Heard D.E. (editor) Analytical Techniques for Atmospheric Measurement

Подождите немного. Документ загружается.

172 Analytical Techniques for Atmospheric Measurement

01e + 82e + 8

Altitude (km)

SZA

88 89 90 91 92 93 94 95 96

column density (10

molecules cm

–2

)

LP-DOAS

(

molecules cm

–3

)

Figure 3.13 Vertical proﬁles of NO

obtained from zenith-sky measurements at Mace Head, west coast

of Ireland on 2 August 2002: (a) measured NO

column density versus SZA; (b) NO

concentration

proﬁle up to 50 km retrieved from the data in (a), using an optimal estimation method (Coe et al. 2002).

The solid line in (a) is calculated from the forward model (see text). The lowermost point of the proﬁle in

(b), assuming a 2 km layer height, is in good agreement with a measurement made by LP-DOAS in the

boundary layer. (Data reproduced with kind permission of J.A. Shillito and H. Coe [UMIST].)

initial estimate of the column density time series, y

. This is compared with the observed

y, and the a priori profile is then refined iteratively to yield x, the optimised NO

profile.

The technique yields significant profile information up to an altitude of about 40 km, if

the signal-to-noise ratio is sufficiently high in spectra recorded with the very low light

levels present when the terminator is in the upper stratosphere (i.e. SZA > 95



). The

vertical resolution achievable is of the order of 3 km, depending on altitude.

Figure 3.13(b) shows the application of the OEM to the time series of NO

column

measurements in Figure 3.13(a). In this case, the bulk of NO

is located below 5 km in the

atmosphere, with a smaller peak between 25 and 30 km in the stratosphere. Also shown in

Figure 3.13(b) is a measurement of NO

made in the same location with an LP-DOAS at

the base of the boundary layer, showing that on this occasion NO

is apparently well mixed

in the first 2 km. This situation contrasts with measurements made under very clean condi-

tions in the Southern Ocean, where emission of dimethyl sulphide, which reacts rapidly

with NO

, creates a large gradient of NO

through the boundary layer (Allan et al., 2002).

UV-Visible Differential Optical Absorption Spectroscopy 173

3.4.3 Multi-axis DOAS

The multi-axis DOAS (MAX-DOAS) technique involves recording scattered light spectra

at several different zenith angles, either simultaneously or in rapid sequence, in order to

infer the coarse height resolution of an atmospheric species (Honninger & Platt, 2002).

First developed in 2002, it has been reviewed in 2004 (Honninger et al., 2004a; Wittrock

et al., 2004). The technique consists of receiving scattered photons from multiple viewing

directions above the horizon, allowing for the retrieval of better spatial information of

the species of interest. The technique is especially powerful for monitoring atmospheric

trace gases present in layers close to the earth’s surface. Figure 3.14 shows the viewing

geometry of a MAX-DOAS instrument where scattered sunlight is collected over a range

of elevation angles . The zenith-scattered light describes a relatively longer path in

the stratosphere than in the troposphere, compared to the scattered light collected by

the lower elevation angle telescopes. Furthermore, when the SZA is large, trace gas

absorption in the lowest layers of the atmosphere is enhanced because the photons

traverse a longer absorption path through the lowermost atmospheric layer (the scattered

sunlight also gets brighter at smaller , so that the signal-to-noise ratio is increased).

Thus, absorption spectra collected at large  with a telescope pointing at low  contain

information mostly about lower tropospheric constituents, whereas spectra taken with a

telescope at small (or zero)  contain enhanced absorption by stratospheric constituents

(Honninger & Platt 2002).

In order to achieve the collection of scattered sunlight at different viewing direc-

tions, different experimental set-ups can be deployed. First, observation at different

elevation angles is obtained by sequentially moving one off-axis telescope, while a second

= 90°

= 20°

= 10°

= 5°

Solar zenith

angle ϑ

Trace gas

layer

Observer

Elevation

angle α

Figure 3.14 Schematic drawing of the viewing geometries of a ground-based MAX-DOAS instrument

with four telescopes at elevation angles 5



,10



,20



and 90



above the horizon. The scattering points at

a give altitude where solar rays intercept the ﬁeld of view of the telescopes are shown as A

−A

. The

absorption path of the photons collected by the different telescopes decreases as a function of 1/ sin.

(Honninger and Platt, 2002, reproduced with permission from Elsevier.)

174 Analytical Techniques for Atmospheric Measurement

fixed telescope is sometimes employed pointing to the zenith. This approach has been

employed in various studies since 2000 (Honninger & Platt, 2002; Bobrowski et al.,

2003; Leser et al., 2003; Heckel et al., 2004; Honninger et al., 2004b,c). The set-up

consists of one or two telescopes controlled by stepper motors coupled to a single

spectrometer–detector system. The system is relatively simple to operate and therefore

suitable for making measurements in remote locations. Alternatively, a number of static

telescopes can be fixed at different elevation angles to allow simultaneous observations

at more than two viewing geometries (Wagner et al., 2002; Bobrowski et al., 2003).

The telescopes are then coupled to a single spectrometer by optic fibres, and dispersed

onto a two-dimensional CCD (Section 3.3.1.1) where the spectra are separated as strips

on different rows of the CCD. This offers the advantage of real-time observations at

several elevation angles. The disadvantage of this approach is that the spectra may have

slightly different line shapes, and the pixel-to-pixel variability across the CCD chip will

not be removed when dividing the spectra since they are recorded on different portions

of the chip. The former difficulty can be overcome to a degree by convoluting each

spectrum with the corresponding line shape of the spectrum by which it is to be divided;

the latter can be minimised by applying a pixel-to-pixel correction factor obtained by

measuring diffuse (non-dispersed) light (e.g. from a tungsten lamp with a diffuser plate)

across the CCD.

The MAX-DOAS technique yields the SCD of the species monitored at the different

elevation angles. The VCD is then obtained by applying an AMF (Section 3.4.1). The

AMF needs to be estimated from a radiative transfer model that determines the altitude

of significant light scattering, for the SZA and elevation angle employed. For instance,

for a stratospheric absorber there will be a strong dependence of the AMF on .In

contrast, for an absorber in the boundary layer the AMF will have a small dependence

on , but a strong dependence on . Figure 3.15 illustrates an example where the BrO

column abundances were strongly dependent on , indicating that a significant fraction

of BrO was in the lowest 2 km of the troposphere. This conclusion is explained by

7580859095

BrO SCD (molecules cm

–2

)

+ 14

8e + 14

+ 15

Zenith

40°

15°

10°

5°

SZA

70 75 80 85 90 95

Figure 3.15 Example of the BrO slant column density recorded at ﬁve different elevation angles, as a

function of the SZA. The strong dependence of the column abundance with elevation angle indicates

that most of the BrO lies in the boundary layer. These spectra were recorded at Barrow, Alaska, on 25

March 2003 (Saiz-Lopez and Plane, unpublished data).

UV-Visible Differential Optical Absorption Spectroscopy 175

Figure 3.16, which compares the scattering geometries for an absorber present in the

stratosphere and in the boundary layer (Honninger & Platt, 2002). For the case of single

scattering occurring below the trace gas layer (stratospheric constituent), the AMF can

be approximated as AMF ≈ 1/cos . In contrast, for scattering above the trace gas layer

(lower troposphere species) AMF ≈ 1/ sin . Of course, the measured SCD contains

contributions from both the stratosphere and the troposphere if the molecular species

is present in both regions. However, for fixed  the slope of a plot of SCD against

AMF ≈1/ sin  provides an estimate of the vertical column of the trace gas below the

scattering altitude (i.e. in the lower troposphere), whereas the intercept is an estimate of

the stratospheric component (Honninger & Platt, 2002).

Figure 3.16 Illustration of an off-axis viewing geometry for single scattering, where the absorbing trace

species occurs (A) in the stratosphere and (B) in the boundary layer. The observed SCD (along ds)is

converted into VCD (along dz) by applying an AMF. (Honninger et al., 2004a, reproduced with permission

from the European Geosciences Union.)

176 Analytical Techniques for Atmospheric Measurement

3.4.3.1 Tomographic DOAS

The LP-DOAS technique measures the column abundance of an atmospheric species

along an optical path, but does not provide information about the distribution in

the atmosphere. However, the spatial distribution of a trace species can be obtained

in two or even three dimensions by using a number of intersecting light paths and

applying tomographic inversion (Hashmonay et al., 1999; Price, 1999). This is known as

tomographic DOAS (Tom-DOAS) (Pundt et al., 2004; Laepple et al., 2004). Instrumen-

tally, it requires one or several artificial light sources in conjunction with a number of

retro-reflectors. Additionally, a two-dimensional CCD together with an imaging spectro-

graph is needed to record simultaneously the spectra of light collected from a number

of telescopes. Tom-DOAS can also be used to collect scattered sunlight by means of a

number of mirrors that will collimate the light into different optic fibres in a similar way

to MAX-DOAS.

Tom-DOAS has been applied to construct a two-dimensional map of NO

from a

vehicle emission plume above a motorway (Pundt et al., 2004). Figure 3.17 shows the

experimental set-up of the Tom-DOAS above the motorway: two telescopes together with

650 m

180 m

800

Figure 3.17 Experimental set-up of a Tomographic-DOAS to observe emissions around a motorway.

The two DOAS telescopes housed at ground level on each side of the motorway can be steered to each

of eight retroreﬂector arrays, positioned at different heights in two groups. This arrangement therefore

provides a total of 16 light paths. (Reproduced with permission from Pundt et al., 2004.)

UV-Visible Differential Optical Absorption Spectroscopy 177

Figure 3.18 A contour plot of NO

as a function of altitude and orthogonal to the motorway (indicated

by a solid black rectangle), measured by the Tom-DOAS system in Figure 3.14. GTI and HMT correspond

to two different geographical locations from where the measurements are carried out. The reconstruction

of the NO

concentration ﬁeld was made using a simultaneous algebraic reconstruction technique.

Note that the higher NO

concentrations are on the right-hand side, downwind from the motorway.

(Reproduced with permission from Pundt et al., 2004.)

eight retro-reflectors at different heights were used, providing 16 different light paths

for evaluation. Figure 3.18 illustrates the two-dimensional NO

distribution determined

using a simultaneous algebraic reconstruction technique (SART) (Pundt et al., 2004). As

expected, the higher values of NO

are observed downwind of the motorway.

3.4.3.2 Balloon-borne DOAS

Balloon-borne UV/visible spectrometers have been utilised to investigate vertical distri-

butions of atmospheric species including O

, BrO, NO

, OClO, OBrO, IO, OIO,

and H

O, by collecting direct sunlight or scattered sunlight at solar occultation

(Pommereau & Goutail, 1988; Pommereau & Piquard, 1994; Ferlemann et al., 1998;

Harder et al., 1998; Pundt et al., 1998; Erle et al., 2000; Ferlemann et al., 2000; Fitzen-

berger et al., 2000; Harder et al., 2000; Pfeilsticker et al., 2001; Lumpe et al., 2002; Pundt

et al., 2002; Bosch et al., 2003). In fact, this area is large enough to be reviewed in its own

right, and only a brief description is given here. The principle of balloon-borne DOAS is

similar to that of passive DOAS methods, in that scattered sunlight spectra are collected

at different SZAs. The observations can be performed as the balloon-gondola ascends or

by solar occultation once the balloon reaches its float altitude ∼30 km. Solar occultation

measurements begin as the sun moves down below the horizon. Pressure sensors and

GPSs can track the position of the gondola precisely. During its ascent the instrument

probes nearly the entire profile of the species of interest whereas while floating only the

fraction of the profile above the balloon altitude is probed. This is because scattered light

is collected in a near-horizontal direction with respect to the gondola. Hence, as the solar

terminator moves upwards after sunset, scattering comes from higher (and further away)

178 Analytical Techniques for Atmospheric Measurement

in the atmosphere. Generally, the float-occultation method provides stronger observable

SCDs (Pommereau & Piquard, 1994).

The optical layout of a balloon-borne UV/visible spectrometer is similar to ground-

based DOAS instruments. However, there are difficulties inherent in balloon flight. For

instance, spectral shifts can be caused by rotation of the gondola, and therefore the

entrance slit, with respect to the solar position. Some instruments employ a diffuser plate

in the optical entrance to minimise this problem. Note that for tropospheric applications

direct sunlight is often collected, without a diffuser, as the light describes a well-defined

path in the troposphere (Fitzenberger et al., 2000; Harder et al., 2000). Pressure and

temperature changes during flight can also cause spectral drift. This problem can be

overcome by enclosing the entire spectrometer in a pressurised and thermostated unit.

Additionally, the balloon needs to be maintained at a fixed azimuth angle, requiring an

azimuthal suspension system on the gondola.

Finally, the spectral analysis by DOAS requires the solar features that dominate the

absorption spectra to be removed (Section 3.4.1). The vertical distributions of species can

be retrieved by inversion methods (Rodgers, 1976; Pommereau & Piquard, 1994), with a

height resolution of about 1 km (Pundt et al., 2002).

3.4.3.3 Airborne MAX-DOAS

Absorption spectroscopy measurements can also be performed from airborne platforms

(Wahner et al., 1989, 1990a,b; Schiller et al., 1990; Glatthor et al., 1998; McElroy et al.,

1999; Petritoli et al., 2002, see also Chapter 2 for infrared applications). For instance, the

column abundances of species such as NO

, BrO and OClO have been monitored from

airborne instruments.

The technique is based on the collection of UV/visible scattered sunlight, as a

function of SZA, by a telescope–spectrometer system installed on an aircraft. The spectral

analysis is performed using the scattered sunlight DOAS procedure (Sections 3.1.2

and 3.4.1). Numerous ‘off-axis’ scattered sunlight measurements with an airborne

UV/visible spectrometer have been made (Petritoli et al., 2002). Indeed, MAX-DOAS

can also be applied from an airborne platform at different viewing directions above

and below the aircraft (termed AMAX-DOAS) (Heue et al., 2003; Wang et al., 2003).

Figure 3.19(a) compares the atmospheric observation geometries for an aircraft (assuming

single scattering), compared to a ground-based zenith-sky instrument (Figure 3.19(b).)

This shows the different paths taken by scattered light as it passes through trace

gas layers in the troposphere and stratosphere. The upward-viewing geometry scans

the atmosphere above the flight altitude whereas the downward-viewing telescopes

collect scattered sunlight from above and below the aircraft. Following the MAX-

DOAS approach, telescopes pointing to the altitude flight horizon (limb-viewing) will

observe larger absorptions than those either looking towards the zenith or nadir.

Depending on the scattering altitude two different modes of viewing can be employed.

The nadir-viewing mode is used when the scattering occurs close to, or at, the earth’s

surface; limb-viewing is used when light is scattered above the surface. Therefore,

depending on the flight altitude, the combination of spectra collected at different

SZA by telescopes pointing at various elevation angles allows for discrimination

between the contributions of the stratospheric and the tropospheric components to the

UV-Visible Differential Optical Absorption Spectroscopy 179

tangent

height

(TH)

(a)

(b)

Figure 3.19 Different scattered sunlight paths seen from an airborne instrument (a), compared to those

observed with a ground-based zenith-sky spectrometer (b). Note that the photon paths are shown for the

simple case of single scattering. The upper and lower grey layers in panel (b) represent stratospheric and

tropospheric components respectively. (Reproduced with permission from Prof. Ulrich Platt [University

of Heidelberg].)

observed column density. The vertical profile information can then be derived by applying

the corresponding AMF as in ground-based MAX-DOAS analysis.

3.5 Summary and conclusions

The aim of this chapter has been to describe the very significant advances that have

been made in using UV/visible differential optical absorption spectroscopy to study the

atmosphere, since the last major reviews of the field in 1994/95 (Platt, 1994; Plane & Smith,

1995). And, of course, we have excluded here a discussion of the remarkable progress that

have been made in space-based DOAS observations of atmospheric constituents, which

are covered briefly in Chapter 1. To summarise, the strengths inherent in the DOAS

180 Analytical Techniques for Atmospheric Measurement

method are that any species with a highly structured spectrum in the UV/visible/near-IR

region can be observed if its concentration is significant (Table 3.1), and the retrieved

column density is directly traceable to an absolute laboratory spectrum. Since the late

1990s, several new species have been observed (uniquely by DOAS). These are IO, OIO

and I

. BrO and ClO have been observed for the first time in the lower troposphere.

There have also been several major advances in DOAS techniques. These include the

development of CRDS for making in situ measurements, and techniques such as MAX-

DOAS and Tom-DOAS for obtaining spatial structure. Of particular note here has been

the retrieval of vertical profiles of BrO and NO

. In conclusion, the rate of progress in

developing the theory and application of the DOAS method has been so rapid since 2000

that one can look forward with great anticipation to where the field will be in 2015.

Further reading

For more information on general atmospheric chemistry and spectroscopy, the following

books are recommended:

Wayne, R.P. (2000) Chemistry of Atmospheres, Oxford University Press, Oxford.

Finlayson-Pitts, B.J. & Pitts, J.N. (2000) Chemistry of the Upper and Lower Atmosphere,

Academic Press, San Diego.

Hollas, J.M. (1996) Modern Spectroscopy, John Wiley & Sons Ltd, New York.

References

Alicke, B., Hebestreit, K., Stutz, J. et al. (1999) Iodine oxide in the marine boundary layer, Nature,

397, 572–573.

Aliwell, S.R. & Jones, R.L. (1996a) Measurement of atmospheric NO

. 1. Improved removal of

water vapour absorption features in the analysis for NO

. Geophys. Res. Lett., 23, 2585–2588.

Aliwell, S.R. & Jones, R.L. (1996b) Measurement of atmospheric NO

. 2. Diurnal variation of

stratospheric NO

at midlatitude, Geophys. Res. Lett., 23, 2589–2592.

Aliwell, S.R., Jones, R.L. & Fish, D.J. (1997) Mid-latitude observations of the seasonal variation of

BrO. 1. Zenith-sky measurements, Geophys. Res. Lett., 24, 1195–1198.

Aliwell, S.R. & Jones, R.L. (1998) Measurements of tropospheric NO

at midlatitude, J. Geophys.

Res.-Atmos., 103, 5719–5727.

Aliwell, S.R., Van Roozendael, M., Johnston, P.V. et al. (2002) Analysis for BrO in zenith-sky

spectra: An intercomparison exercise for analysis improvement, J. Geophys. Res.-Atmos., 107,

art. no.-4199.

Allan, B.J., Carslaw, N., Coe, H. et al. (1999) Observations of the nitrate radical in the marine

boundary layer, J. Atmos. Chem., 33, 129–154.

Allan, B.J., McFiggans, G., Plane, J.M.C. et al. (2000a) Observations of iodine monoxide in the

remote marine boundary layer, J. Geophys. Res.-Atmos., 105, 14 363–14 369.

Allan, B.J., McFiggans, G., Plane, J.M.C. et al. (2000b) The nitrate radical in the remote marine

boundary layer, J. Geophys. Res.-Atmos., 105, 24 191–24 204.

Allan, B.J., Plane, J.M.C., Coe, H. et al. (2002) Observations of NO

concentration profiles in the

troposphere, J. Geophys. Res.-Atmos., 107, art. no.-4588.

Allan, B.J., Plane, J.M.C. & McFiggans, G. (2001) Observations of OIO in the remote marine

boundary layer, Geophys. Res. Lett., 28, 1945–1948.

UV-Visible Differential Optical Absorption Spectroscopy 181

Armerding, W., Spiekermann, M. & Walter, J. (1996) Multipass optical absorption spectroscopy:

A fast-scanning laser spectrometer for the in situ determination of atmospheric trace-gas compo-

nents, in particular OH, Appl. Optics, 35, 4206–4219.

Arpag, K.H., Johnston, P.V., Miller, H.L. et al. (1994) Observations of the Stratospheric BrO

Column over Colorado, 40 Degrees N, J. Geophys. Res.-Atmos., 99, 8175–8181.

Ashworth, S.H., Allan, B.J. & Plane, J.M.C. (2002) High resolution spectroscopy of the OIO radical:

Implications for the ozone-depleting potential of iodine, Geophys. Res. Lett., 29, art. no.-1456.

Atkinson, D.B. & Hudgens, J.W. (1997) Chemical kinetic studies using ultraviolet cavity ring-

down spectroscopic detection: Self-reaction of ethyl and ethylperoxy radicals and the reaction

→ C

, J. Phys. Chem. A, 101, 3901–3909.

Axelsson, H., Galle, B., Gustzsson, K. et al. (1990) A transmitting/receiving telescope for DOAS-

measurements using retroreflector technique, Dig. Top. Meet. Opt. Remote Sens. Atmos., OSA,

4, 641–644.

Ball, S.M., Povey, I.M., Norton, E.G. et al. (2001) Broadband cavity ringdown spectroscopy of the

radical, Chem. Phys. Lett., 342, 113–120.

Ball, S.M. & Jones, R.L. (2003) Broad-band cavity ring-down spectroscopy, Chem. Rev., 103,

5239–5262.

Bobrowski, N., Honninger, G., Galle, B. et al. (2003) Detection of bromine monoxide in a volcanic

plume, Nature, 423, 273–276.

Bosch, H., Camy-Peyret, C., Chipperfield, M.P. et al. (2003) Upper limits of stratospheric IO and

OIO inferred from center-to-limb-darkening-corrected balloon-borne solar occultation visible

spectra: Implications for total gaseous iodine and stratospheric ozone, J. Geophys. Res.-Atmos.,

108, art. no.-4455.

Brauers, T., Hausmann, M., Bister, A. et al. (2001) OH radicals in the boundary layer of the Atlantic

Ocean. 1. Measurements by long-path laser absorption spectroscopy, J. Geophys. Res.-Atmos.,

106, 7399–7414.

Brauers, T., Hausmann, M., Brandenburger, U. et al. (1995) Improvement of differential optical-

absorption spectroscopy with a multichannel scanning technique, Appl. Optics, 34, 4472–4479.

Brown, S.S., Stark, H., Ciciora, S.J. et al. (2001) In-situ measurement of atmospheric NO

and

via cavity ring-down spectroscopy, Geophys. Res. Lett., 28, 3227–3230.

Brown, S.S. (2003) Absorption spectroscopy in high-finesse cavities for atmospheric studies, Chem.

Rev., 103, 5219–5238.

Brown, S.S., Stark, H., Ryerson, T.B. et al. (2003) Nitrogen oxides in the nocturnal boundary layer:

Simultaneous in situ measurements of NO

,NO

, NO, and O

, J. Geophys. Res.-Atmos.,

108, art. no.-4299.

Camy-Peyret, C., Bergqvist, B., Galle, B. et al. (1996) Intercomparison of instruments for tropo-

spheric measurements using differential optical absorption spectroscopy, J. Atmos. Chem., 23,

51–80.

Carslaw, N., Carpenter, L.J., Plane, J.M.C. et al. (1997a) Simultaneous observations of nitrate and

peroxy radicals in the marine boundary layer, J. Geophys. Res.-Atmos., 102, 18 917–18 933.

Carslaw, N., Plane, J.M.C., Coe, H. et al. (1997b) Observations of the nitrate radical in the free

troposphere at Izana de Tenerife, J. Geophys. Res.-Atmos., 102, 10 613–10 622.

Chance, K.V. & Spurr, R.J.D. (1997) Ring effect studies: Rayleigh scattering, including molecular

parameters for rotational Raman scattering, and the Fraunhofer spectrum, Appl. Optics, 36,

5224–5230.

Coe, H., Allan, B.J. & Plane, J.M.C. (2002) Retrieval of vertical profiles of NO

from zenith sky

measurements using an optimal estimation method. J. Geophys. Res.-Atmos., 107, art. no.-4587.

Comes, F.J., Armerding, W., Grigonis, R. et al. (1992) Tropospheric OH – Local measurements

and their interpretations. Ber. Bunsen-Ges. Phys. Chem. Chem. Phys., 96, 284–286.

de Beek, R., Vountas, M., Rozanov, V.V. et al. (2001) The ring effect in the cloudy atmosphere.

Geophys. Res. Lett., 28, 721–724.