Stratospheric Ozone Observations



The accurate knowledge of the distributions of ozone (03) in the global atmosphere is important for several reasons. First, the amount of ozone in the atmosphere plays a significant role in determining the amount of biologically damaging ultraviolet (UV) radiation that can reach Earth's surface. Second, ozone both absorbs and emits radiation in the atmosphere; this must be accounted for in atmospheric circulation models if they are to correctly represent the temperature and wind distributions in the atmosphere, especially in the upper troposphere and lower stratosphere. Finally, ozone together with the hydroxyl (OH) radical formed in the atmosphere in ozone photochemistry are key atmospheric oxidants. Hydroxyl plays a particularly important role by initiating much of the chemistry associated with air pollution and by being the primary destruction mechanism for several long-lived chemical compounds that contribute to global warming.

Unlike any other atmospheric phenomenon, the U.S. Congress has mandated that the National Aeronautics and Space Administration (NASA) prepare reports describing the status of our current understanding of the upper atmosphere (Public Law 101-549). In accordance with this mandate, several documents have been issued since 1985 in the form of World Meteorological Organization reports; these summaries contain a plethora of information about stratospheric ozone data as well as supporting measurements of other trace gases critical to the destruction of the stratospheric ozone layer. Much of the information in this section can be found in the last few of these reports (Albritton and Watson, 1991; Albritton et al., 1994, 1998), and the reader is referred to these studies for additional in-depth information.

Handbook of Weather, Climate, and Water: Atmospheric Chemistry, Hydrology, and Societal Impacts, Edited by Thomas D. Potter and Bradley R. Colman. ISBN 0-471-21489-2 © 2003 John Wiley & Sons, Inc.

For the purposes of understanding surface UV radiation, the main quantity for which knowledge is needed is the total column amount of ozone (the integrated ozone amount in a single "column" of the atmosphere). For atmospheric circulation models and both air pollution and chemical oxidation studies, the distribution of ozone as a function of altitude (the vertical profile) must be known. Accurate, long-term knowledge of ozone distributions is important if changes in surface UV flux, upper atmosphere temperature distributions, and concentrations of both long- and short-lived pollutants are to be quantitatively understood.

For ozone distributions to be used in quantitative studies, they must be measured with high accuracy and precision. A particularly important form of precision is long-term measurement stability, as there is strong evidence for long-term changes in ozone distribution (both total column and vertical profile) over much of Earth's surface. Without excellent (and well-characterized) long-term measurement stability, it is difficult to differentiate gradual long-term changes in actual ozone distributions from inaccuracies or drift in the measurement systems.

The measurement of ozone distributions in the atmosphere presents both a challenge and an opportunity. The challenge comes mainly from its limited amount—the amount of ozone in a given volume of air is always quite small, nearly never exceeding a mixing ratio (mole fraction) of 10 parts per million by volume (ppmv) in the stratosphere and on the order of 100 parts per billion by volume (ppbv) in the polluted troposphere. Amounts in the unpolluted troposphere can be significantly smaller than the latter figure (see Chapter 3.) The mixing ratio of ozone can vary significantly with altitude, including a significant increase with altitude on going from the troposphere into the stratosphere, as well as the existence of thin layers (laminae) of air with concentrations of ozone that differ from those of the surrounding air. In regions of severe ozone depletion (such as the lower stratosphere over the Antarctic in the Austral spring), ozone may be nearly completely absent.

The total column amount of ozone typically varies in the range of 3 x 1018 molecules/cm2 to ~1.5 x 1019 molecules/cm2. This column amount is expressed in Dobson units (DU), which correspond to the thickness of the layer of ozone (in thousandths of a centimeter) that would be formed if all the ozone in a column of air were brought to the surface. The conversion factor is 1 DU = 2.69 x 1016 molecules/cm2. The amounts given above correspond to the Antarctic in Austral springtime (~ 100 DU) and the Arctic region in late winter (~500 DU).

Other challenges to ozone measurement can stem from potential interferences. Measurement techniques based on chemical reactivity may be affected by the presence of other oxidizing species, such as sulfur dioxide (S02), while spectroscopic measurements may be affected by the presence of species with absorption features in the same wavelength regions as those of ozone. Spectroscopic measurements may also be significantly affected by the presence of aerosol particles, and most such techniques (microwave and far infrared measurements being the exception) cannot penetrate clouds. Finally, it is worth noting that ozone is present in several isotopic forms. The dominant one (~99%) consists of three atoms of the dominant isotopic form of oxygen atoms (160), but there are also forms involving I80 and 170, whose chemical abundance is ~0.3 and ~0.04% of that of 160.

Although this chapter focuses on the measurement of ozone, it is important to remember that ozone measurements cannot be understood (and especially, the causes of observed changes be understood) if measurements of related parameters, such as temperature, aerosol distributions, and other chemical constituents, are not also made. Indeed, many measurement networks, aircraft-based research platforms, and satellites make several of the measurements together to allow for maximum utility of the information obtained.


The ozone molecule contains three oxygen atoms and is shaped in the form of an isosceles triangle, with a bond angle of 117° and the length of the bond is ~0.13 nm. The electronic spectroscopy of ozone is very rich owing to the multiplicity of electronic states that arise from the combination of a triply degenerate oxygen molecule [02(3Sg )] with a ninefold degenerate oxygen atom [0(3P)], as well as the presence of relatively low-lying states of both atomic oxygen [Of' D)] and molecular oxygen [02('A), 02('S)]. The resulting electronic spectra consist of several important band systems covering the range from the ultraviolet to the near infrared. Figure 1 shows the absorption properties of the ozone molecule in the ultraviolet and visible parts of the electromagnetic spectrum. These strong spectral features—the Hartley and Huggins bands in the ultraviolet and the Chappuis band in the visible— demonstrate the potential for use of ozone spectra in its measurement. In particular, the sharp variation in ozone absorption near 320 nm shows that ultraviolet measurements shortward of this wavelength should be very useful for ozone measurements. The much weaker Chappuis band (near 600 nm) may be useful where long path lengths are available or where ozone amounts are sufficiently high that near saturation could occur with shorter wavelengths.


The measurement of the total column amount of ozone in the atmosphere goes back more than 70 years with the development of an ultraviolet technique by Dobson. This technique, still used today, has formed the backbone of all global measurement programs for ozone columns. The basic physics of this technique is relatively straightforward. UV radiation from the sun will be absorbed by ozone in the atmosphere, so ground-based measurements of surface UV flux will contain information about the integrated ozone amount in the atmosphere. As noted in the previous section, other processes, such as Rayleigh and Mie scattering involving atmospheric aerosol particles will also affect ozone measurements. The Dobson technique involves the use of pairs of UV wavelengths corresponding to features with different strengths in ozone's UV spectrum. Since the wavelength sensitivity over a relatively

Wavelength (nm)

Figure 1 Electronic absorption spectrum of ozone: (a) Hartley band and (h) Huggins bands in the ultraviolet, and (c) the weaker Chappuis band, centered near 600 nm, in the visible.

Wavelength (nm)

Figure 1 Electronic absorption spectrum of ozone: (a) Hartley band and (h) Huggins bands in the ultraviolet, and (c) the weaker Chappuis band, centered near 600 nm, in the visible.

short spectral region of the scattering by aerosols is much less than that of ozone's UV absorption spectrum, a much improved estimate of total ozone amounts can be retrieved (assuming knowledge of the ultraviolet flux from the sun is available at the two wavelengths used). Different pairs of wavelengths may be used for different amounts of ozone; typically used pairs include 312/331 nm and 318/340 nm.

The distribution of Dobson instruments increased dramatically in the 1950s, and now there is excellent coverage over much of the world with Dobson-type instruments (which included not only those operating on the above principle using a limited number of fixed wavelengths, but also other instruments such as the Brewer spectrophotometer and filter photometers used primarily in the former Soviet Union). Like all surface-based instruments, the Dobson network lacks coverage over much of the ocean-dominated Southern Hemisphere and has fewer stations in developing countries than in industrialized nations. As noted above, such surface-based measurements will not provide data in the presence of clouds, which can be a significant limitation in the tropics, where cloudiness associated with the upward part of the Brewer-Dobson circulation is a common occurrence.

Although the Dobson technique is the most common surface-based one for measurements of the total ozone column, other approaches have been used in the past. These include those using both infrared and visibie/UV wavelengths. In the latter, a variant of the Dobson technique, known as differential optical absorption spectroscopy (DOAS) is used. These other techniques have the advantage of not requiring direct sunlight to make measurements, which may be of particular importance in attempts to measure ozone in polar night (when moonlight can be sufficient for ozone measurements).

Space-Based Measurements

The primary space-based measurement technique used for measurements of total column ozone is the backscatter ultraviolet (BUV) technique. This is really a spaceborne analog of the Dobson technique, except when used from space one must account for the fact that the UV radiation passes through the atmosphere at least twice—oncc on the way from the sun to the surface (or scattering/absorbing layer) and oncc on its return to the measuring spacecraft. Account must also be taken for the UV reflectivity of the underlying ground or cloud surface. A schematic diagram of how satellite measurements are taken using four different methods is given in Figure 2. The backscatter ultraviolet (as its name implies) utilizes the ozone absorption characteristics in the ultraviolet portion of the spectrum. Occultation techniques use the properties of ozone absorption in the visible and ultraviolet wavelengths whereas the limb emission and limb scattering techniques use a knowledge of ozone absorption in the infrared and microwave portions of the electromagnetic spectrum.

In the BUV technique, the solar flux can be measured directly, although to reduce the flux to manageable levels for the observing instrument, a iLdiffuser plate" is typically deployed when the instrument looks at the sun (it is retracted for Earth t

Backscatter Ultraviolet


Backscatter Ultraviolet


Limb Emission

Limb Scattering

Figure 2 Schematic diagram showing the spacecraft and atmosphere geometry for the four common methods of acquiring trace gas measurements from satellite instruments. See ftp site for color image.

viewing). In applying the BUV technique, it is also helpful to have the measurement time close to local noon, so that the optical path lengths through the atmosphere are close to their potential minimum for the corresponding surface location. In its simplest form, the instrument looks straight down (nadir viewing) to make measurements below the satellite track (Fig. 2a). Maps of ozone column can be created by either scanning the instrument's field of view across the orbital track or using some sort of imaging detector so that observations are made corresponding to different ground locations.

The first use of this technique was on the BUV instrument aboard the Nimbus 4 satellite launched in 1970. This instrument, which was a purely nadir-viewing one, obtained data for several years. The data gave an excellent picture of both the latitudinal and seasonal nature of global column ozone distributions. These data are still of scientific interest; recently they were reexamined to help characterize Antarctic ozone amounts in the early 1970s and show that there was no evidence for significant depletion of Antarctic ozone in the springtime then.

The most significant application of the BUV techniques has been in the total ozone mapping spectrometer (TOMS) and solar backscatter ultraviolet (SBUV) series of instruments. The TOMS instruments use measurements at six wavelengths to measure ozone. For the first two TOMS instruments (one on the Nimbus 7 satellite that obtained data from October 1978 to May 1993 and one aboard a Russian Meteor-3 satellite that obtained data from September 1991 to December 1994), the wavelengths used were 312.5, 317.5, 331.2, 339.8, 360, and 380nm. The latter two are essentially unaffected by the presence of ozone and were used to provide information on surface UV reflectivity. The TOMS instruments were also shown to have information about concentrations of sulfur dioxide, especially during times of enhancement following large volcanic eruptions, aerosols (both tropo-spheric aerosols including those from biomass burning and volcanic dust, among other types and stratospheric aerosols following large volcanic eruptions), and surface UV radiative flux.

The ground resolution of these instruments is approximately 50 x 50 km at nadir (resolution is degraded as the instrument field of view scans sideways). The orbit of the Nimbus 7 satellite (sun synchronous, polar orbiting) was excellent for TOMS measurements, while that of the Meteor-3 satellite was less so since it was not sun synchronous. Roughly half the time the Meteor-3 orbit led to TOMS observations at local times sufficiently far away from noon that results must be used at great care if at all.

The newer TOMS instruments, which operated aboard the Japanese ADEOS spacecraft (Aug. 1996-May 1997) and NASA's Earth Probe (EP) satellite, use a slightly different wavelength set from the previous TOMS instruments. For these new instruments, there is an additional channel to help in the measurement of ozone at high solar zenith angles, as well as a channel to monitor the behavior of the TOMS instrument. The Earth Probe TOMS instrument was originally launched into a relatively low (—500 km) orbit to provide for better ground resolution (—26 x 26 km at nadir) than that of the ADEOS TOMS instrument (42 x 42 km), which flew aboard the higher orbiting ADEOS spacecraft. At the lower altitude, TOMS could not obtain full daily maps over the entire sunlit Earth, however, as there were interorbit gaps equatorward of approximately 60° latitude. Following the ADEOS failure, the EP satellite was boosted into a higher orbit (—750 km) to allow for near global spatial coverage.

The SBUV series of instruments includes the original SBUV instrument, which flew aboard the Nimbus 7 satellite, and updated instruments (SBUV/2) that flew aboard several of the operational meteorological satellites of the American National Oceanic and Atmospheric Administration (NOAA) on the NOAA-9, NOAA-11, and NOAA-14 satellites, to date. The SBUV instruments, which also have a capability to determine ozone vertical profile (see Section 4 of this chapter) do not have any cross-track scanning capability and thus do not obtain contiguous daily maps as do the TOMS instruments; they simply obtain data along the daytime subsatellite tracks (the nature of the BUV technique, which requires the presence of sunlight, precludes nighttime data). Long-term calibration information for the SBUV instruments was provided by the Shuttle-borne SBUV (SSBUV) instrument, which flew eight times on the Space Shuttle from 1989 to 1996.

The TOMS and SBUV series of instrument have provided an invaluable database on the total ozone distribution of Earth's atmosphere and its many variations. In Figure 3, a two-dimensional representation (latitude-time) of total ozone distributions derived from the TOMS satellite is shown. Key elements of the total ozone distribution are evident— low total ozone with little seasonal variation in the tropics, highest total ozone values in late winter at high northern latitudes, and lowest total ozone values associated with Antarctic ozone depletion at high south em latitudes during the Austral spring. The long-term changes in the global amount of total ozone determined from Nimbus-7 TOMS is shown in Figure 4a. These data have gone through extensive calibration procedures and comparisons with ground-based Dobson stations to ensure the greatest possible accuracy. The greatest changes have occurred over the Antarctic continent and is seasonal in extent. On the other hand, no statistically significant ozone depletion has been noted in the tropics (see

Figure 3 (see color insert) Two-dimensional (latitude/season) representation of total column ozone as measured by TOMS for the period 1978 to 1993. See ftp site for color image.



Jan Apr Jul Oct

Figure 3 (see color insert) Two-dimensional (latitude/season) representation of total column ozone as measured by TOMS for the period 1978 to 1993. See ftp site for color image.

86 Year

Trend * -2.20% per decade Uncertainty = 1.13% per decade Con = 296.16 DU

Figure 4 («) Long-term trend determined from TOMS data over the period from 1978 to 1994 a tier correcting for seasonal cycles, the II-year solar eyele, and the quasi-biennial oscillation; (b) month vs. latitude analysis of ozone trend plotted in contours of percentage loss per year. Shaded areas indicate areas where trend is not definitive. See ftp site for color image.

86 Year

Trend * -2.20% per decade Uncertainty = 1.13% per decade Con = 296.16 DU

TOMS Total Ozorte for October 16,1999

orotn -J o ro m -j orooi -Jorooi -jo ooiooiouiom ouiocnooiocno

Oobson Units Dark Gray < 100, Red > 500

Figure 5 (see color insert) Map of total column ozone over the Antarctic as determined from TOMS October 16, 1999. See ftp site for color image.

Fig. 4b). An example of the daily mapping capability of the TOMS instruments is shown in Figure 5, in which a contour map of total ozone distributions during the height of the Antarctic ozone depletion season is presented. The regions of ozone depletion, surrounded by the "collar" region of higher ozone amounts, are clear.

Several limitations of TOMS and SRU V data are worth noting. In particular the UV wavelengths used do not penetrate clouds and arc less sensitive to ozone in the lowest tew kilometers of the troposphere. Thus, variations in cloudiness or the nature of the tropospheric ozone profile can alTect the retrieved ozone column amounts. Improved understanding of these limitations has been an important goal of much recent research.

Another measurement technique used to obtain measurement of a "total ozonelike quanlily" is infrared emission. The TIROS operational vertical sounder (TOVS)

instruments measure infrared radiation at 9.6 |im (corresponding to one of the fundamental vibrational modes of ozone). TOVS is mainly sensitive to lower stratospheric ozone and as such does not provide a true total column measurement. However, the correlations between total ozone and lower stratospheric ozone are well established (since it is the lower stratosphere where most ozone is found), so the TOVS product has many of the same characteristics as TOMS total ozone. Since the TOVS measurement uses infrared emission, data can be obtained in regions without sunlight, such as high latitudes during polar night.

A space-based version of the DOAS technique has been implemented aboard the European Space Agency's ERS-2 satellite using the Global Ozone Monitoring Experiment (GOME). ERS-2 was launched in April 1995. The GOME instrument uses a broader wavelength range than do the TOMS or SBUV instruments, including longer wavelengths.


The first ground-based ozone profiling technique to be used was the Umkehr method in which the solar zenith angle dependence of Dobson-type measurements is used to determine the vertical profile. This technique obtains data at ~5 km vertical resolution. It provides little information on the lowest ~20 km of the atmosphere, however. In the middle and upper stratosphere, the Umkehr data record has provided an important source of information, especially on long-term ozone trends.

Another ground-based technique for obtaining information on the ozone vertical profile is the use of microwave emission. Since ozone molecules occupy a broad range of rotational states at atmospheric temperatures, there are numerous transitions that will take place for which emission-based remote sensing may be used. Information on the vertical profile comes from the shape of the observed emission lines because of the pressure-broadened nature of the emission lines—emission from ozone in the upper stratosphere will take place near the center of the spectral band while that from lower down will occur in the wings of the line. Although the vertical resolution of this technique is somewhat limited, it can provide valuable information, especially when measured together with distributions of ozone-destroying free radicals such as CIO or H02.

Another technique for ground-based measurement of ozone is lidar. In the lidar technique, a pulse of laser light is sent up from the ground and the scattered signal that returns to the ground provides information on the composition of the air mass being observed, while the time delay between the laser shot and the return signal is used to provide altitude information. Since the air mass being sampled will interact with laser light by other processes besides ozone absorption (such as molecular Rayleigh scattering, as well as aerosol scattering), lidar systems typically employ two laser wavelengths, one of which is more strongly absorbed by ozone than the other. The wavelength dependence of the aerosol and Rayleigh scattering is typically much less than that of ozone (and relatively well understood) allowing for retrieval of ozone amounts. The wavelengths used will depend to some extent on the altitude range at which ozone measurements are desired. For stratospheric measurements, where larger ozone abundances are typically observed than in the troposphere, wavelengths with a smaller absorption cross section are needed than for tropospheric measurements. Typical wavelength pairs used are 308 and 355 nm for stratospheric ozone lidar and 288 and 299 nm for tropospheric lidar. Lidar can also be implemented from aircraft, in both upward and downward looking configurations.

In Situ Measurement Techniques

The primary in situ measurement technique used for determination of the ozone vertical profile is that used on ozonesondes. In one standard implementation, an iodine/iodide redox concentration cell is used. An electric current is generated when air containing ozone is pumped into the cell, with the amount of current being related to the partial pressure of ozone in the air mass being sampled. This technique is capable of providing excellent vertical resolution, and is unparalleled at determining the existence of "tongues" or "laminae" of air masses with ozone contents that differ from those of their surroundings (see examples in Chapter 1). Because of the limitations of the balloon on which they are flown, ozonesondes rarely rise above ~30 km. Ozonesonde measurements are usually only made from a fairly limited set of observing stations, and except during certain intensive field campaigns, are typically made at most weekly. Ozonesondes can be flown at various locations and do not require the presence of sunlight. Ozonesondes from the South Pole have provided an important part of our knowledge of the vertical distribution of ozone over Antarctica, for instance, especially on its seasonal variation in springtime. In Figure 6, a plot of ozone vertical profiles over Antarctic measured before and during the presence of the ozone hole are shown. The ozonesondes provide clear evidence for the near total absence of ozone in the 12 to 22 km altitude range during the period of ozone depletion.

The measurement technique used by ozonesondes requires very careful emphasis on calibration and intercomparison. In some cases due to uncertainties about operations, ozonesonde profiles are "normalized to Dobson" so that the observed profile is modified based on a scaling of the calculated integrated ozone column to that observed at a co-located or nearby Dobson station. There are also several different types of ozonesondes, whose operational characteristics differ slightly. In spite of these uncertainties, the ozonesonde record has been critical in the assessment of ozone trends in the lower stratosphere, a region (~ 15 to 20 km) that is very difficult to observe at high accuracy using space-based instruments.

Other in situ techniques for ozone measurement also exist. One used extensively aboard research aircraft is a spectroscopic technique in which the absorption of air at UV wavelengths is accurately determined. This is a very accurate technique, as the spectral information is well known and there is little opportunity for interference because of the significantly smaller abundance of most potential contaminants. This technique has been used aboard NASA's ER-2 aircraft in its flights in the lower

Temperature (deg C) —100 —90 —80 —70 —60 —50 —40 —3I

Temperature (deg C) —100 —90 —80 —70 —60 —50 —40 —3I

Ozone Partial Pressure (mPa)

Figure 6 (see color insert) Plot of vertical profile of ozone (blue and red lines) over the South Pole as measured torn ozonesondes during austral winter (July 28) and spring (October 16), 1999; temperature profile for October 16 is also shown (green line). See ftp site for color image.

Ozone Partial Pressure (mPa)

Figure 6 (see color insert) Plot of vertical profile of ozone (blue and red lines) over the South Pole as measured torn ozonesondes during austral winter (July 28) and spring (October 16), 1999; temperature profile for October 16 is also shown (green line). See ftp site for color image.

stratosphere and upper troposphere, for instance. A chemiluminescent system has also been used.

0 0

Post a comment