Retrieval of volcanic SO2 column abundance from Atmospheric Infrared Sounder data



[1] The concentrations of volcanic sulphur dioxide (SO2) in the upper troposphere and lower stratosphere are inferred using new infrared measurements made by the Atmospheric Infrared Sounder (AIRS). Column abundance of SO2 is derived using the strong SO2 absorption feature near 1362 cm−1 (ν3-band). The retrieval takes into account interference from water vapor across the band. Examples from several recent volcanic eruptions are given to illustrate the technique and the retrievals are compared to contemporaneous SO2 ultraviolet measurements from the Ozone Monitoring Instrument (OMI), Global Ozone Monitoring Experiment (GOME) and Total Ozone Mapping Spectrometer (TOMS).

1. Introduction

[2] Volcanic eruptions can emit large quantities of sulphur dioxide and other gases into the atmosphere. These gaseous emissions are mostly confined to the lower troposphere (below ∼3 km), but can also reach the upper troposphere (heights > 3 km) or lower stratosphere (heights >8–15 km) where the gases may have long lifetimes (>days) and, in the case of SO2, convert to acid aerosol and affect the radiative budget of the atmosphere. It is of some interest to be able to quantify the amount of SO2 emission to the atmosphere from volcanoes and also to determine the eventual height of the gas in the atmosphere. Because of the sporadic nature of volcanic eruptions and the large geographic extent of volcanoes, measurements from orbiting satellites are well suited to this task. The Total Ozone Mapping Spectrometer (TOMS) and more recently, the Global Ozone Monitoring Experiment (GOME) and the Ozone Monitoring Instrument (OMI), use reflected ultraviolet sunlight to determine column SO2 amounts. TOMS has been delivering global SO2 measurements for nearly 26 a. These SO2 measurements have restricted spatial resolutions, about 50 × 50 km2 for TOMS and 320 × 40 km2 for GOME, and temporal frequency, once per day for both TOMS and GOME, daylight hours only. OMI provides better spatial resolution (17 × 17 km2), but at the time of writing only limited SO2 products are available to the research community.

[3] Recently, Prata et al. [2003] have shown that the High-Resolution Infrared Sounder (HIRS) on board the operational NOAA polar orbiters can be used to infer SO2 column abundance by utilizing absorption measurements in a water vapor channel centered near to 7.34 μm (1362 cm−1), which also happens to cover the very strong antisymmetric stretch absorption feature (ν3-band) of the SO2 molecule. This lucky coincidence has permitted the use of this channel to determine global SO2 by a completely different method to that used by TOMS, OMI and GOME. The HIRS satellite data record extends back to 1979. Since the method uses infrared radiation from a polar orbiter, the temporal frequency is twice that of the ultraviolet sensors, and the spatial resolution is about 17 × 17 km2 at nadir. A similar SO2 retrieval scheme has been devised by A. J. Prata et al. (Volcanic sulphur dioxide concentrations derived from infrared satellite measurements, submitted to Journal of Geophysical Research, 2007) using Moderate resolution Infrared Spectroradiometer (MODIS) data, which has 1 × 1 km2 spatial resolution at nadir and up to four samples per day from two polar platforms (Terra and Aqua).

[4] Carn et al. [2005] have described a retrieval scheme for AIRS based on detailed radiative transfer calculations and have provided examples for the 2002 eruption of Mt Etna, Sicily. In this paper we demonstrate that the AIRS instrument can be used to infer SO2 column abundance, by using off-line radiative transfer and an ad hoc spectral correlation technique. The infrared retrieval makes use of the same SO2 band as that exploited for HIRS and MODIS, but because of the high spectral resolution of AIRS (up to 2378 channels) a different approach is taken to the retrieval. The paper begins with a description of the AIRS instrument, followed by an explanation of the retrieval scheme. The retrieval is then illustrated using examples from several recent eruptions. A comparison with OMI, TOMS and GOME is included and the use of trajectory modeling is advocated as a useful constraint on the retrieval as well as providing a means for validation. A comprehensive error budget for the retrieval is beyond the scope of this paper, but we provide an indication of the error bounds.

2. AIRS Instrument

[5] AIRS is an echelle grating spectrometer operating at infrared wavelengths between 650 and 2700 cm−1 [Chahine et al., 2006]. The instrument is on board the EOS-Aqua polar orbiting satellite at an altitude of about 705 km above the Earth's surface in a Sun-synchronous orbit with local equator crossing times of 1330 Local time (LT) and 0130 LT. The primary goal is to provide global atmospheric retrievals of temperature, moisture and other gases for numerical weather prediction and climate. AIRS scans a swath of ±49° from nadir with an instantaneous field of view of 1.1° providing nadir pixels with dimensions 15 × 15 km2, increasing to 18 × 40 km2 at the swath edge. There are 90 pixels along each scan line and each image granule is divided into 135 along-track lines for convenience in processing and product delivery. A rigorous quality control procedure that ensures accurate wavelength calibration for each of the 2378 channels and a rejection protocol for substandard pixels (poor system response function) is applied. Absolute spectral calibration accuracy is estimated to be less than 0.5% of the detector system response function (SRF) full width at half maximum. Preliminary validation indicates an absolute radiometric accuracy of better than 0.5 K, but this does vary with channel. A full description of the AIRS instrument and data processing can be found at and references therein.

3. Data Analysis

[6] The AIRS data used in this study have been downloaded via anonymous FTP from the AIRS data processing center at NASA Goddard. These data are level 1b v4.0 calibrated and navigated radiances on an image grid corresponding to 2378 channels by 90 pixels by 135 lines [Manning, 2002]. Channels that are out of specification or noisy are removed from the analysis; this amounts to about 275 channels and does not affect the retrieval scheme. No other data quality control procedures are applied. The SRFs for the channels used in the generation of the layer SO2 absorption cross sections (see next section) were obtained via anonymous ftp from (S. Hannon et al., unpublished manuscript, 2002) (available from

4. Retrieval Scheme

[7] The retrieval scheme is a two-step process. In the first step pixels that contain SO2 are identified. In the second step a least squares procedure is used to find the amount of SO2 in the pixel, based on off-line radiative transfer calculations. The upwelling radiance received by AIRS is assumed to consist of emission from the surface and from the atmosphere,

equation image

[8] Iν,s is the radiance emitted from the surface, Bν is the Planck function, ν is wave number (cm−1), T(z) is the temperature as a function of height z, τ is the transmittance and qi(z), i = 1…n are constituent profiles of SO2, H2O, CO2, O3 etc. The aim of the analysis is to retrieve the column abundance, u1 of SO2, which is related to the constituent profile by,

equation image

[9] A portion of the AIRS spectrum between 1295 and 1405 cm−1 is used for the retrieval. Although SO2 has absorption features at 2500 cm−1 and 1160 cm−1 these are not used in the retrieval. The 2500 cm−1 feature is often discernible in the data, while only the long wavelength side of the 1160 cm−1 feature can be identified in AIRS because there are no channels between 1170 and 1180 cm−1. By using the 1295–1405 cm−1 interval, the surface emission term in (1) can be neglected, because the atmosphere is effectively black in this interval. This is demonstrated in Figure 1a where the transmittance at the top of the atmosphere as a function of wave number is shown for a US standard atmosphere, with background SO2 (and other gases) using the MODTRAN-3 radiative transfer model. The transmittance profile of the SO2ν3-band is also shown. Except for two small regions between 1320–1335 cm−1 and 1342–1358 cm−1, the atmosphere appears to be opaque, and this is due principally to water vapor absorption. As the water vapor resides mostly in the lowest layers, closest to the surface, higher in the atmosphere the transmittance is higher. At some point in the atmosphere, which depends on the water vapor and SO2 amounts, the atmosphere is sufficiently transparent and the SO2 amount sufficiently large that the signal from the SO2 dominates over that from the water vapor, and if the signal is greater than the instrumental noise, AIRS is able to detect and quantify the SO2 absorption.

Figure 1.

(a) Top-of-atmosphere transmittance as a function of wave number in a background atmosphere and for atmospheres containing enhanced SO2 gas. The ν3 SO2 absorption feature is also depicted. (b) Ratio of path radiance to total radiance as a function of height in the atmosphere for a background atmosphere and for an atmosphere containing a single layer of 50 DU (column abundance) SO2 at 5 km. The radiance difference between these profiles, above the SO2 layer, gives an indication of the change in radiance above the SO2 layer due to the layer.

[10] The total column background atmospheric SO2 in the absence of volcanic activity is typically less than 1 Dobson unit (DU) (1 DU = 2.6849 × 1016 molecules cm−2) (<0.2 DU in the boundary layer). We assume that the SO2 lies in a layer at z = z1 to z = z2 (above the boundary layer) so that equation (1) can be written,

equation image

[11] The AIRS retrieval relies on being able to correctly identify pixels within the image granule that are affected by SO2. To do this we assume that the transmission of radiation within this restricted band, for each pixel, follows the Beer-Bougier-Lambert law:

equation image

where Iν is the radiance at wave number ν leaving the SO2 layer measured at the satellite (term 2 in equation (3)), Iν,0 is the radiance entering the SO2 layer from below and is equivalent to term 1 in (3), and k is the absorption coefficient. We assume that the radiance contributions from the atmosphere above the SO2 layer (term 3 in (3)) are the same with or without the SO2 layer. The radiance contributions reflected off the SO2 layer are assumed to be negligible. Figure 1b shows the ratio of the path radiance to the surface emission at the top of the atmosphere for a background atmosphere, as a function of height, together with the same ratio for an atmosphere containing ∼50 DU of SO2 placed in a single layer at ∼5 km. This is the ratio of term 2 on the right-hand side of (1) to Iν, with the lower limit of the integral in term 2 replaced with z. Above the SO2 layer the radiances are almost identical and small (less than 10% of the total emission). It can be seen that the difference between the radiance profiles with and without the SO2 layer are very small, and we are justified in neglecting this contribution in the light of more serious sources of error. We treat the radiation field as isotropic and assume that the atmosphere is locally horizontally homogeneous below and above the SO2 layer. From (4) we can determine the absorbance spectrum,

equation image

[12] For a single spectral Lorentz line or band and a single absorbing gas, the absorption can be written,

equation image

where α is the line half-width and ν0 is the location of the line center. For a homogeneous path, the integral can be solved to yield [Goody, 1964],

equation image

where S is the line strength, and L0 and L1 are modified Bessel functions. The weak, and strong absorption limits give a linear and square-root dependence respectively, of the absorption (A) on the absorber amount (u). The absorption due to a single spectral line or spectral band is often referred to as the equivalent width measured in units of cm−1. In practice the absorbance is due to many lines, the path is inhomogeneous and there may be lines due to multiple absorbing gases, some with overlapping lines. Because the path is inhomogeneous there is a temperature and pressure dependence of the line parameters with height. A series of MODTRAN-3 simulations was carried out to examine how well this absorption model holds for the 7.3 μm SO2ν3-band. Figure 2 shows a plot of the absorption (cm−1) versus absorber amount (milli atm-cm or DU) for an SO2 layer inserted at two heights in the atmosphere, for amounts up to 120 DU. Up to about 40 DU the relationship is linear, and follows the weak-line absorption limit. Beyond about 50 DU, the absorption is nonlinear and follows the square-root dependence of the strong-line limit. At high absorber amounts, the absorption tends to a constant and presumably the band is saturated resulting in a low sensitivity. Despite the presence of other interfering gases, the simulations suggest that the Lorentz band model is quite accurate. Calculation of the absorbance spectrum requires identification of a background or reference pixel (pr, lr), from which the reference radiance Iν,0 is determined. This pixel is found by comparing the absorbance spectrum of each pixel with a synthetic spectrum calculated using library line strengths (the HITRAN 96 database is used; Rothman et al. [2003], MODTRAN-3 and a standard atmosphere perturbed by 100 DU of SO2. Spectra were calculated using varying amounts of SO2 (from 10 DU up to about 120 DU) and all have similar shapes and are highly correlated. For different atmospheres with different amounts of interfering gases, there are only small changes in the spectral features as determined by these simulations. The method of determining the optimal reference pixel relies solely on the degree of correlation between the absorbance spectrum computed from,

equation image

and that computed from,

equation image

where Is is the synthetic radiance spectrum (1295–1405 cm−1) with 100 DU of SO2 and I0 is the synthetic spectrum with background SO2. Ipt,lt and Ipr,lr are measured AIRS radiance spectra (functions of ν; the reference to ν has been dropped for notational convenience) for the target pixel [pt, lt] and the reference pixel, [pr, lr], respectively, and p and l represent pixel and line number. The R2 correlation is calculated from,

equation image
Figure 2.

Equivalent width or absorption as a function of absorber amount determined from radiative transfer calculations for single SO2 layers at 5.3 km and 17.3 km. For SO2 columns <40 DU, the variation of absorption with absorber amount is approximately linear, while above 50 DU there is a square-root dependence.

[13] The ordinates of the spectrum occur at discrete values of wave number, νi that are determined by the AIRS instrument characteristic, and n is the number of channels used (n ≈ 140). equation image and equation image are the normalized measured and synthetic absorbance spectrums, defined as,

equation image
equation image

[14] The reference pixel is deemed to be that pixel which produces the highest R2 correlation over all other reference pixels in the image. Generally, this pixel is geographically close to the target pixel and thus the assumption that the atmospheres of the target and reference pixel be similar is likely to be met for each SO2 pixel. The correlation is lowest when other gases interfere with the spectral matching, or when the SO2 column amount is low or when clouds and collocated water vapor exist within the SO2 layer or above it. The relation between the R2 correlation and the column SO2 need not necessarily be linear or even positive. High SO2 amounts may occur in the presence of enhanced water vapor loadings and the spectral matching may, in this case, produce a low correlation. Note that it is the spectral shape that determines the correlation, not the absolute absorber amount. For thin SO2 layers in very dry atmospheres over cold surfaces many of the underlying assumptions of the retrieval scheme break down and the R2 correlations can be quite low regardless of the amount of SO2 present.

[15] The procedure is time consuming, but is objective and optimal in some sense. Once the reference pixels have been found (there will be potentially up to one less reference pixels as there are target pixels), the absorbance spectra for all pixels with R2 correlations above a specified value are calculated and a correlation image (containing the R2 correlations between the synthetic and observed spectra for all pixels) determined. The correlation image and the reference pixel arrays are retained during the analysis procedure. Figure 3 shows a comparison between the observed absorbance spectrum (dotted black line) and the synthetic spectrum (continuous red line) for six pixels with different degrees of calculated correlation for the 10 May 2003 eruption of Anatahan (see later). Also shown in Figure 3 are the retrieved SO2 column amounts derived during the second stage of the retrieval process (see later). The correlations vary from R2 = 0.46 (Figure 3a) to R2 = 0.96 (Figure 3f); in this case the pixel with the highest R2 correlation also contains the highest SO2 column amount. The location (pixel number and geographic coordinates) are also given on Figure 3 and it can be seen that in all cases but one, the reference pixel is on the same line number as the target pixel and in all cases the reference pixel is clear of any retrieved SO2 (see Figures 5a and 5b in section 5.1). There are some possible scenarios where the spectral matching might produce erroneous results. These might be when the reference pixel contains some SO2, leading to an underestimate of the target SO2 or when the reference pixel is geographically distant from the target pixel, and hence the atmospheric environment might be different. From experience with processing large amounts of AIRS data it seems these cases are rare. When SO2 is present in a reference pixel it is expected that the R2 correlation will be lower than for another SO2-free pixel. However, it is possible to have an SO2 cloud so large that it covers the entire AIRS granule (in that case one might choose to use a second contiguous granule) and we caution that there are other scenarios one could imagine that might affect the correlations.

Figure 3.

(a–f) Spectral matching plots for measured absorbance spectra (dotted black lines) and synthetic absorbance spectra (red lines) for R2 correlations ranging from 0.46 to 0.96. The gaps in the measured spectra between 1348 and 1354 cm−1 are measurement gaps and are excluded in the correlation analyses. The locations of the target pixel and the reference pixel are given in each panel. Note that the retrieved SO2 column shown has been derived from least squares estimation and not by integrating the absorbance spectrum.

[16] The column SO2 is now determined for pixels exceeding a specified R2 from the absorbance spectrum using a linear least squares method. In this method a set of precomputed spectra are determined at discrete levels in the atmosphere at 2 km intervals starting at 6 km and ending at 20 km. These are linearly combined to produce a least-squares “best fit” between the measured and computed spectra. The computed spectra include the effects of a constant predefined water vapor distribution and the least-squares is improved by providing an estimate of the height of the cloud layer, which is assumed to contain only SO2 and enhanced water vapor (a constant amount for all retrievals). There is a strong dependence between absorbance and the height of the SO2 layer in the atmosphere, which is unknown. The height can be specified from a trajectory model run, or from some other independent source.

[17] The amount of SO2 retrieved using the 7.3 μm band is dependent on the location of the SO2 layer in the atmosphere. There is a temperature dependence of the line strengths for this band. There is also a strong vertical dependence because of the interfering effects of water vapor. We define the cutoff height for SO2 retrievals as the height in the atmosphere where the signal-to-noise ratio reaches unity. The NEΔT for AIRS channels between 3.7 and 13.5 μm is ∼0.2 K at 250 K, which gives NEΔI ≈ 0.1 mW/(m2 sr cm−1) for channels in the region 1295–1405 cm−1. The signal strength can be calculated from simulations,

equation image

where Iν,z is the spectral radiance from an atmosphere with a prescribed amount of SO2 placed at height z′ in the atmosphere, and Iν,0 is the spectral radiance from an unperturbed, background atmosphere. The signal to noise ratio (SNR) is,

equation image

We define the cutoff height zc = z′ when SNR = 1.0. Figure 4 shows the variation of SNR with height (z′) for a tropical atmosphere with 50 DU placed at different levels in the atmosphere from the surface up to about 5 km. In this case the SNR = 1 at z′ ≈3 km. It is apparent that for this case, below ∼2.5 km, the noise is twice the signal strength. For drier and wetter atmospheres, zc decreases or increases, respectively; likewise for emplacements of larger masses of SO2 the signal strength would be larger and the cutoff height would decrease. Under most atmospheric conditions absorption by water vapor across this band is significant below ∼3 km and we maintain that the AIRS instrument is effectively “blind” to SO2 emissions in the boundary layer, and often below ∼3 km (but note that this cutoff height depends on the SNR). This constraint suggests that the 7.3 μm AIRS channels behave like a filter to reveal mostly upper troposphere/lower stratosphere (UTLS) SO2, which is more likely to be climatically significant.

Figure 4.

Computed signal-to-noise (SNR) ratio as a function of height for a tropical atmosphere with ∼40 DU SO2 column amount placed in a single layer at various heights in the atmosphere up to ∼5 km.

[18] Another approach to assess the information content of the AIRS channels in the 1295–1405 cm−1 interval is to use weighting functions. The weighting function [e.g., Rodgers, 2000] is

equation image

which appears in the integrals of (3). Weighting functions for channels near 7.3 μm typically peak between 400 and 600 hPa depending on the water vapor, temperature and SO2 vertical profiles. This suggests that this waveband has greatest sensitivity to SO2 in the mid to upper troposphere (4–8 km). Given knowledge of these three profiles it is straightforward to calculate the Ws for AIRS channels and determine which atmospheric layers are contributing to the measured radiances. The value in this approach is that it permits an objective means for selecting specific (high information content) AIRS channels for retrieving SO2. Here we employ an ad hoc “two-step” method of first identifying SO2 pixels through spectral matching and then retrieving the column abundance through a least-squares procedure and off-line radiative transfer calculations, utilizing all AIRS channels between ∼1320 cm−1 to ∼1395 cm−1. We sacrifice any potential vertical SO2 profile information in pursuing this approach.

[19] Equation (5) in matrix notation may be written.

equation image

where x = {xoxN} is a vector of layer absorber amounts to be determined, K is an MxN matrix consisting of absorption cross sections at M wave numbers and N atmospheric layers, and y is a vector of measured absorbances at M wave numbers. The absorption cross sections at 8 levels, starting at 6 km and ending at 20 km in 2 km steps, are predetermined from a detailed radiative transfer program [Griffith, 1996]. The least-squares solution to (10) is [Rodgers, 2000],

equation image

[20] Equation (11) is underdetermined because the basis functions K are approximately linear functions of each other. The dominant processes affecting the shapes of the spectral lines within the band are pressure and Doppler (thermal) broadening. Temperature decreases through the UTLS (upper troposphere–lower stratosphere) in an almost linear manner, giving absorption cross sections in different layers which are very nearly linear functions of each other. To stabilize K, we found it necessary to include the effects of a second gas: water vapor. Thus the basis functions consist of absorption cross sections for a standard profile of water vapor with a single layer of enhanced SO2 and H2O at a prescribed level. The retrieval produces a layer abundance of SO2, which we treat as a column amount. The retrieval scheme produces layer amounts in eight layers, which are integrated to obtain a column amount. The information content in the layers is insufficient to expect an accurate vertical profile. Improvements to this scheme would include a better specification of the water vapor profile (potentially this could be determined from AIRS standard retrieval products) and simultaneous retrieval of SO2 and H2O. This would also help to identify enhancements in “in plume” water vapor associated with the volcanic cloud. In principle it is also possible to determine some vertical profile information on SO2 by judicious choice of “microwindows” within the ν3-band.

[21] The SO2 is retrieved on a pixel by pixel basis using (11) for only those pixels that exceed a specified value of R2. The units of SO2 amount are molecules cm−2, which we convert to milli atm cm or DU. The use of DU for SO2 column abundance is done for consistency with the OMI, TOMS and GOME retrievals. The total mass loading (in Tg) is evaluated from:

equation image

where i is pixel number, u absorber amount (in DU), β is the area of a pixel (in km2) evaluated assuming elliptical pixels on a spherical Earth, and θ is the AIRS scan angle subtended at the Earth's surface.

[22] Each SO2 retrieval is accompanied by an R2 correlation map. R2 ≈0.3 was found to delineate the boundary for SO2 retrievals of 6 DU, which is considered to be the lower limit of SO2 detection from AIRS. In this work we compute total SO2 mass loadings for pixels with R2 = 0.1 and R2 = 0.7, which provides a range of certainty and an error bound for the retrieved products. Finally, the height of the SO2 cloud is required as input to the retrieval. This information is derived using wind trajectories determined using the HYSPLIT model In the cases shown, there is no ambiguity in setting the height of the cloud from the trajectory forecast, but we acknowledge that this may not always be the case and it would be preferable to determine the height uniquely by some independent means.

5. Examples and Comparison

[23] The next subsections provide several examples of AIRS SO2 retrievals for various eruptions encompassing the years 2002–2006, a wide latitude range and covering heights in the atmosphere from a few kilometers above the surface up to 20 km, well into the stratosphere. These examples have been chosen to illustrate features of the AIRS retrievals and also to provide comparisons with independent data sources.

5.1. Anatahan, Northern Mariana Islands

[24] Anatahan (16.35 °N, 145.67 °E, 790 m) had its first recorded eruption on 10 May 2003, emitting an ash and SO2 plume which reached heights of up to 14 km or more. We use this eruption to illustrate some of the aspects of the retrieval scheme. AIRS observed the complete SO2 cloud on 10 May 2003 at 1641 UT (Figure 5a). Largest concentrations of 60 DU are observed south of the volcano, moving toward Saipan and Guam. The R2 correlation map (Figure 5b) shows excellent correlations (R2 ≥ 0.7) for large parts of the plume. The correlation map shows the regions of the image where pixels have been determined to contain SO2, on the basis of the criterion that the correlation between the synthetic absorbance spectrum and the measured absorbance spectrum is high.

Figure 5.

(a) AIRS SO2 column amounts for the Anatahan eruption of 10 May 2003. (b) Corresponding R2 correlation map for Figure 5a.

[25] TOMS observed the plume on 12 May and for about a month afterward and reported a total mass loading of ∼0.51 Tg, with ∼0.11 Tg emitted between 10 and 12 May [Wright et al., 2005]. No GOME data are available for this eruption, but the MODIS retrieval gives a total mass loading of 0.07 Tg. Table 1 provides a comparison of AIRS, TOMS and OMI retrievals for several recent volcanic eruptions, including this one and the Anatahan activity of April 2005. In each of the examples that follow, we will refer to Table 1.

Table 1. Total SO2 Masses Derived From AIRS Granules for Eruptions From Eight Different Volcanoes Over the Years 2002–2006a
VolcanoDateAIRSTOMS(T)/OMI(O)Altitude (Nearest km)
Time, UTMass, TgTime, UTMass, Tg
  • a

    Where available SO2 masses derived from TOMS or OMI are given. AIRS lower and upper bounds correspond to R2 correlations of 0.7 and 0.1, respectively.

Ruang25 Sep 20021705–17110.025–0.028  16
Ruang26 Sep 200205350.039–0.055∼03150.016(T)16
Reventador4 Nov 200206230.060–0.07116320.084(T)12
Anatahan10 May 200316410.025–0.034  14
Anatahan11 May 200302470.007–0.00900270.010(T)14
Anatahan12 May 200303290.007–0.00801150.099(T)14
Anatahan26 May 200315530.003–0.00401140.035(T)8
Montserrat13 Jul 200316530.071–0.075∼1515∼0.10(T)15
Montserrat14 Jul 200305050.043–0.051∼1515∼0.04(T)15
Montserrat15 Jul 200316410.012–0.037  15
Montserrat16 Jul 200304470.011–0.024  15
Manam28 Jan 200503350.069–0.1170409–05570.093(O)18
Manam29 Jan 200504410.068–0.1260314–06400.051(O)18
Manam30 Jan 200505230.009–0.0360357–07160.039(O)18
Anatahan6 Apr 200503350.045–0.0490212–05360.090(O)6
Anatahan6 Apr 20051547–15530.060–0.077  6
Anatahan7 Apr 20050241–04170.049–0.0710255–04420.085(O)6
Anatahan7 Apr 20051629–16350.022–0.038  6
Anatahan8 Apr 200503230.009–0.0160203–05250.046(O)6
Anatahan8 Apr 200515350.006–0.015  6
Montserrat21 May 200606350.175–0.2081740–19250.196(O)20
Montserrat25 May 20060923–20170.130–0.1551811–21450.164(O)20
Rabaul7 Oct 200603530.105–0.1350234–04120.125(O)17
Rabaul8 Aug 20061347–15170.092–0.1090138–04580.285(O)≤8
Nyamuragira28 Nov 200611470.059–0.0570709–13490.216(O)≥8
Nyamuragira29 Nov 200612230.177–0.1910746–14320.468(O)≤8
Nyamuragira30 Nov 200600410.156–0.1700659–13360.675(O)≤8
Nyamuragira1 Dec 200612110.308–0.3330734–14190.690(O)≤8
Nyamuragira2 Dec 200600290.219–0.2410649–14530.611(O)≤8

[26] The discrepancy between the AIRS/MODIS and TOMS/OMI retrievals may be due in part to the fact that AIRS and MODIS only observe the SO2 residing above ∼3 km or so and hence should always infer less SO2 than TOMS/OMI, whenever there is SO2 below ∼3 km, and when equal spatial coverage is achieved and there is no time difference between the observations. Very often the AIRS instrument does not provide complete horizontal spatial coverage of SO2 clouds, making detailed comparisons with TOMS, OMI and GOME difficult. The cutoff height of AIRS SO2 retrievals restricts the number of comparisons that can be made with independent data sources, because many eruptions are small and the gases often do not reach sufficient height to be detectable by AIRS. However, we have noted that even when the gases lie below ∼3 km, if there is a large SO2 emission (e.g., Nyamuragira, see later) the AIRS SNR can be large enough to render the SO2 signal detectable.

5.2. Etna, Sicily

[27] Mt Etna (37.734 °N, 15.004 °E, 3350 m) is a stratovolcano on the island of Sicily and has a very long and well-documented eruptive history. It has been in near continuous eruption since mid-1995. There was a flank eruption in July–August 2001 and during the period of operation of AIRS a series of midsized ash and gas eruptions during 2002, the most vigorous of these occurring in late October, early November 2003. Carn et al. [2005] used this eruption to illustrate their AIRS SO2 retrievals. We have made AIRS retrievals for the same case illustrated by Carn et al. [2005, Figure 3] for the Etna plume of 30 October 2002. A height of 6 km was assumed, on the basis of the work of Tiesi et al. [2006], which also suggests a plume constrained at heights between 5600 and 6600 m. Carn et al. [2005] assumed a 2 km thick plume centered at 6 km. The AIRS retrieval is shown in Figure 6. The total SO2 mass retrieved for this image is 0.022 Tg (R2 = 0.3), which can be compared with Carn et al.'s [2005] estimate of 0.0245 Tg for the same image. Pixel column amounts are very similar and the shape of the plumes are also very similar. Some plausible reasons for the ∼10% difference between the two retrievals are possible interference effects due to volcanic ash and our conservative approach adopted in the spectral matching. For R2 = 0.1, the total mass retrieved is 0.0291 Tg. Carn et al. [2005] also retrieved significant quantities of ash in this plume. It is not known how much interference ash particles cause on absorption in the 1295–1405 cm−1 interval, but ash does have significant effects in the 8–12 μm (800–1250 cm1) region [e.g., Prata, 1989]. Carn et al. [2005] used 5 AIRS channels across the ν3 SO2 band, while we use ∼140 channels, some of which may be influenced by ash particle effects. In any case, the degree of agreement between the two techniques is encouraging.

Figure 6.

AIRS retrieval for the Etna eruption on 30 October 2002. Total mass retrieved in this image is 0.022 Tg (R2 = 0.3) and this compares favorably with Carn et al.'s [2005] estimate for the same image of 0.0245 Tg.

5.3. Ruang, Sangihe Islands, Indonesia

[28] Ruang (2.28°N, 125.43°E, 725 m) off the northern tip of the island of Sulawesi erupted between 0340 UT and 0400 UT on 25 September 2002, emitting a large cloud of ash and SO2 [Smithsonian Institution, 2002]. The ash cloud was initially reported to have reached heights of 5 km, but later analysis of satellite imagery and wind trajectories suggests that portions of the cloud may have been much higher. The first Aqua overpass over the region to capture the eruption was at 1705 UT, about 13 hours after the initial eruption. Figure 7 shows the AIRS retrieval from two consecutive AIRS granules. Largest SO2 amounts are about 40 DU situated west of the volcano. There is an indication that the main SO2 cloud has traveled west and then southward. There is no indication of an SO2 cloud east of Ruang in this image. GOME and TOMS both observed SO2 from Ruang on the following day, TOMS around midday on 26 September and GOME at 0230 UT, also on 26 September. Aqua passed over the region at around 0530 UT. Figure 8 shows the AIRS (Figure 8a), MODIS (Figure 8b), GOME (Figure 8c) and TOMS (Figure 8d) SO2 retrievals on 26 September. The total SO2 mass loadings estimated in these images are: 0.039 Tg (R2 = 0.7)–0.055 Tg (R2 = 0.1) for AIRS, 0.079 Tg for MODIS, 0.02 Tg for GOME and 0.016 Tg for TOMS. Neither GOME nor TOMS have captured the maximum SO2 concentrations near to 123°E longitude. The larger masses retrieved by MODIS are presumably due to the better coverage by MODIS of the eastern arm of the SO2 cloud. Analysis of AIRS and MODIS images suggests that the SO2 plume had two components: a lower portion moving west and then southward, and a higher, faster moving cloud that traveled eastward to Irian Jaya and New Guinea. Figure 9 shows HYSPLIT wind trajectories for three sources starting at Ruang at 0400 UT on 25 September with initial heights of 5 km, 15 km and 20 km. The trajectory at 15 km corresponds well to the movement of the SO2 cloud observed by AIRS and MODIS. The observations of SO2 moving eastward suggests that some of the gas rose to heights of ∼20 km or higher and hence reached the stratosphere. The incomplete coverage of each of the satellite instruments suggests that none are able to give an accurate measure of the total mass of SO2 ejected from Ruang.

Figure 7.

AIRS SO2 column amounts for the Ruang eruption of 25 September 2002, 1705–1711 UT.

Figure 8.

(a) AIRS SO2 column amounts for the Ruang eruption of 26 September 2002, 1329 LT (local time = UT + 8 hours). (b) Corresponding MODIS retrieval at 1329 LT. (c) Corresponding GOME retrieval at 1030 LT. (d) Corresponding TOMS retrieval at local noon.

Figure 9.

HYSPLIT trajectories for three sources on Ruang starting at heights of 5 km, 15 km and 20 km, at 0400 UT on 25 September 2002.

5.4. Soufrière Hills, Montserrat, West Indies

[29] Eruptive activity at Soufrière hills volcano (16.72°N, 62.18°W, 915 m) on Montserrat increased during 12–13 July 2003, resulting in explosive events, the largest occurring around 2000 UT on 12 July. Sulphur dioxide emissions were measured by the Montserrat Volcanological Observatory and reported by the Smithsonian Institution [2003] as reaching highest levels of 0.172 Tg d−1. A TOMS overpass on 13 July recorded a long plume of SO2 extending northeast from the volcano with a total mass loading of ∼0.1 Tg (S. Carn, private communication, 2005). Six AIRS overpasses detected significant SO2 amounts between 13 and 15 July. Figure 10a shows the AIRS retrieval on 13 July, corresponding to the TOMS retrieval (Figure 10b) of the same day. Table 1 lists the SO2 mass loadings determined from the AIRS data. There is reasonable agreement (within 20–30%) between the AIRS retrievals and TOMS on 13 and 14 July.

Figure 10.

(a) AIRS SO2 column amounts for the Montserrat eruption of 13 July 2003. (b) Corresponding TOMS retrieval at local noon.

5.5. Tavurvur, New Britain, PNG

[30] Tavurvur (4.27°S, 152.20°E, 688 m) is a stratovolcano cone on the rim of the Rabaul caldera, near the town of Rabaul, New Britain and has been restless since September 1994. On 7 October 2006 a large ash and gas eruption occurred reaching heights of at least 16 km and possibly penetrating the tropopause. AIRS captured the eruption at 0353 UT on 7 October providing almost complete coverage of the SO2 cloud. OMI imaged the cloud at the same time (between 0234 and 0412 UT). The excellent coverage by these two instruments and the likely upper tropospheric penetration of the gas cloud make this eruption ideal for intercomparisons of SO2 retrievals. The AIRS and OMI (courtesy Simon Carn) images are shown in Figure 11 on the same geographic projection and scale. The similarity is striking; although there are differences in the detail, for example the region of maximum SO2 amount has a different shape in both images. Table 1 lists the two best coincidences and column SO2 amounts retrieved by AIRS and OMI. The coincidence on 7 October shows that (unusually) AIRS registered higher SO2 amounts than OMI, but the difference is less than 10%. On 8 October OMI is registering nearly 3 times as much SO2 as AIRS. The SO2 cloud had formed two branches: one high-level cloud (>15 km) to the north of New Britain and the other a low-level cloud (<10 km) to the south. AIRS detected much less SO2 in the southerly branch compared to OMI. Additionally, the AIRS swath did not overlap with the westernmost branch of the northern cloud and consequently the AIRS estimated SO2 is an underestimate for this cloud.

Figure 11.

(a) AIRS SO2 column amounts for the Tavurvur eruption of 7 October 2006. (b) Corresponding OMI retrieval at 0324–0412 UT (data courtesy Simon Carn).

5.6. Nyamuragira, DR Congo

[31] One of the most prodigious emitters of volcanic SO2, Nyamuragira (1.41°S, 29.20°E, 3000 m ASL) has a long and well-documented TOMS (and now OMI) record of satellite measurements. Most of the SO2 emissions from this large basaltic volcano are due to passive degassing, interspersed with periods of mild eruptive activity. Between 28 November 2006 and 4 December 2006, Nyamuragira produced a regional-scale SO2 cloud that initially traveled westward before recurving north and then east toward the Red sea. OMI and AIRS captured the SO2 cloud during the whole week of activity. This is an interesting volcano to observe as most of the emissions reside in the lower troposphere, but the large quantities of SO2 emitted make detection by many different satellite sensors feasible. Figure 12 shows the AIRS SO2 retrievals for the period 28 November to 3 December 2006, averaged and rebinned to show the scale and extent of Nyamuragira's SO2 cloud. The total SO2 mass emission calculated from this map is 0.444 Tg. A comparable OMI image showing this cloud over the period 28 November to 4 December is available from Simon Carn Detailed comparisons with OMI retrievals are listed in Table 1. In all cases the AIRS retrievals are lower than the OMI values; in some cases by a factor of 2. Once again, we suggest the cause of this is due to AIRS inability to detect SO2 in the layers closest to the surface. Radiosonde profiles at two locations (see map) in DR Congo (Ouessa) and Cameroon (Douala) for 28 November 2006 were used to calculate the cutoff height for the AIRS band between 1295 and 1405 cm−1. The SNR as a function of height using single-layer emplacement of SO2 of 60–70 DU (column abundance) is shown in Figure 13 for these profiles. The cutoff height appears to be ∼2 km, which is lower than for the tropical atmosphere simulations shown earlier. HYSPLIT runs suggest that trajectories starting at Nyamuragira above 7 km go westward and then recurve north, while those below 4 km travel westward and do not recurve. We conclude that the SO2 detected by AIRS for this eruption must have reached at least 4 km and likely higher, but because the AIRS values are much smaller than the OMI values there must also be amounts of SO2 below 2 km that are not detectable by AIRS.

Figure 12.

AIRS SO2 column rebinned and combined to show the regional-scale SO2 plume from Nyamuragira volcano during 28 November to 3 December 2006. The total SO2 mass estimated from this image is 0.444 Tg. Locations of two radiosonde stations are also shown on the map.

Figure 13.

SNR ratio as a function of height computed using the radiosonde data from Ouesso (DRC) and Douala (Cameroon) (see map in Figure 13 for locations). The profiles were obtained on 28 November 2006 and show that in this case, AIRS could detect SO2 down to ∼2 km above the surface.

6. Error Budget

[32] A comprehensive error budget has not been undertaken for the retrieval scheme. The main sources of error are (1) radiometric accuracy of the AIRS channels; (2) estimation of the background absorbance spectrum; (3) interference from clouds and other gases, principally water vapor; (4) height errors; and (5) errors in the spectroscopic parameters for SO2. Errors due to the wavelength calibration of AIRS, navigational accuracy of the pixels, radiative transfer modeling error and numerical errors from the least-squares procedure and other computational procedures are much smaller than the main sources listed above and are neglected. The main error sources are estimated in terms of a noise equivalent radiance (mW/(m2 sr cm−1)) and then converted to a noise equivalent absorbance, and finally to an equivalent absorber amount error in DU. Table 2 provides a guide to the magnitude of these error sources.

Table 2. Error Budget for the Proposed Retrieval Scheme for a Single Pixel Column Abundance Retrieval
Error SourceNEΔI, mW/(m2 sr cm−1)NEΔA (No Units)δu, D.U.
  • a

    Determined by varying the reference pixel by ±3 pixels and ±3 lines from its optimal value and recomputing the absorber amounts.

  • b

    Determined by computing absorber amount with and without water vapor. The error was then doubled to account for the effect of cloud absorption.

  • c

    Assumes a height error of ±4 km, which is equivalent to ∼15% error in the absorption cross section.

  • d

    Assumes a 10% error in the absorption cross section.

Background referencea±1±0.100∼4
Interference: water vapor and cloudsb±1±0.100∼4
Total (rms)  ∼6

7. Conclusions

[33] A retrieval scheme has been proposed for deriving the column SO2 abundance from the 1362 cm−1 (7.34 μm) antisymmetric stretch absorption v3-band using AIRS spectra. The detailed AIRS measurements permit the effects of water vapor to be accounted for and the spectral resolution allow unambiguous identification of SO2. The scheme has an accuracy of about 6 D.U. (≈1.6 × 1017 molecules cm−2), mostly limited by errors in accounting for the background atmosphere. AIRS radiometric accuracy and lack of knowledge of SO2 spectroscopic parameters contribute smaller error to the retrieval.

[34] Several examples demonstrate that AIRS can detect small eruptions (total SO2 mass less than 0.1 Tg) that penetrate into the UTLS. The inability of instruments relying on absorption in the 7.34 μm SO2 band to measure SO2 below ∼3 km is a fundamental restriction. One consequence of this is that AIRS (or HIRS and MODIS) measurements provide a natural filter for identifying SO2 in the UTLS and that may be climatically significant. With its high spectral resolution, AIRS will be able to “see” deeper into the atmosphere than either HIRS or MODIS. Limited comparisons between AIRS, MODIS, GOME and TOMS measurements suggest that none of the instruments can provide a reliable estimate of the total mass loading of SO2 because of the lack of complete spatial coverage. More detailed comparisons between AIRS and OMI measurements indicate that OMI detects more SO2 whenever the SO2 cloud remains below ∼8 km. Figure 14 summarizes the measurements listed in Table 1 showing that for eruptions above ∼8 km there is a high degree of agreement between AIRS and OMI, while below ∼8 km AIRS values must be increased by a factor of 2–3 to obtain agreement. We stress here that these comparisons are not intended to provide a validation of either instrument, rather we show them in order to demonstrate that AIRS can quantitatively detect SO2, albeit with certain caveats. A validation data set is being assembled under a separate research activity in collaboration with the University of Maryland (Simon Carn).

Figure 14.

AIRS column abundance versus OMI/TOMS column abundance for all coincidences listed in Table 1. The spread in AIRS estimates corresponds to masses computed with R2 = 0.1 and R2 = 0.7; those for OMI/TOMS have been set at ±5%. SO2 clouds above 8 km are indicated by circles, while those below 8 km are shown as triangles. The correlation between AIRS and OMI/TOMS estimates is much better for high clouds than low clouds. The lack of correlation for some points is due to the incomplete horizontal coverage of AIRS, which often leads to an underestimate by AIRS. For low clouds, AIRS values must be increased by a factor of 2–3 to get better agreement with OMI/TOMS.

[35] Inspection of AIRS spectra also shows that there may be vertical SO2 height information. SO2 absorption has also been identified in the AIRS spectra at 2460–2525 cm−1 and 1070–1125 cm−1. The spectral matching technique may also be applied to this spectral region as Figure 15 shows for an SO2 cloud detected on 25 June 2003. The SO2 was released from a fire at the Al-Mishraq sulphur plant in Iraq [Carn et al., 2004]. Improvements to the retrieval scheme will include better specification of the water vapor profile, simultaneous gas retrieval, use of the 4.0 μm channels and investigation of retrieving the vertical profile of SO2. Radiance in the 4.0 μm window originates from much lower levels in the atmosphere compared to radiance at 7.3 μm, and a combination of AIRS measurements at 4 μm, 7.3 μm and 8.6 μm (lower wavelength part only) could provide sufficient information for a vertical SO2 profile retrieval from the surface to the stratosphere.

Figure 15.

Spectral matching plot for the 4 μm band of SO2. The red line shows a synthetic absorbance spectrum and the black line shows AIRS measurements. The data are for an AIRS image of the Al-Mishraq (Iraq) SO2 plume on 25 June 2003.


[36] We especially thank Simon Carn and the OMI science team for generously sharing OMI retrievals and Simon Carn for access to the TOMS retrievals. Andreas Richter is thanked for the GOME SO2 retrievals used in this paper. The authors also gratefully acknowledge the NOAA Air Resources Laboratory (ARL) for the provision of the HYSPLIT transport and dispersion READY Web site used in this publication. The anonymous reviewers are thanked for their helpful comments, and in particular one reviewer who made several suggestions with regard to improvements in discussion of the retrieval technique and the R2 correlation.