Some recent studies carried out in our laboratory are described where laser flash photolytic production of reactant free radicals has been combined with reactant and/or product detection using time-resolved optical techniques to investigate the kinetics and mechanisms of important atmospheric chemical reactions. Discussed are (1) a study of the radical-radical reaction O + BrO yields Br + O₂ where two photolysis lasers are employed to prepare the reaction mixture and where the reactants O and BrO are monitored simultaneously using atomic resonance fluorescence to detect O and multipass UV absorption to detect BrO; (2) a study of the reaction of atomic chlorine with dimethylsulfide (CH₃SCH₃) where atomic resonance fluorescence detection of Cl is employed to elucidate the kinetics and tunable diode laser absorption spectroscopy is employed to investigate the HCl product yield; and (3) a study of the aqueous phase chemistry of Cl₂^- radicals where longpath UV absorption spectroscopy is employed to investigate the kinetics of the Cl₂^- + H₂O reaction.

A new, compact Fourier Transform Michelson interferometer (FTUV) with an apodized resolving power greater than 300,000 at 300 nm, high throughput and wide spectral coverage has been developed. The objectives include atmospheric spectroscopy (direct solar absorption and solar scattering) and laboratory spectroscopy of transient species. In this paper, we will briefly describe the prototype FTUV instrument and the results of preliminary laboratory investigations of OH and ClO spectra in emission and absorption.

IR absorption using tunable diode laser spectroscopy provides a sensitive and quantitative detection method for laboratory kinetic studies of atmospheric trace gases. Improvements in multipass cell design, real time signal processing, and computer controlled data acquisition and analysis have extended the applicability of the technique. We have developed several optical systems using off-axis resonator mirror designs which maximize path length while minimizing both the sample volume and the interference fringes inherent in conventional 'White' cells. Computerized signal processing using rapid scan (300 kHz), sweep integration with 100 percent duty cycle allows substantial noise reduction while retaining the advantages of using direct absorption for absolute absorbance measurements and simultaneous detection of multiple species. Peak heights and areas are determined by curve fitting using nonlinear least square methods. We have applied these techniques to measurements of: (1) heterogeneous uptake chemistry of atmospheric trace gases (HCl, H2O2, and N2O5) on aqueous and sulfuric acid droplets; (2) vapor pressure measurements of nitric acid and water over prototypical stratospheric aerosol (nitric acid trihydrate) surfaces; and (3) discharge flow tube kinetic studies of the HO2 radical using isotopic labeling for product channel and mechanistic analysis. Results from each of these areas demonstrate the versatility of TDL absorption spectroscopy for atmospheric chemistry applications.

The UV absorption spectrum and the self-reaction rate constants of the ethylperoxy (C₂H₅O₂) radical and of two brominated peroxy substitutes, 2-bromo-ethylperoxy (BrCH₂CH₂O₂) and 3-bromo-sec-butylperoxy [(CH₃)BrCHCH(CH₃)O₂], were measured via the molecular modulation technique in conjunction with both single wavelength and diode-array UV spectroscopy. The UV spectrum of the three peroxy radicals shows broad absorption with a (sigma) _max near 4.3 X 10^-18 cm² molecule^-1 at about 240 nm. Analysis of decay profiles of C₂H₅O₂ at 250 nm enabled an overall removal rate, k_obs, of (1.1 +/- 0.1) X 10^-13 cm³ molecule^-1 s^-1 at 298 K to be determined. Between 250 and 330 K, k_obs may be represented by: k_obs (250 - 330 K) equals (1.85 +/- 0.2) X 10^-13 exp -[(147 +/- 30)/T] cm^{3/ molecule(superscript -1} s^-1. Self-reaction rate constants, k_obs, of (6.2 +/- 0.1) X 10^-12 and (9.6 +/- 1.9) X 10^-13 cm³ molecule^-1 s^-1 were obtained at 298 K for (BrCH₂CH₂O₂) and [(CH₃)BrCHCH(CH₃)O₂], respectively.

Fourier transform infrared (FTIR) spectroscopy has been used to study nitric-acid/ice films representative of type I polar stratospheric clouds (PSCs). These studies reveal that in addition to amorphous nitric acid/ice mixtures, there are three stable stoichiometric hydrates of nitric acid: nitric-acid monohydrate (NAM), dihydrate (NAD) and trihydrate (NAT). We also observe two distinct crystalline forms of the trihydrate, which we denote (alpha) - and (beta) -NAT. These two forms appear to differ in their concentration of crystalline defects, but not in their chemical composition. In addition to probing the composition of type I PSCs, we have also used FTIR spectroscopy to study the interaction of HCl with model PSC films. In this work we find that for HCl pressures in the range 10^-5 - 10^-7 Torr, HCl is taken up by ice at 155 K to form a thin layer of HCl (DOT) 6H₂O. At 193 K, the uptake of HCl by ice was consistent with <EQ monolayer coverage. Uptake of HCl by (alpha) - and (beta) -NAT at 175 K was also consistent with <EQ monolayer coverage.

The tunable diode laser absorption spectrometric technique was employed to measure H₂O₂ and HCHO during the Mauna Loa Observatory Photochemistry Experiment II fall and winter intensives. Comparisons were made between the measurements made with the TDLAS technique and several other methods. The half-hourly average mixing ratios varies between 1300 pptv and the instrument detection limit of 50 - 100 pptv for H₂O₂ and between 400 pptv and the instrument detection limit approximately 25 pptv for HCHO. Daily average mixing ratios for H₂O₂ were 550 +/- 79 pptv (fall) and 540 +/- 80 pptv (winter), and for HCHO were 130 +/- 21 pptv (fall) and 200 +/- 35 (winter). HCHO showed strong diurnal behavior with maxima in the 12:00 - 17:00 time period whereas H₂O₂ showed essentially no diurnal variation. Clean tropospheric air sampled during downslope conditions between 20:00 and 06:00 had HCHO mixing ratios averaging 80 pptv in the fall and 160 pptv in the winter. These values are considerably lower than model predictions.

We recently have conducted measurements of distributions and fluxes of methane in several small towns in the U.S. These measurements are made with a real-time methane instrument recently developed at Aerodyne Research. This instrument is based on absorption of infrared light from a Zeeman-broadened HeNe laser operating at 3.39 micrometers . A non-uniform magnetic field applied to the laser plasma results in a nearly flat gain profile, which allows piezoelectric tuning of the laser on and off a methane line. Long path (47 m) absorption occurs in a multipass cell. The instrument has a noise level of 5 ppbv (rms, 1 sec.) and a gas sampling time of 6 sec. The instrument is mounted in a van with a suite of other real-time instruments (CO₂ and SF₆) which can function while moving at roadway speeds. We map the methane distribution in towns, identify the primary methane sources, and measure fluxes with tracer methods. The simultaneous measurement of CO₂ with a commercially available NDIR device is useful in identifying some sources of methane. An SF₆ instrument (electron capture) allows precise location and measurement of tracer plumes during flux measurement experiments.

Air pollution monitoring requires sensitive and selective detection schemes. This paper reviews the main theoretical and experimental characteristics of gas phase laser photoacoustic spectroscopy. A mobile CO₂ laser photoacoustic system is described in more detail. Examples demonstrate the great potential of this technique. Various compounds could successfully be detected in urban and rural environments with excellent time resolution. The sensitivity permits the detection of trace gases at ppb concentrations and the selectivity in some cases even permits the differentiation between isomers. Present studies include measurements with isotopic CO₂ lasers as well as the implementation of a continuously tunable high pressure CO₂ laser.

Gaseous or aqueous ammonia and halogen species were determined by chemiluminescence (CL) method based on the new CL reaction of ammonia and halogen species using the gas-liquid contact type reaction vessel. CL produced in the reaction of ammonia and chlorine has the maximum intensity at 690 nm and that with bromine has the maximum intensity at 850 nm. Detection limit of ammonia was 7 ppb (V/V) and relative standard deviation for 5 measurements at 4 ppm (V/V) was 2.5%. This method could be applied for the determination of ammonium cation or hypochlorite anion in liquid phase. Detection limit of ammonium cation was 20 ppb (W/W) and standard deviation for 5 measurements at 5 ppm (W/W) was 1.2%. Gaseous chlorine and bromine could be continuously determined using this CL reaction. Ammonium in acid rain was measured by the developed method.

The chemistry of the atmosphere is substantially influenced by a wide range of chemical processes which are primarily driven by the action of ultraviolet radiation of wavelengths shorter than 320 nm (UV-B) on ozone and water vapor. This leads to the formation of hydroxyl (OH) radicals which, despite very low tropospheric concentrations, remove most gases that are emitted into the atmosphere by natural and anthropogenic processes. Therefore, although only about 10% of all atmospheric ozone is located in the troposphere, through the formation of OH, it determines the oxidation efficiency of the atmosphere and is, therefore, of the utmost importance for maintaining its chemical composition. Due to a variety of human activities, especially through increasing emissions of CH₄, CO, and NO_x, the concentrations of tropospheric ozone and hydroxyl are expected to be increasing in polluted and decreasing in clean tropospheric environments. Altogether, this may be leading to an overall decrease in the oxidation efficiency of the atmosphere, contributing to a gradual buildup of several longlived trace gases that are primarily removed by reaction with OH. In the stratosphere, especially due to catalytic reactions of chlorine-containing gases of industrial origin, ozone is being depleted, most drastically noted during the early spring months over Antarctica. Because ozone is the only atmospheric constituent that can significantly absorb solar radiation in the wavelength region 240 - 320 nm, this loss of ozone enhances the penetration of biologically harmful UV-B radiation to the earth's surface with ensuing negative consequences for the biosphere. Several of the aforementioned chemically active trace gases with growing trends in the atmosphere are also efficient greenhouse gases. Together they can exert a warming effect on the earth's climate about equal to that of carbon dioxide.

Ground-based measurements of scattered light were collected in the Antarctic fall, winter, and spring at McMurdo Station during the winter and spring of 1991. These measurements yielded values of the slant column amounts of the dioxides of chlorine and nitrogen found in the stratosphere. Two different viewing schemes were used to collect this data; a zenith viewing mode to collect light scattered from directly overhead, and a new off-axis viewing mode to collect light scattered along a path at 80 degree(s) zenith angle toward the sun. This new viewing geometry allowed measurements to be made much further into twilight and polar winter than was previously possible with only the zenith viewing mode. Results of this analysis showed that OClO levels were very high in late July and early August, declining through September, and dropping below detection levels in early October.

A method is described which makes possible the recording of absolute atmospheric absorption spectra. The method is based on optical long path absorption spectroscopy and thus integrates the optical properties over a measurement path, typically one kilometer. Cancellation of instrument factors are achieved by ratioing spectra from two nearly collinear paths of different length recorded close in time. The method is demonstrated for measurement of spectrally resolved particle extinction in the wavelength region 275 - 300 nm, a physical parameter needed for the correction of LIDAR ozone measurements. By the use of differential optical absorption spectroscopy, (DOAS) it is possible to obtain concentrations of O₃ and SO₂ and by a subtraction technique their contribution to the absorption are eliminated. The method was tested during an ozone LIDAR intercomparison campaign, TROLIX, in the Netherlands, 1991. In this field experiment spectrally resolved particle extinction was measured as well as O₃ and SO₂ concentrations, and the results were compared with LIDAR and point monitor measurements. The dual path method also has the potential to improve the detection limit in long path absorption measurements of gas concentration. By ratioing spectra from two different pathlengths instrument factors are cancelled and detection limits may be improved. An example demonstrating improvement in ozone long path measurements is given.

The EPA is preparing to issue new regulations calling for enhanced monitoring of ozone and oxides of nitrogen and additional monitoring of volatile organic compounds (VOC), including aldehydes, as well as meteorological parameters. Ozone non-attainment areas classed as serious or worse would be required to establish photochemical assessment monitoring stations (PAMS). Long path, remote sensing methods are not deemed by the EPA to be sufficiently accurate and comprehensive in providing fully speciated VOC measurements to be included in the regulations. This paper reports on the performance of a representative long path system based on ultra-violet absorption spectroscopy, and assesses the potential benefits and limitations for the future use of such systems in the PAMS network and in related air quality management applications. The basis for our assessment is field experience with a commercially available U-V spectroscopic long path instrument, using differential optical absorption spectroscopy (DOAS), as well as reference to the published experience of others using similar long path methods for measuring air pollutants. The database for our analysis consists of air quality measurements, obtained from the DOAS instrument and an existing network of point monitors (as archived in the AIRS database of EPA), in the Washington DC region during the summer of 1991. The goal of this continuing study is to suggest ways in which the long path spectroscopic methods might be used in a cost-effective manner to complement the more detailed batch sampling methods in support of urban air quality management.

Concentrations of SO₂, NO₂, H₂CO, and O₃ have been measured regularly since October 1990 at the urban site of the Campus of the Universite Libre de Bruxelles, using the differential optical absorption spectroscopy (DOAS) technique associated with a Fourier Transform Spectrometer. The experimental set up has already been described elsewhere (Vandaele et al., 1992). It consists of a source (either a high pressure xenon lamp or a tungsten filament) and an 800 m long path system. The spectra are recorded in the 26,000 - 38,000 cm^-1 and 14,000 - 30,000 cm^-1 spectral regions, at the dispersion of 7.7 cm^-1. The analytical method of the DOAS technique is based on the fact that in atmospheric measurements, it is impossible to obtain an experimental blank spectrum. Therefore, the Beer-Lambert law has to be rewritten as: I equals I'_oe^n(Delta (sigma) d) where I is the measured intensity, I_o the measured intensity from which all absorption structures have been removed, n the concentration, d the optical path length, and (Delta) (sigma) the differential absorption cross section of the molecule. Numerous methods for determining I'_o exist. Fourier transform filtering has been used in this work. This method defines I'_o as the inverse Fourier transform of the lower frequencies portion of the power spectrum of the experimental data. A least squares procedure is then applied in order to determine the concentration of the desired molecules.

DOAS measurements of NO₃, NO₂, and O₃ were performed on the French Atlantic coast in June 1989. The use of two light paths of 6 km length covering the altitude intervals of 5 m - 10 m and 5 m - 40 m allowed the observation of vertical NO₃ concentration gradients. Since the NO₃ precursors (NO₂ and O₃) did not show significant differences on the two light paths altitude dependent sinks must be considered. Simultaneously measured DMS concentrations at 1 m and 40 m altitude showed a vertical gradient with higher values at ground level during all nights. For 3 (of 10) nights measured DMS concentrations and gradients could explain the observed NO₃ behavior.

During the summer of 1991, simultaneous measurements of O₃, NO, NO₂, and SO₂ were obtained by an open path differential optical absorption spectrometer (DOAS) and conventional point methods at two sites (one urban and one rural) in the central Piedmont of North Carolina. The monitoring was conducted as part of the Southern Oxidant Study (SOS) program. In addition, the DOAS was configured to measure HNO₂, NH₃ HCHO, benzene, toluene, and p-xylene. The DOAS measurements were each integrated over a one minute period and the full set of measurements was obtained over a seven minute cycle time. NO and NH₃ were measured over a 200 meter folded path. The remaining gases were measured over a 400 meter folded path. Additional point measurements included continuous measurements of NO_y, H₂O₂, and integrated measurements of speciated hydrocarbons, aldehydes, and HNO₃. This paper examines the relationship between the path integrated, and point measurements of O₃, NO, NO₂, and SO₂.

Two systems developed for measuring gases in the boundary layer are described. A portable long-path absorption instrument which uses an incandescent infrared source with an oscillating filter and a mobile differential-absorption lidar system that performs range-resolved measurements in the ultraviolet, visible, and infrared spectral regions. Measurements of industrial and urban pollution performed with both of these systems are described.

A fully computer-controlled differential optical absorption spectroscopy (DOAS) system for atmospheric air pollution monitoring is described. A receiving optical telescope can, via a large planar mirror controlled by stepping motors, sequentially tune in to light beams from a number of distant lamp light sources to cover the area. The light is directly coupled into a rapid scanning spectrometer from the computer-controlled secondary mirror in the Newtonian Telescope. The beam-finding servo system and automatic gain control allows long unattended measurements. Using an astronomical code, celestial sources can also be searched and tracked. By computer fitting to stored laboratory spectra the path-averaged concentration of a number of important pollutants such as NO₂, SO₂, and O₃ can be evaluated. A measurement of NH₃ and NO close to the UV-limit is also demonstrated. Evaluated data are stored together with meteorological data and other system parameters.

The `multiplex advantage' of diode array detectors over optomechanical spectral scanning devices promises to improve DOAS instruments by reducing measurement time about two orders of magnitude. Alternatively the signal to noise ratio can be improved by adding several scans. Unfortunately, the use of diode arrays gives rise to new problems. Two thin layers on the semiconductor (the protective SiO₂ layer plus deposits of vapor) produce spectral interference structures and a thermal recombination current in the diode junction is superimposed to the light signal. Interferences in the protective layer of the diode array impose spectral structures, which are subjected to slow changes due to deposition of vapor upon the diode array and rapid changes caused by varying illumination of the array due to air turbulence in the light path. Detailed model calculations reproduce the etalon structure measured in the laboratory by assuming the existence of a layer of SiO₂ (index of refraction n approximately equals 1.6) and a second layer with n approximately equals 1.3 (probably ice). A model considering the geometry of the spectrograph detector system is presented, which describes the influence of changing illumination of the diodes (i.e., caused by atmospheric turbulence) on the etalon structure. The dark current can, in principle, be reduced by cooling the diode array. However, low temperatures increase the complexity of the detector and enhance deposition of vapor, aggravating the etalon problem. The dark current, depending on the charge of the diode, is a complex function of light intensity and exposure time. Thus, subtracting the signal of the darkened diode array to remove the dark current signal leads to apparent nonlinearities. A model for the behavior of the dark current is presented, which allows the use of diode arrays without extensive cooling.

Although several methods exist for the determination of the flux of an atmospheric species, the airborne eddy correlation method has the advantage of providing direct flux measurements that are representative of regional spatial domains. The design criteria pertinent to the construction of chemical instrumentation suitable for use in airborne eddy correlation flux measurements are discussed. A brief overview of the advantages and limitations of the current instrumentation used to obtain flux measurements for CO, CH4, O3, CO2, and water vapor are given. The intended height of the measurement within the convective boundary layer is also shown to be an important design criteria. The sensitivity, or resolution, which is required in the measurement of a scalar species to obtain an adequate species flux measurement is discussed. The relationship between the species flux resolution and the more commonly stated instrumental resolution is developed and it is shown that the standard error of the flux estimate is a complicated function of the atmospheric variability and the averaging time that is used. The use of the recently proposed intermittent sampling method to determine the species flux is examined. The application of this technique may provide an opportunity to expand the suite of trace gases for which direct flux measurements are possible.

The hydroxyl radical (OH) is important for many processes involved in tropospheric chemistry. For instance, it initiates the photochemical degradation of gases that cause global climate change, such as methane and the chlorofluorocarbon substitutes (HCFCs). Because of its reactivity, its abundances are less than 0.1 pptv. Thus, OH has been very difficult to measure accurately, despite its importance. Techniques have evolved, however, so that good measurements of tropospheric OH abundances are now possible. One of these techniques that is adaptable to aircraft measurements is the laser induced fluorescence detection of the OH radical in a detection chamber at low pressures. The current ground-based instrument, which can be readily adapted to aircraft, can detect OH abundances of 1.4 x 10 exp 5 OH molecules/cu cm with S/N = 2 in 30 sec, and 5 x 10 exp 4/cu cm in 5 min.

Filter radiometers sensitive from 280 to 320 nm and from 280 to 400 nm, respectively, were used for measurements of the actinic flux in the stratosphere. Since the instruments are calibrated for absolute spectral sensitivity the data can be compared with model calculations of the actinic flux. Data were obtained during seven balloon flights during the European Arctic Stratospheric Ozone Experiment (EASOE).

A new remote sensing instrument, the Submillimeterwave Limb Sounder, has been developed to simultaneously measure ClO, HCl, O3 and HO2 from high altitude balloon and aircraft platforms. The instrument is a double sideband heterodyne radiometer measuring atmospheric thermal emission spectra near 640 GHz from the Earth's limb. The instrument was flown in April and October, 1991 and in February, 1992 on high altitude balloon platforms over the southwestern United States.

A cryogenic high spectral resolution Fourier Transform Spectrometer for limb emission measurements from a balloon platform in the mid-IR-region is described. The payload consists of a line-of-sight stabilization system, a telescope, a double-pendulum-interferometer, a detector system, and fast onboard signal-processing electronics. Sufficient sensitivity is achieved by cooling the optical system to 200 K with solid carbondioxide, by using liquid helium-cooled detectors and by splitting up the spectral coverage in bands. Calibration is performed by using space and blackbody measurements. Four successful balloon flights have been undertaken from Aire sur l'Adour, France and Esrange, Sweden between May 1989 and March 1992. All flights have provided information about a variety of trace gases involved in the ozone depletion process.

The Far Infrared Limb Observing Spectrometer (FILOS) is an instrument designed to measure chemical species in the upper atmosphere using limb emission in the FIR region of the spectrum. FILOS uses three Fabry-Perot etalons in series to obtain a resolution of 0.0017/cm near 100/cm (100 microns). It is compact and has low power and low data rate requirements so that it may be flown as an auxiliary balloon payload with larger instruments. FILOS has two 0.05/cm bandwidth channels which are currently tuned to a HCl line at 104.2/cm and a pair of OH lines at 101.3/cm. The instrument is described in further detail and results are presented from two recent balloon flights in which OH was measured as a function of time on one hour centers from sunrise to sunset.

The JPL MkIV interferometer, a Fourier transform IR (FTIR) spectrometer designed specifically for atmospheric remote sensing, made measurements of the composition of the Arctic stratosphere in January, February, and March 1992. These measurements were made from the NASA DC-8 aircraft as part of the AASE2 campaign. The data reveal that despite 5 to 6 km of subsidence inside the vortex, which more than doubled the vertically integrated column amounts (burdens) of HF and HNO₃ with respect to outside the vortex, considerable losses of NO₂, HCl, and ClNO₃ were evident by mid-January. Temporary freeze-out of HNO₃ was observed only on one occasion, Jan. 19, and was accompanied by substantial reductions in HCl and ClNO₃. During February and March, ClNO₃ and NO₂ amounts increased dramatically. HCl also recovered but at a much slower rate, so that by March ClNO₃ was the major reservoir of inorganic chlorine, at times exceeding HCl by a factor 2.

Reflecting the increasing importance of atmospheric chemistry associated with the ozone problem and the potential role of observations from space on a global scale, the European Space Agency has included within its plans for future missions the provision of a number of instruments capable of contributing to work in this area. The first of these is the GOME (Global Ozone Monitoring Experiment) approved for flight on ERS-2. Following this a number of chemistry instruments are being considered for flight on the European polar platforms. These include nadir viewing as well as limb sounding instruments, operating in absorption as well as in emission. All parts of the electromagnetic spectrum are addressed including the ultraviolet, visible, infrared and microwave regions.

The Global Ozone Monitoring Experiment (GOME) and the Scanning Imaging Absorption Spectrometer for Atmospheric Chartography (SCIAMACHY), are two European experiments which will fly on satellite platforms in the 1990s. GOME is a small scale version of SCIAMACHY and together they form the SCIAMACHY scientific project. The principal scientific objective is the global determination of the distributions of atmospheric constituents: trace gases, aerosol, and cloud. Special emphasis is placed in this project on stratospheric and tropospheric measurements. GOME observes between 240 and 790 nm in nadir sounding, whereas SCIAMACHY will sound the atmosphere in nadir, limb and solar, and lunar occultation viewing geometries between 240 and 2380 nm.

ATMOS is an investigation of the chemistry and composition of the middle atmosphere using a modified Michelson interferometer designed to be carried on board the Space Shuttle. During orbital sunsets and sunrises it obtains high resolution infrared solar spectra every 2 seconds. The instrument was first flown on the Spacelab 3 mission in April, 1985, and is being reflown as part of the ATLAS series of payloads which started with the ATLAS-1 flight in March 1992. A summary of the results from the Spacelab 3 mission will be presented. These results included several first detections of critical atmospheric species in addition to the thirty or more constituents for which profiles were derived at altitudes between 10 and 150 km. Preliminary results from the ATLAS-1 mission will be described as part of an update to the status of this long term effort.

Measurements of the atmospheric backscattered UV albedo have been used from satellites for more than 20 years to measure ozone. The longest continuous record has been from the Solar Backscattered Ultraviolet instrument (SBUV) and TOMS on the Nimbus 7 satellite, which have been in operation since November of 1978. Because of degradation in space of the diffuser plate used to measure extraterrestrial solar flux, it has been necessary to develop new techniques to maintain the calibration of these instruments. Calibration is maintained by requiring that ozone measured by different wavelength pairs be consistent, and by requiring that ozone measured at different solar zenith angles be consistent. This technique of using a geophysical quantity, ozone, as a transfer standard for wavelength calibration is very powerful. The recalibrated data have been used to measure total ozone trends to an accuracy of +/- 1.3 percent 2(sigma) error over ten years. No significant trends are found near the equator, but significant trends larger than predicted by homogeneous chemistry are found at middle to high latitudes in both hemispheres. In addition, UV albedo data have been used to measure SO2 using band structure in the 300-310 nm range, and to measure nitric oxide in the upper stratosphere and mesosphere using the (10) and (02) NO gamma band fluorescence features.

The improved stratospheric and mesospheric sounder (ISAMS) is one of the instruments on the NASA Upper Atmosphere Research Satellite which was launched in September 1991. ISAMS is a limb-viewing infrared radiometer which measures thermal emission in 24 pass bands (some of which are obtained by gas correlation). This enables the daily mapping over much of the Earth of temperature, the concentrations of 8 chemical species (water vapor, methane, ozone, nitric acid, nitrogen dioxide, nitric oxide, dinitrogen pentoxide, carbon monoxide), and aerosol opacity in the stratosphere and mesosphere. The instrument has eight separate focal planes, each consisting of a 4-element detector array, which are cooled by two mechanical coolers developed specifically for the instrument. The instrument uses a moveable mirror to scan the limb in elevation and to view at a variable azimuth angle to avoid Doppler shifts; the view may be to either side of the spacecraft in order to improve the geographical coverage.

This paper evaluates pulsed photolysis techniques for studying gas-phase radical-molecule reactions that are important in atmospheric chemistry. Reactions of OH are discussed. The pulsed photolysis-pulsed laser induced fluorescence (PP-PLIF) technique is described and compared with other detection techniques. As an example of application of this technique the kinetics of the OH reaction with Methyl Vinyl Ketone (MVK) and Methacrolein (MACR) over the temperature range 232-378 K are presented. Measurements of the UV absorption cross sections of MVK and MACR over the wavelength range 190-450 nm using a diode array spectrometer are also described. These data are used to calculate the tropospheric lifetimes of MVK and MACR.

Relative rate constants for OH reactions with some HFCs have been determined at 298 K by a technique which measures the loss of HFC compared to that of a reference reactant. The experiments were done in a slow-flow or stopped-flow photochemical reactor. The OH was produced by photolysis of H₂O vapor at 185 nm. The HFCs studied were CF₂HCF₃ (HFC-125), CFH₂CF₃ (HFC-134a), and CH₃CF₂H (HFC-152a). Methane and ethane were used as the reference reactants. Concentrations were monitored by FTIR. The following ratios were determined: k(152a)/k(CH₄) equals 5.2 +/- 0.5, k(CH₄)/k(125) equals 3.9 +/- 0.5, k(CH₄)/k(134a) equals 2.1 +/- 0.2, k(134a)/k(125) equals 2.0 +/- 0.2, and k(C₂H₆)/k(152a) equals 6.2 +/- 1.0. These results are in good agreement with literature values for the absolute rate constants, except for HFC 134a, in which case a slower rate constant (factor of about 1.3) is indicated.

A liquid jet technique is used to measure gas phase loss and liquid phase uptake of nitrogen dioxide (NO₂) and nitrous acid (HNO₂) caused by heterogeneous reactions on liquid surfaces. The experimental results are compared with two-dimensional model calculations. In the case of NO₂ the predictions of the model are in agreement with the experimental observations for both gas and liquid phase (i.e., conservation of mass) assuming an effective Henry constant at least eight times larger as the physical Henry constant. In addition the comparison yields a mass accommodation coefficient (alpha) >= 2 * 10^-4 for NO₂ and (alpha) > 1 * 10^-3 for HNO₂.

The kinetics of the reactions of OH with i-butanol and t-butanol have been studied by laser flash photolysis-laser induced fluorescence (LP-LIF), under pseudo-first-order conditions. The values, k(OH + i-butanol) equals (9.0 +/- 0.9) X 10^-12 cm³ molecule^-1 s^-1 and k(OH + t-butanol) equals 1.8 +/- 0.2) X 10^-13 cm³ molecule^-1 s^-1 at 292 K were obtained. As a check on the technique measurements were also made of the OH + CH₄ reaction. The value obtained, k(OH + CH₄) equals (5.4 +/- 0.2) X 10^-15 cm³ molecule^-1 s^-1 at T equals 292 K, is in excellent agreement with the most recent literature value.

Absolute rate constants for the reactions of OH radicals with CCl₃CHO and CF₃CHO have been determined as a function of temperature using a pulsed laser photolysis-resonance fluorescence technique. The rate constant for reaction with CCl₃CHO, k(CCl₃CHO) equals 1.4 +/- 0.2 cm³ molecule^-1 s^-1, was found to be virtually independent of temperature over the range 233 - 313 K while the Arrhenius expression obtained for reaction with CF₃CHO was k(CF₃CHO) equals 3.5 +/- 1.0 X 10^-12 exp [-(488 +/- 57)/T] cm³ molecule^-1 s^-1. Reactivity trends and atmospheric implications are discussed.

Multireflection cells based on the development by White (1942, 1976) are becoming increasingly popular for studying the chemical composition of the ambient atmosphere by long path UV-visible absorption spectroscopy (DOAS). This is partly due to the special property of the design to eliminate negative effects of vibration and atmospheric fluctuations on the optical quality of the cell. The main parameters limiting the number of reflections (and thus achievable optical length) as well as the vibration tolerance are the size of the main mirror and the number of loci that can be fitted on its surface. Also limiting can be the total reflectivity of the mirrors. Here we present a design variant simplifying the optical alignment of the cell by replacing two pairs of about 90-deg mirrors by quartz prisms and adding a third prism to further increase the stability of the system by a factor of two. An application of a practical design based on those principles to the simultaneous measurement of NO2, NO3, and oxygen dimers (O4) is presented.

The episodes when the boundary-layer ozone concentrations in the Arctic drop from the normal 30-40 ppb to less than 3 ppb levels have been found to be associated with high concentrations of 'filterable bromine', and were suggested (Barrie et al., 1988; Bottenheim et al., 1990) to be caused by BrO radicals present in the filterable bromine. This paper describes direct measurements of BrO made in April 1992 during the Polar Sunrise Experiment in the Canadien Arctic, using long-path differential optical absorption spectroscopy (LP-DOAS). A new design of the LP-DOAS spectrometer was employed, based on the principle of Platt and Perner (1983), which uses a holographic flat field grating, a slotted disk as scanning device, and a photomultiplier as a detector. A comparison of ozone concentrations measured with the LP-DOAS with in-situ measurements made by a short-path UV-absorption instrument showed agreement within a few ppb. Measurements for April 1992 indicated BrO concentrations between about 3 ppb to 17 ppb. However, no anticorrelation was found between concentrations of ozone and BrO.

An absorption method for the monitoring of tropospheric trace gases is presented which is based on the fast scanning of a laser system. The 1 km absorption path is folded by a 6m White cell with an open path set up. In situ measurements of tropospheric OH have been carried out under different environmental conditions. Due to the small dimensions of the volume under study a controlled variation of the photon flux density can be realized. This 'open air laboratory' allows kinetic measurements of tropospheric photochemistry under field conditions.

We describe the UV laser long path absorption technique that was used to measure the concentration of free hydroxyl radicals in the troposphere. The paper discusses the experimental setup and the detection limit that was obtained in field experiments and presents an example of the deconvolution process to evaluate the OH concentration from the spectra. We found that under the conditions of field experiments the observed SNR of the long path air absorption spectra was considerably smaller than values obtained with a short path in the laboratory and predicted from statistical calculations of the detector performance. We attribute this to absorption contributions of yet unknown atmospheric trace species. Present detection limits depend on the absortion path length and range from (0.5-4) x 10 exp 6 OH/cu cm.

The results of limb radiance calculations in the 9.6 micrometers ozone absorption band are presented for the tangent heights 30 - 110 km. Deviations from local thermodynamic equilibrium (LTE) are taken into account for 20 vibrational transitions forming (nu) ₁ and (nu) ₃ bands. The monochromatic limb radiance in line contours has been investigated for fundamental and hot bands for LTE and non-LTE. Variations of radiance stipulated by variations of kinetic, vibrational temperatures and ozone concentration have been analyzed. For day- and night-time limb radiance spectra have been calculated for different spectral resolutions. The approach is suggested to retrieve ozone profiles under non-LTE conditions using high-resolution limb spectral measurements (e.g., obtained by space-borne devices of MIPAS type).

Remote sensing methods (space-borne, air-borne, and ground-based) are presently widely used for obtaining information on atmospheric composition and thermal regime. Modern IR remote sensing methods are based on the validity of LTE, which is not true in the upper atmosphere (height of breaking of LTE depends on gas, absorption band and time under consideration). The non-LTE limb radiance has been calculated in the 15-micron CO2 band for tangent heights 50-110 km. The inverse problem has been formulated including the retrieval of vibrational temperature profiles for the lower vibrational states of the CO2 molecule. The method is based on measurements of high-resolution limb radiance spectra. The retrieval accuracy has been investigated depending on measurement error, spectral resolution, and spectral region used.

The paper presents a detailed software simulation package for the Global Ozone Monitoring Experiment (GOME) instrument which will fly on ERS-2 in 1994. The GOME instrument is a nadir-viewing spectrometer designed for measurements of ozone and related trace gases such as NO, NO2, ClO, BrO, OClO, HCHO, SO2, O2/O4, and H2O. Examples are presented of input and output spectra and signal-to-noise calculations for normal viewing mode (nadir observations) and for sun and moon calibration mode. The GOME instrument simulating program can be used for a variety of purposes during instrument development, such as tests and calibrations, and tests of the so-called Zero-to-One processing step. The scheme could be adapted to other optical instruments.

The Global Ozone Monitoring Experiment (GOME) is a diode array spectrometer that will make remote sensing global measurements of atmospheric constituents from the ERS-2 satellite due for launch in October 1994. Retrieval of atmospheric profiles from radiance measurements by GOME requires an accurate radiative transfer model of the atmosphere under nadir viewing conditions. The main inputs to such a forward model are the viewing geometry parameters, an appropriate choice of model atmospheric profiles, and an accurate spectral data base for the retrievable quantities over the GOME range (240-790 nm). In this work, we shall discuss aspects of the forward model for GOME, and present some studies of theoretical precisions for retrieved parameters, based on a plane-parallel nadir viewing radiative transfer model.

The hydride VPE technique was used to grow In(x)Ga(1-x)As/InAs(y)P(1-y) detectors optimized for 1.8 micron (x = 0.58, y = 0.08), 2.2 microns (x = 0.71, y = 0.36), and 2.6 microns (x = 0.82, y = 0.60) using several compositional grading techniques and several doping levels in the heteroepitaxial layers. Additionally, 1.7 micron In(0.53)Ga(0.47)As detectors were grown on GaAs and Si substrates to study the possibility of fabricating a monolithic linear InGaAs array. Single element detectors with 75 and 500 micron diameter, and 256 and 512 element detector arrays of (25 x 500) micron pixel sizes were fabricated. The best results include a room temperature leakage current of 500 pA at 10 mV back bias for a 2.2 micron cutoff, (25 x 500) micron array pixel. High reliability (12,000 hours at 125 C) has been observed for both In(x)Ga(1-x)As/InAs(y)P(1-y) and graded In(0.53)Ga(0.47)As detector structures grown on GaAs and Si substrates. The relationship between dislocation density, leakage current, and reliability is also discussed.

The paper describes the development of spectroscopic methods for monitoring atmospheric gases, which are based on tunable diode lasers (TDLs). The problem of complexity of TDL-based systems was solved using several approaches. These include the development of a pulse-periodical mode of laser operation with high-frequency scanning rate; simplification of optical and cryogenic parts using mid-IR optical fibers; and the development of specialized electronics and automatic measurements methods. A block diagram of the system is presented together with tables listing information on the sensitivity limitations of the TDL-based spectrometer and detectivity levels for various atmospheric gases.

A simple tunable diode laser based system for routine monitoring of concentrations of atmospheric gases was developed. The instrument operation is based on a PbSnSe diode laser emitting radiation at about 4.7 microns, and InSb photodiodes. Usually, the laser operates in multimode regime with an optical power of about 0.1 mW in the mode. The system consists of the emitter-receiver unit (including optical elements and a cryostat with liquid nitrogen, which contains the diode laser and two InSb photodetectors) and the retroreflector. The instrument has two limitations: (1) the gas should have absorption lines within the mid-IR spectral region (4-8 microns) and (2) the gas should be stable enough while in the instrument's cell to allow periodic calibration of the instument. The instrument, which in the manufacturing stage, is presently used for carbon monoxide monitoring.

Recent progress in the development of submillimeter-wave instrumentation has provided access to a new spectral range (v = 625-655 GHz) to study upper atmospheric chemistry. A number of strong emission features, in particular of HCl, ClO and O3 have been measured in this frequency band with the SUMAS (Submillimeter-wave Atmospheric Sounder) experiment. The instrument has been operated successfully during the European Arctic Stratospheric Ozone Experiment (EASOE), in the period December 1991 to March 1992. We give a description of the SUMAS experiment and report on fast results obtained during EASOE.

Absorption cross-sections for HNO₃ and N₂O₅ have been measured in the wavelength region 220 - 450 nm, using a dual beam diode array spectrometer with a spectral resolution of 0.3 nm. The results for both compounds are in good agreement with recommended values at room temperature. Cross-sections of both HNO₃ and N₂O₅ show a marked reduction with decreasing temperature in the range 295 - 233 K.

Significant ozone loss due to reactive chlorine from man-made chemicals has occurred near the poles in the last decade. In this paper, we describe a novel star-pointing UV-visible spectrometer to measure amounts of some reactive gases in the ozone layer and discuss its advantages. The instrument has the capability of measuring stratospheric amounts of O3, NO2, NO3 and OClO at night. By using the most modern cooled array detectors, good signal-to-noise ratios can be obtained with a modest telescope and a short integration time. By using a two-dimensional array, light from the atmosphere adjacent to the star is measured simultaneously and subtracted from the stellar light. As with measurements using the sun as a source of light, before spectral analysis the observed spectrum at low elevation must be divided by the spectrum of the same star measured at higher elevation. This removes absorption features due to gases in the atmosphere of the star itself. The amount of absorbing constituent in the earth's atmosphere is proportional to the ratio of the slant path to the vertical path through the atmosphere. This air-mass factor is maximized, and the random error in the measurement minimized, at elevation angles close to the horizon. The instrument was deployed at Abisko in Northern Sweden during the 1991/92 European Arctic Stratospheric Ozone Expedition. Despite unusually cloudy conditions, many spectra were recorded.

The resonant and non-resonant photoacoustic (PA) detection of atmospheric trace gases is discussed with special respect to field measurements. Continuous in-situ methane measurements in a cow shed by a non-resonant PA detector applying an HeNe laser are reported, which allow the model based determination of the methane production rate of cows.

A detection system for the measurement of tropospheric OH radicals by laser-induced fluorescence has been developed. Ambient air is expanded through a nozzle into a fluorescence cell and is irradiated at low pressure by a pulsed frequency-doubled dye laser. The laser wavelength is tuned to selectively excite the OH radicals on a single rovibronic transition at 308 nm. The OH-resonance fluorescence, emitted mostly between 307 and 311 nm, is detected by a gated photomultiplier/photon counter assembly. This excitation/detection method reduces interferences due to laser generated OH efficiently far below the projected limit of detection. Calibration of our present system yields a detection limit (SNR = 2) of 8.2 x 10 exp 6 OH/cu cm for a 5-min on-resonance and 5-min off-resonance signal integration period at a laser pulse repetition rate of 20 Hz. A considerable improvement of the detection limit to 3.7 x 10 exp 5 OH/cu cm is anticipated by replacing the currently available laser system by a copper-vapor laser pumped dye laser allowing a higher repetition rate of 10 kHz. This would allow useful in situ OH measurements for testing current tropospheric chemistry models.

Intracavity laser spectroscopy is one of the most sensitive techniques for absorption measurements. The main feature of this technique is that the narrow line absorber is placed in the cavity of a laser with large homogeneously broadened gain. Laser gain compensates only broadband cavity loss, such as mirror transmission, but is not affected by narrow line intracavity absorption (ICA). The laser light passes through the absorber many times and ICA is accumulated in its spectrum like in a multipass cell. The ultimate sensitivity of the emission spectrum of a multimode laser to ICA is limited by the laser pulse duration, nonlinear mode coupling, or by quantum noise. Multimode lasers which have been applied to ICA measurements, such as dye lasers, solid state lasers, and diode lasers show different sensitivity limits. These limits are determined by specific laser parameters and can be optimized individually for each laser. The highest sensitivity limit, 3 x 10 exp -10/cm, has been achieved with CW dye lasers so far. It is determined by nonlinear coupling of laser modes in the gain media. Quantum-limited sensitivity of this laser, which corresponds to 10 exp -12/cm, allows the measurement of the absorption of single atoms in the cavity.

SF6 is a synthetic chemical which is used in many industrial applications and as a meteorological tracer. This paper describes determinations of spectral absorption coefficients k-nu of SF6 measured in the central Q-branches of the nu(3)-fundamental at 947/cm at various temperature-pressure combinations representing tangent heights in solar-occultation experiments or layers in the atmosphere. Spectral data were also obtained for C2H4 and NH3 fundamental bands. Measurements were made with the Doppler-limited spectral resolution of a tunable diode laser spectrometer with a cryogenically cooled absorption cell, described by Varanasi and Chudamani (1992).

A new tunable diode laser absorption spectrometer has been developed capable of long term operations with reduced daily maintenance, space and power requirements. This compact, self aligning unit can be easily adopted for use in laboratory, industrial, mobile or aircraft environments. Designed with fast response (10 Hz) capability, it could be easily utilized to monitor any species of interest in the 3-5 micron spectral range. When sampling at 10 Hz, the system routinely demonstrates a sensitivity of about 5 ppb for CH4.

The concept of intermodulated photoacoustics combined with the laser electric resonance spectroscopy for accurate, selective, and sensitive measurements of atmospheric pollution is discussed. Polar molecules such as ammonia, nitric acid, and water vapor adsorb to solid walls and the sampling feedline causing concentration loss and hence erroneous readings. A survey of other novel techniques for studies of ammonia and other gases is given.

The second European Remote-Sensing Satellite (ERS 2) will follow the first one, which was successfully launched in July 1991. GOME is the only new experiment on board of ERS 2, which will be launched in 1994. ERS 2 will fly on a sun synchronous polar orbit in about 785 km with a descending equator crossing time at 10:30 local time. GOME is a nadir-viewing spectrometer which observes solar radiation transmitted through or scattered from the earth atmosphere or the surface. In normal operation mode the scene is scanned across track in 3 periods of 1.5 sec and a backward scan of 1.5 sec. The instantaneous field of view of 2.8 deg x 0.14 deg is swept in this period through a default optical angle of +/- 31 deg, resulting in a ground swath of 960 km. In the ERS-2 orbit a global coverage will be possible every three days. The spectrometer covers a wavelength range from 240 to 790 nm with a spectral resolution of 0.2 to 0.4 nm. This range will be recorded by 4 RETICON diode arrays, simultaneously. Additionally broadband polarization detectors are included. Using the differential optical absorption spectroscopy technique the detection of column densities of several tropospheric and stratospheric trace gases (e.g., ozone, nitrogendioxide, water vapor, oxygen/oxygen dimer, bromine oxide) will be possible without an absolute radiometric calibration. However, an absolute calibration is possible and hence the advantage to switch the instrument in a SBUV mode for comparison with this technique.

A ranging UV-visible spectrometer to make simultaneous vertical profile measurements of O3, H2O, NO2, and other constituents throughout the troposphere from the ground is described. This technique brings together benefits of both the UV-visible and the lidar methods, creating a UV-visible spectrometer with ranging capabilities. The central feature of the instrument is a two-dimensional CCD detector array which is used to provide the time discrimination necessary to obtain range information in one dimension, as well as the spectral information in the other. The prototype instrument has a potential vertical resolution of the order of 1 km. It would be capable of providing useful profiles throughout the troposphere and under some circumstances into the low stratosphere with integration times of 1-20 minutes, depending on range. In this paper the concept will be outlined, and predicted performance characteristics of the prototype ranging spectrometer described.