Richard W. Stewart

Effect of chemical kinetics uncertainties on calculated constituents in a tropospheric photochemical model

The imprecision of photochemical reaction rates as measured in the laboratory introduces significant uncertainty into trace species concentration calculated in a photochemical model. We have evaluated uncertainties in tropospheric concentrations calculated with a one-dimensional photochemical model, using a Monte Carlo technique to introduce random uncertainty into the model rate coefficients. Correlations between rate coefficients and species and between species and species have been determined to answer the following questions. Which are the most critical kinetic processes in determining constituent distributions? Which are the strongest species-species correlations? The most critical reactions turn out to be the primary photodissociations of O3 and NO2, which initiate ozone destruction and formation, respectively. The reaction between OH and methane is critical, as is the rate of nitric acid formation, which removes both odd nitrogen (NOx) and odd hydrogen (HOx). Species-species correlations reveal anticorrelation between HOx and NOx and positive correlation between OH arid peroxides, acids, and aldehydes. Particular attention is given to ozone and to the transient OH, which is difficult to measure and is therefore frequently calculated using photochemical models and observations of more stable trace gases. A set of Monte Carlo computations is performed for conditions simulating several distinct chemical environments because imprecision in computed species are nonlinear and depend strongly on mean chemical composition. For low NOx, low hydrocarbon, low O3 levels, as in the remote troposphere, the 1 σ imprecision in computed boundary layer OH may be as low as 20%, with that for HO2 at 15% and H2O2 at 25%. At higher NOx and O3 levels, the 1σ imprecision in boundary layer OH and HO2 is 70%, and H2O2 is 90%. The 1σ imprecision in computed O3 is ∼15% (7–10 ppbv) in both cases. The implications of model computed OH imprecision for predictive and diagnostic calculations are explored. Averaging over the regionally differing results suggests that a typical estimate of global OH is ∼25% uncertain due to kinetics imprecision. This limits the certainty with which lifetimes for numerous natural and anthropogenic trace gases can be calculated with a photochemical model. The imprecision in a given determination of computed OH can be cut to 20% or less with simultaneous high-precision measurements of O3, CO, CH4, and NO2. Several experimental strategies for optimizing the deduction of OH are described, including one based on measurement of the HO2 radical.

Sensitivity of tropospheric oxidants to global chemical and climate change

A photochemical model has been used to quantify the sensitivity of the tropospheric oxidants O3 and OH to changes in CH4, CO and NO emissions and to perturbations in climate and stratospheric chemistry. Coefficients of the form ∂1n[O3]/∂1n[X] and ∂1n[OH]/∂1n[X], where [X] = flux of CH4, CO, NO; stratospheric O3 and H2O have been calculated for a number of “chemically coherent” regions (e.g. nonpolluted continental, nonpolluted marine, urban) at low and middle latitudes. Sensitivities in O3 and OH vary with regional emissions patterns and are nonlinear within a given region as [X] changes. In most cases increasing CH4 and CO emissions will suppress OH (negative coefficients) and increase O3 (positive coefficients) except in areas where NO and O3 influenced by pollution are sufficient to increase OH. Stratospheric O3 depletion will tend to decrease O3 (except in high NOx areas) and increase OH through enhanced u.v. photolysis. Increased levels of water vapor (one possible outcome of a global warming) will also decrease O3 and increase OH. We conclude that in most regions, NO, CO and CH4 emission increases will suppress OH and increase O3, but these trends may be opposed by stratospheric O3 depletion and climate change. A regional survey of OH and O3 levels suggests that the tropics have a pivotal role in determining the earths future oxidizing capacity.

Impact of temperature‐driven cycling of hydrogen peroxide (H2O2) between air and snow on the planetary boundary layer

Hydrogen peroxide (H2O2) contributes to the atmospheres oxidizing capacity, which determinesthe lifetime of atmospheric trace species. Measured bidirectional summertime H2O2 fluxes fromthe snowpack at Summit, Greenland, in June 1996 reveal a daytime H2O2 release from thesurface snow reservoir and a partial redeposition at night. The observations also provide the firstdirect evidence of a strong net summertime H2O2 release from the snowpack, enhancing averageboundary layer H2O2 concentrations approximately sevenfold and the OH and HO2concentrations by 70% and 50%, respectively, relative to that estimated from photochemicalmodeling in the absence of the snowpack source. The total H2O2 release over a 12-day periodwas of the order of 5 * 10(13) molecules m(-2) s(-1) and compares well with observed concentrationchanges in the top snow layer. Photochemical and air-snow interaction modeling indicate thatthe net snowpack release is driven by temperature-induced uptake and release of H2O2 asdeposited snow, which is supersaturated with respect to ice-air partitioning, approachesequilibrium. The results show that the physical cycling of H2O2 and possibly other volatilespecies is a key to understanding snowpacks as complex physical-photochemical reactors and hasfar reaching implications for the interpretation of ice core records as well as for thephotochemistry in polar regions and in the vicinity of snowpacks in general.

Physically based modeling of atmosphere‐to‐snow‐to‐firn transfer of H2O2 at South Pole

Quantitative interpretation of ice core chemical records requires a detailed understanding of the transfer processes that relate atmospheric concentrations to those in the snow, firn, and ice. A unique, 2 year set of year-round surface snow samples at South Pole and snow pits, with associated accumulation histories, were used to test a physically based model for atmosphere-to-firn transfer of H2O2. The model, which extends our previous transfer modeling at South Pole into the snowpack, is based on the advection-dispersion equation and spherical diffusion within representative snow grains. Required physical characteristics of the snowpack, such as snow temperature and ventilation, were estimated independently using established physical models. The surface snow samples and related model simulations show that there is a repeatable annual cycle in H2O2 in the surface snow at South Pole. It peaks in early spring, and surface snow concentration is primarily determined by atmospheric concentration and temperature, with some dependence on grain size. The snow pits and associated model simulations point out the importance of accumulation timing and annual accumulation rate in understanding the deposition and preservation of H2O2 and δ18O at South Pole. Long-term snowpack simulations suggest that the firn continues to lose H2O2 to the atmosphere for at least 10–12 years (∼3 m) after burial at current South Pole temperatures and accumulation rates.

Multiple steady states in atmospheric chemistry

The equations describing the distributions and concentrations of trace species are nonlinear and may thus possess more than one solution. Several authors have suggested that the steady-state equations describing tropospheric and stratospheric chemistry may have multiple solutions, but the existence of such solutions has not been completely demonstrated. This paper develops methods for searching for multiple physical solutions to chemical continuity equations and applies these to subsets of equations describing tropospheric chemistry. The calculations are carried out with a box model and use two basic strategies. The first strategy is a “search” method. This involves fixing model parameters at specified values, choosing a wide range of initial guesses at a solution, and using a Newton-Raphson technique to determine if different initial points converge to different solutions. The second strategy involves a set of techniques known as homotopy methods. These do not require an initial guess, are globally convergent, and are guaranteed, in principle, to find all solutions of the continuity equations. The first method is efficient but essentially “hit or miss” in the sense that it cannot guarantee that all solutions which may exist will be found. The second method is computationally burdensome but can, in principle, determine all the solutions of a photochemical system. Multiple solutions have been found for models that contain a basic complement of photochemical reactions involving Ox, HOx, NOx, and CH4. In the present calculations, transitions occur between stable branches of a multiple solution set as a control parameter is varied. These transitions are manifestations of hysteresis phenomena in the photochemical system and may be triggered by increasing the NO flux or decreasing the CH4 flux from current mean tropospheric levels.

Physically based inversion of surface snow concentrations of H2O2 to atmospheric concentrations at South Pole

Inversion of chemical records archived in ice cores to atmospheric concentrations requires a detailed understanding of atmosphere-to-snow-to-ice transfer processes. A unique year-round series of surface snow samples, collected from November, 1994 through January, 1996 at South Pole and analyzed for H2O2, were used to test a physically based model for the atmosphere-to-snow component of the overall transfer function. A comparison of photochemical model estimates of atmospheric H2O2, which are in general agreement with the first measurements of atmospheric H2O2 at South Pole, with the inverted atmospheric record (1) demonstrate that the surface snow acts as an excellent archive of atmospheric H2O2 and (2) suggest that snow temperature is the dominant factor determining atmosphere-to-surface snow transfer at South Pole. The estimated annual cycle in atmospheric H2O2 concentration is approximately symmetric about the summer solstice, with a peak value of ∼280 pptv and a minimum around the winter solstice of ∼1 pptv, although some asymmetry results from the springtime stratospheric ozone hole over Antarctica.

Temperature distributions in the lower atmosphere of Mars from Mariner 6 and 7 radio occultation data

Mars lower atmosphere vertical temperature distributions from Mariner 6 and 7 radio occultation data, using improved trajectory estimates

Kinetic data imprecisions in photochemical rate calculations: Means, medians, and temperature dependence

The uncertainty in computed species concentrations resulting from measurement imprecision in reaction rate components is investigated in two tropospheric photochemical models. In this study, which extends Thompson and Stewart [1991], we perform statistical analysis on reaction rate coefficients to focus on two aspects of model uncertainty : (1) the change in the magnitude of concentration uncertainty as temperature varies throughout a model grid and (2) the difference resulting from selection of mean, as opposed to median, rates for model calculations. Reaction rates are treated as realizations of random variables having statistical properties given by component terms and their imprecisions. These assumptions lead to expressions for probability distributions of bimolecular rates and for the high- and low-pressure limits for termolecular and thermolytic processes. They also imply that bimolecular rates and high- and low-pressure limits used in photochemical models correspond to median values taken from a lognormal distribution. We derive analytic expressions for mean values, which are always larger than the medians, an intrinsic property of lognormal variables. We suggest that comparison of species concentrations computed using median and mean rates can provide some measure of the effect of rate imprecision in multidimensional models. We also find that the temperature dependence derived for the imprecision in bimolecular rates differs from that given in standard references, with our expression giving a smaller increase in imprecision as temperatures deviate from 298 K. Results are illustrated using a box and a one-dimensional (1-D) model.

Perturbations to tropospheric oxidants, 1985–2035: 2. Calculations of hydrogen peroxide in chemically coherent regions

Abstract Increasing global emissions of trace gases NO, CH4, and CO, along with perturbations initiated by changes in stratospheric O3 and H2O, may cause tropospheric hydrogen peroxide (H2O2) levels to change. Specific scenarios of CH4CONO emissions and global climate changes are used to predict HO2 and H2O2 changes from 1985 to 2035 in a one-dimensional model that simulates different chemically coherent regions (e.g. urban, non-urban continental and marine mid-latitudes; marine and continental low latitudes). If CH4 and CO emissions continue to increase throughout the troposphere at current rates (1% yr−), there will be large increases in H2O2, for example, more than 100% in the urban boundary layer from 1985 to 2035. Globally, H2O2 will increase 22% with HO2 increasing 8% and O3 increasing 13%. When CH4, CO and NO emissions are specified on a regionally varying basis and are parameterized for high and low potential growth rates, globally averaged increases in surface concentrations are 12% for H2O2 and 18% for O3. A global warming (with increased H2O vapor) or stratospheric O3 depletion superimposed on CH4, CO and NO emissions changes will cut O3 increases but add to peroxide, increasing levels as much as 150% above present day in some regions. Both globally uniform and region-specific scenarios predict a 10–15% loss in global OH from 1985 to 2035. Thus, conversion of OH to HO2 and H2O2 in the atmosphere may signify a loss of gaseous oxidizing capacity in the atmosphere and an increase in aqueous-phase oxidizing capacity.

The Natural and Perturbed Troposphere

It is now generally accepted that the troposphere is a region of great chemical complexity and that many human activities may alter the chemical structure of the region. A prerequisite to any realistic assessment of human impacts on pollution and climate is an understanding of the natural budgets of atmospheric gases. This requires a detailed knowledge of physical, chemical, and biological processes within the various reservoirs which are involved in the cycles of these gases. This paper first reviews the processes important in establishing the concentrations of a number of tropospheric species and discusses gaps in our current understanding of these processes. We identify the points at which man may intervene in the major cycles of atmospheric gases and describe the possible consequences of such interventions. Pollutants released into the troposphere may adversely affect the environment by virtue of their chemical interactions with other atmospheric species, their radiative properties, or both. Problems discussed in this review include the growth of atmospheric CO2 resulting from the burning of fossil fuels and its possible climatic effects, the consequences of increased levels of CO emission on the self-cleansing ability of the troposphere and on the radiation budget, and possible changes in the stratospheric odd nitrogen and ozone amounts due to increased use of fertilizers in agriculture. The magnitude of the perturbations predicted by various model studies are reviewed with particular attention to uncertainties which may affect the results.