. Improper removal of this interference affects the accuracy of the ozone results. Our approach is two-fold: (1) develop a forward model which simulates in detail the SAGE instrument response; and (2) develop an inversion algorithm which has the capability of removing aerosol contamination from the ozone signal. Our results indicate that the current version of the SAGE II ozone data are biased by aerosol contamination. The figure shows the results of the DOE/LASP inversion compared to the standard SAGE II (or Langley Research Center, LaRC) results. The first three panels compare the zonal average ozone profiles from the DOE/LASP and LaRC inversions for three days corresponding to different aerosol loading conditions: low (left panel), moderate (second panel), and high (third panel). The fourth panel shows the percent difference between the DOE/LASP and LaRC algorithms, 100(LASP-LaRC)/LaRC, for the low (blue), moderate (green) and high (red) aerosol loading examples given in the first three panels. Both inversions compare well for low aerosol loading conditions, but show larger differences for the cases of higher aerosol loading. In particular, these results indicate that the LaRC algorithm consistently overestimates the ozone densities below about 18 km under moderate or high aerosol-loading conditions, consistent with previously published comparisons between the SAGE II data and balloon-borne ozonesonde data [Veiga et al., JGR 100, 9073-9090, 1996]. Validation of the DOE/LASP algorithm is in progress. The entire SAGE II database will then be inverted with the DOE/LASP algorithm, after which we will determine the ozone trends.
Comparison of DOE/LASP Inversion to LaRC Inversion
Comparisons of direct sun observational N-values at Dobson wavelengths showed very small biases between SBUV-2, SSBUV, and Dobson 83 surface-based observations. The comparisons of zenith sky umkehr N-values exhibited large differences in the C-pair N-values. It has been shown that total ozone derived from the double wavelength pairs AC, CD, and AD yielded the same value of total ozone, which is two percent larger than the Dobson AD double wavelength pair value. The source of the umkehr N-value differences was further investigated by using TOMS overpass ozone measurements to calculate zenith sky N-values for Dobson A, C, D, and SBUV-2 A and B wavelength pairs. The results showed the measured N-values agreed closely with those calculated. The zenith sky umkehr measurements at A, C, and D Dobson wavelength pairs have been inverted to obtain total column ozone values. Double wavelength pairs AC and AD yield total ozone which is one percent higher than Dobson 83 direct sun values. Single wavelength pairs yield total ozone values which are about four percent larger than Dobson 83 direct sun values. The double wavelength pairs reduce the effects of aerosols on the inverted ozone values. These results illustrate further that the SBUV type instrument measurements from space and the ground are not different from those from the world standard Dobson instrument No. 83 due to calibration errors and that zenith sky radiance measurements can be inverted to obtain very accurate total ozone values.
Our work with the Umkehr retrieval algorithm thus accomplished has led to insightful conclusions; however, a number of thoughts and questions in the general area of uncertainties and peculiarities in the performance of the two algorithms have surfaced. Our present work is aimed at probing deeper into the Umkehr phenomenon to gain satisfactory answers to our questions. Most of the scientific tools such as fast radiative transfer codes, fast computers are available. Moreover, stratospheric aerosol data, and ozone profile data obtained by satellites, aircraft, and ground-based observing methods are also available. Each ozone observing method has its strengths and weaknesses. The combination acts to improve the ultimate credibility of the scientific conclusions that are drawn from the analyses of the various data sets. The inclusion of the Umkehr method presents an inexpensive and important international contribution to the array of present methods for observing ozone profile.
Our investigative work with the Umkehr method for observing ozone profiles has focused on many facets of the method that need to be understood and taken into account. The analysis involves scrutinizing every aspect of the mathematical approaches that are followed in the development of ozone retrieval methods and corrections for the errors caused by elevated levels of stratospheric aerosols. The topics of our ongoing research activities that we have been conducting are listed as follows:
This project is working on analyzing existing and emerging data to determine changes in ozone and UV levels. This group, which includes two atmospheric scientists and two statisticians has a long history of deriving credible ozone trends and is now branching into UV as well. The most recent publication has been a re-analysis of the Robertson-Berger meter UV monitoring network. The results of this work, soon to be published in JGR, reveal problems with the data which make them inappropriate for trend analysis. Discrete changes in the data imply either instrumental or site changes which have a dramatic impact on the derived trends. This directly negates previous work which showed a decrease in UV for the U.S. Some of the authors of the original paper have been shown the more recent work by this group and have supported the results. More recent work has focused on determining the number of years to derive a trend. This work highlights the importance of continuity in datasets and ancillary data for the derivation and interpretation of trends. Future work will include an assessment of when the expected "turn-around" in ozone may be observed and what criteria should be used for assessing that such a turn-around is evident. It is best that such criteria be developed before any expected turn-around to lessen the likelihood of false claims of recovery or false claims of failure of the Montreal Protocol efforts.
*CIRES, University of Colorado, Boulder
Total ozone trends at northern midlatitudes are largest in winter. These trends significantly exceed the predictions of current 2D models that include heterogeneous chemical losses associated with increases in anthropogenic chlorine and bromine. In addition, the latter models predict trends that are largest at the pole while the observed winter trends are largest at 45 degrees north. By analyzing the longitude dependence of the observed ozone trends as well as the long-term variability of lower stratospheric temperature and geopotential height, evidence has been obtained that a large contribution to NH midlatitude ozone trends is the result of long-term changes in upper tropospheric and lower stratospheric circulation. Possible sources of the inferred circulation changes are decadal climate change in the troposphere and secondary dynamical feedbacks resulting from chemical ozone depletion. Using an empirical model for total ozone variability induced by planetary wave forcing of the lower stratosphere, the ``dynamical'' contribution to total ozone trends in the NH winter has been estimated. This contribution has a longitude dependence consistent with that of the observed trends and also has a latitude dependence that agrees with the observed trends, peaking at midlatitudes and decreasing at higher latitudes. Subtracting the estimated wave forcing contribution from the observed trends yields a residual meridional trend profile that agrees more closely with two-dimensional stratospheric model estimates.
In collaboration with X. Tie and G. Brasseur of NCAR, we have also recently investigated the impact of observed temperature trends on the rate of heterogeneous chemical loss of ozone in the NH. At 45 degrees north in February, the decadal ozone change predicted by the NCAR 2D model increases from about -3 Dobson Units to about -4 Dobson Units. At 75 degrees north, the predicted ozone change increases from about -18 DU to -22 DU. These calculations demonstrate the necessity of taking into account lower stratospheric temperature trends when estimating the chemical contribution to midlatitude and high-latitude ozone trends.
K. Stamnes
Ultraviolet (UV) quanta are sufficiently energetic to break many molecular bonds. In the troposphere, the resulting photo-fragments drive smog formation, control diurnal and seasonal cycles of oxidants such as ozone and hydroxyl radicals, and determine the lifetimes of many climatically important carbon, sulfur, and halogen species. In the biosphere, UV-induced damage to living tissues can impact adversely human health, and may cause complex alterations of terrestrial and marine ecosystems.
We have developed a global three-dimensional transport/chemical model of the stratosphere (called STARS: Study of Transport And Chemical Reactions in the Stratosphere) that includes a representation of the formation of Polar Stratospheric Clouds (PSCs) and heterogeneous reactions on the surfaces of PSCs and sulfate aerosols. The formation of the observed springtime "Antarctic ozone hole" is well reproduced by the model. Calculated ozone and chlorine concentrations are consistent with satellite observations. After the breakdown of the polar vortex in December, air with depleted ozone is transported to mid-latitudes in the Southern Hemisphere, resulting in a 2
4% ozone decrease at 50°S in December and a 1% decrease in the subtropics. Ozone-poor airmasses are also transported to the troposphere and produce a significant decrease in upper tropospheric ozone.
The New DOE/LASP Inversion Algorithm for SAGE Data
D.W. Rusch, C.E. Randall, M.T. Callan, M. Horanyi, S.C. Solomon, and R.T. Clancy*
University of Colorado, Boulder
rusch@sertan.colorado.edu
Comparisons of Total Ozone from SBUV-2 and SSBUV Surface-Based Direct Sun and Zenith Sky Observations with Those from Dobson 83 Direct Sun Observations
D. F. Heath and Z. Wei
RSI, Boulder, CO
dheath@csn.org
Umkehr Ozone Profile Retrieval Method Uncertainties
J. DeLuisi and I. P. Petropavlovskikh
NOAA/ARL/SRRB, Boulder, CO
deluisi@srrb.noaa.gov
* the procedures for determining stratospheric and tropospheric aerosol errors to the Umkehr retrieved ozone profiles,
* the accuracy of the forward calculation radiative transfer computer codes used in the new and old retrieval algorithms and the calculation of aerosol error,
* first-guess ozone profile effects on retrieved ozone profiles,
* the nature of the vertical profile of the aerosol error to the ozone profile as affected by aerosols at different altitudes,
* characterization of the biases between the SBUV, the New Umkehr, and the Old Umkehr,
* the capability of the New Umkehr and the Old Umkehr to retrieve a subtle changes in the ozone profile such as caused by the solar cycle,
* the ozone profile error caused by tropospheric aerosols,
* the collection and application of stratospheric aerosol information, including lidar data, SAGE data, sunphotometer data, dustsonde data, and in situ aircraft data taken by NASA Ames
* the uncertainties in Umkehr ozone profile trend analyses caused by non-varying first- guess profiles in the Umkehr retrieval algorithm,
* the set-up and testing of a summertime mid-latitude case for calculating the stratospheric aerosol error for the Brewer Umkehr measurement (calculations were made for five wavelengths and were done in collaboration with the Canadian Atmospheric Environment Service), and
* in anticipation of the new US EPA Brewer network consisting of up to 20 stations, we are consulting with Canadians on implementation of the Brewer Umkehr in the EPA network, with the intent to standardize the entire north American ensemble of Brewers making Umkehr measurements.
Ozone and UV Trend Analysis
E. C. Weatherhead1, J. E. Frederick2, G. C. Tiao2, G. C. Reinsel3
1CIRES, University of Colorado, Boulder
2Department of Geophysical Science
University of Chicago
3Department of Statustics
University of Wisconsin, Madison
betsy@srrb.noaa.gov
Analysis of TOMS (Version 7) Total Ozone Data: (a) Daily Variability, (b) Correlation of Ozone Variability with Measured Surface UVB Radiation, and (c) Initial Application of a CTM
R. A. Reck, D. Allen, L.e Carlstrom, K. Schmidt, and E. C. Weatherhead*
Argonne National Laboratory
Ruth_Reck@qmgate.anl.gov
Our results show an average daily variation of total ozone (denoted DW) from 15-30 Dobson Units (5-10%) in the midlatitudes and much smaller variations in the tropics (see Fig. 1a). Zonal asymmetries are evident near 50° in the Northern Hemisphere, with the RMS maxima slightly to the east of the 14.5-year average of total ozone maxima near Hudson Bay and over the east coast of Asia, while the maximum near 50° S, 60° E is nearly colocated with the Southern Hemisphere midlatitude maximum. At two nearby locations in Washington, D.C. at which surface ultraviolet radiation has been measured, fractional daily changes in surface ultraviolet B radiation (not sorted by zenith angle, i. e. time of day) are 5 times larger than fractional changes in the TOMS ozone data. The seasonal cycle of DW (Fig. 1b) shows maxima from the late fall to early spring, with much smaller values in the summer.
The temporal scale of the short-term ozone variability is believed to be related to wave activity in the troposphere and stratosphere. We have determined the contribution to daily total ozone variability from planetary- (waves 1-3) and medium-scale (waves 4-7) waves (see Fig. 1c,d). To fully understand the forcing (chemistry and transport) of short-term ozone variability requires a detailed 3-dimensional chemical transport model. This capability is now being pursued.
Total Ozone Trends at Northern MidLatitudes: Interpretation and Model Comparisons
L. L. Hood and J. P. McCormack
University of Arizona
lon@lpl.arizona.edu
Ultraviolet Radiation Climatology of the Earth's Surface and Lower Atmosphere
S. Madronich and J. Zeng
National Center for Atmospheric Research, Boulder, Colorado
sasha@acd.ucar.edu
University of Alaska, Fairbanks
We are developing a modeling system for estimating atmospheric photodissociation rate coefficients (J values) and biologically effective radiation (dose rates) at any geographical location, altitude, season, and time of the day. A prototype version, the Tropospheric Ultraviolet and Visible (TUV) radiation model, is now available via anonymous ftp at 128.117.32.22, and offers a number of flexible features such as variable altitude and wavelength gridding, easy loading of spectral data, interchangeable radiation schemes, and direct user-control of atmospheric input data. The model includes a novel fast pseudo-spherical method to improve the accuracy of calculations at low sun, and is being used for a number of theoretical studies, including the wavelength dependence of cloud transmission, and the differential absorption by stratospheric vs. tropospheric ozone.
Currently the model computes J values for 47 different photo-reactions of interest to tropospheric and stratospheric chemistry, as well as surface dose rates of common interest (e.g., DNA damage, erythemal irradiance). Critical evaluations of input molecular absorption cross sections and quantum yield data have been carried out for a number of molecules. In one case, for the reaction O3 + hv -> O2 + O(1D), comparisons between the TUV model results and direct actinometric measurements have led to revision of the quantum yield data for this reaction, with substantial implications for the global distribution of the hydroxyl radical production rates. Comparisons with spectro-radiometric measurements obtained at Lauder, New Zealand also show agreement (for cloud-free skies) of about 5-10% for erythemal radiation.
Work underway includes development of a highly optimized version for on-line use in global chemistry/transport models, and assimilation of satellite measurements of ozone, clouds, and aerosols to help establish a global UV climatology and allow model evaluation against the growing data base from surface UV monitoring networks.
The Effects of Heterogeneous Reactions on Ozone in the Lower Stratosphere and Upper Troposphere: A Three dimensional Model (STARS) Study
X.X. Tie and G. Brasseur
National Center for Atmospheric Research, Boulder, CO
xxtie@ra.cgd.ucar.edu
Heterogeneous conversion of bromine reservoirs (BrONO2, HOBr) on the surface of aerosol particles in the lower stratosphere has important effects on the concentrations of Cl) and ozone at mid-latitudes of the lower stratosphere. The model simulations suggest that, owing to the bromine heterogeneous interactions with sulfate aerosols, the increase in total bromine loading from 1960 (12 pptv) to 1990 (21 pptv) has led to enhanced concentration of ClO and consequently to an ozone depletion of 2-3% in the lower stratosphere at middle and high latitudes under post-volcanic conditions.
The model is also used to study the effects of a large volcanic eruption on the ozone mass exchange between the stratosphere and the troposphere. The results suggests that the resulting heterogeneous chemical reactions occurring on the surface of sulfate aerosols have a significant impact on the ozone mass fluxes across the tropopause. The annually global ozone mass transported from the stratosphere into the troposphere could be reduced by approximately 15% after a large volcanic eruption such as that of Mt. Pinatubo.
Return to ACP Homepage
