AMTAtmospheric Measurement TechniquesAMTAtmos. Meas. Tech.1867-8548Copernicus GmbHGöttingen, Germany10.5194/amt-8-195-2015Observations of volcanic SO2 from MLS on AuraPumphreyH. C.hugh.pumphrey@ed.ac.ukhttps://orcid.org/0000-0003-0747-1457ReadW. G.LiveseyN. J.YangK.School of GeoSciences, The University of Edinburgh, Edinburgh, UKJet Propulsion Laboratory, California Institute of Technology, Pasadena,
CA, USADepartment of Atmospheric and Oceanic Science, University of Maryland,
College Park, MD, USAH. C. Pumphrey (hugh.pumphrey@ed.ac.uk)12January20158119520929May201431July201417November20144December2014This work is licensed under a Creative Commons Attribution 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/This article is available from https://www.atmos-meas-tech.net/8/195/2015/amt-8-195-2015.htmlThe full text article is available as a PDF file from https://www.atmos-meas-tech.net/8/195/2015/amt-8-195-2015.pdf
Sulfur dioxide (SO2) is an important atmospheric constituent,
particularly in the aftermath of volcanic eruptions. These events
can inject large amounts of SO2 into the lower stratosphere,
where it is oxidised to form sulfate aerosols; these in turn have a
significant effect on the climate. The MLS instrument on the Aura
satellite has observed the SO2 mixing ratio in the upper
troposphere and lower stratosphere from August 2004 to the present,
during which time a number of volcanic eruptions have significantly
affected those regions of the atmosphere. We describe the MLS
SO2 data and how various volcanic events appear in the data. As
the MLS SO2 data are currently not validated we take some initial
steps towards their validation. First we establish the level of
internal consistency between the three spectral regions in which MLS
is sensitive to SO2. We compare SO2 column values calculated
from MLS data to total column values reported by the OMI
instrument. The agreement is good (within about 1 DU) in cases
where the SO2 is clearly at altitudes above 147 hPa.
Introduction
Sulfur dioxide (SO2) is an important minor constituent of the atmosphere.
Natural tropospheric sources include volcanoes, while anthropogenic sources
include combustion of fossil fuels and smelting of sulfur-containing metal
ores. Tropospheric emission of SO2 has a variety of detrimental effects on
air quality and ecosystems; in particular it can be a major contributor to
acid rain . The high solubility of SO2 in water which
leads to acid rain means that very little of the SO2 emitted at the
surface reaches the stratosphere. Sulfur dioxide in the stratosphere may be
produced in situ by chemical reactions involving other, less soluble,
sulfur-containing molecules – mostly carbonyl sulfide (OCS)
. Sulfur dioxide may also reach the upper troposphere or
stratosphere if it is emitted directly from the Earth's surface by a process
which is sufficiently energetic to loft it to that altitude. A volcanic
eruption is the only such process which has been observed to lead to enhanced
SO2 in the lower stratosphere.
Once in the stratosphere, SO2 becomes an important component of the
climate system . It is oxidised on a timescale of about a
month, becoming aerosol particles which have a lifetime in the stratosphere
of over a year. These particles alter the albedo of the Earth, reflecting a
fraction of sunlight back into space and thereby reducing the Earth's
temperature. Based on six of the largest eruptions of between 1875 and 1991,
show that the cooling can be on the order of 0.1 to
0.2 ∘C.
A variety of techniques exist for the remote sensing of atmospheric SO2
from satellites. Nadir sounding provides good horizontal resolution but
little or no vertical resolution. Both thermal emission in the infrared (IR)
and backscattered sunlight in the ultraviolet (UV)
can be used. Note that SO2 layer heights may also be
retrieved from hyper-spectral UV observations when the columns are
sufficiently large . Limb-sounding instruments provide
vertically resolved profiles but with limited horizontal resolution.
Thermally emitted radiation in both the infrared and
microwave regions can be used for limb sounding.
The Microwave Limb Sounder, or MLS , is one of the four
instruments on NASA's Aura satellite . MLS measures the
concentrations of a suite of 14 chemical species in the upper troposphere and
middle atmosphere. Sulfur dioxide (SO2) is one of the species measured.
The sensitivity of the measurement is not sufficient to detect the background
levels of SO2, but the enhanced levels present following a
sufficiently large volcanic eruption are detected; it is these observations
which we report in this paper. We describe the instrument and the measurement
process in more detail in Sect. , give an overview of the
volcanic events observed in Sect. , and examine three of the
larger events in more detail in Sect. . In
Sect. we estimate the total mass of SO2 injected into
the stratosphere by the larger volcanic eruptions of the last decade. In
Sect. we make an initial attempt to validate the data.
DataThe MLS instrument
Aura was launched in July 2004, and the MLS instrument
has operated with little interruption from August 2004 to
date. The satellite orbits at an altitude of 705 km, performing
approximately 14.5 orbits per day. The MLS instrument consists of a 1.6 m
parabolic dish antenna feeding heterodyne radiometers operating at 118, 190,
240 and 640 GHz. A separate small antenna feeds another radiometer operating
at 2.5 THz. The output of the radiometers is analysed by banks of filters.
The antenna looks forward from the Aura platform, in the plane of the orbit,
and is scanned across the Earth's limb 240 times per orbit. As the orbit is
inclined at 98∘ to the Equator, the instrument observes a latitude
range from 82∘ S to 82∘ N every day. The observations are of
thermal emission from the atmosphere and can therefore be made day and night.
The orbit is sun-synchronous, so the observations are always made at the same
two local times for a given latitude. The radiances reported by the filter
banks are used as input to a software package which
estimates profiles of temperature, and of the mixing ratios of the target
chemical species. Most MLS estimated profiles, including SO2, are reported
on pressure levels spaced at six levels per pressure decade, a spacing of
approximately 2.7 km in altitude. The estimated profiles are spaced
1.5∘ (167 km) apart along the orbit track. The software produces an
estimate of the precision of every quantity estimated. All mixing ratios in
this paper are produced by version 2 (V2) of this software package. The
precisions and accuracies of these data are summarised in a data quality
document .
The MLS SO2 measurement
The SO2 molecule, like the H2O and O3 molecules, is a
nonlinear triatomic molecule. A combination of this shape, a
moderately large dipole moment (similar to that of water) and a large
moment of inertia leads to a spectrum with a large number of strong
lines and an even larger number of weak ones. There are
reasonably strong SO2 lines present in the passband of all MLS
radiometers. The software makes separate attempts to estimate the
SO2 mixing ratio from the 190, 240 and 640 GHz radiometers, but only
the 240 GHz product is of sufficient quality for general use.
Three versions of the MLS data have been released to the public: V1.5, V2 and
V3. Although V3 is an improvement on V2 in most respects, CO and SO2 data
in the upper troposphere are distinguished less well from clouds in V3 than
in V2. For this reason we use the V2 data in this paper. The pressure levels
on which SO2 is reported and where significant amounts have been observed
are 316, 215, 147, 100, 68 and 46 hPa. At altitudes below the 316 hPa level
no retrieval is attempted. Retrieval is attempted between 316 hPa and
1 hPa; the sensitivity and vertical resolution of the retrieval become
rapidly poorer at altitudes above 10 hPa. Between 215 and 10 hPa the
vertical resolution is close to the spacing between the pressure levels,
about 3 km, and the estimated precision is approximately 3 ppbv. Although
retrieval is attempted at 315 hPa the data at this level are not recommended
for general use. We present them in this paper but are cautious about drawing
any conclusions from them. Further details on the data quality are given in
.
Figure shows data from a single pressure level, for all
profiles on two different days. On 7 August 2008 the SO2 mixing ratio, as observed by
MLS, was very small compared to the measurement noise. The additional scatter
seen in the tropics is caused by high clouds interfering with the retrieval.
The standard deviation of the data is generally 2–4 times larger than the
estimated precision, suggesting that there are contributions to the random
error which are not accounted for in the retrieval software. On
11 August 2008 there are a collection of points with large and positive
values; as we show in more detail below these points represent SO2 emitted
from the eruption of Kasatochi on 8 August 2008. Those points which are more
than 7.7 standard deviations above the mean are marked in red. The mean and
standard deviation used are for the unmarked points only; an iterative
procedure was used to separate the marked points. The choice of 7.7 standard
deviations as a cut-off is arbitrary and was chosen by trial and error in
order to have only a small number of false positives while still identifying
the major volcanic events. Figure suggests that there are a
number of points which are affected by volcanic SO2 but which are below
this threshold. The 7.7 standard deviation threshold is typically in the
50–100 ppbv range; this is about 100 times larger than the background
SO2 mixing ratio as measured by other instruments (see e.g.
).
Scatter plot of MLS SO2 mixing ratio
against latitude for 7 and 11 August 2008 at 215 hPa. All profiles
(and hence the entire 82∘ S to 82∘ N latitude range) are
shown. An eruption of the volcano Kasatochi in the Aleutian Islands
occurred on 8 August. Points that are more than 7.7 standard
deviations above the zonal mean, and which therefore lie within the
SO2 plume caused by the eruption, are marked with red circles.
Figure also suggests that the daily zonal mean MLS SO2
mixing ratio is not quite zero. Inspection of a time series (not shown) shows
that it varies somewhat throughout the year. At most altitudes the annual
cycle is largest at the poles, where it has an amplitude of about 2 ppbv.
The data therefore have a bias which is dependent upon both time and
latitude. We do not consider this feature of the data to be due to SO2 in
the atmosphere for several reasons. Firstly, recent observations of the
background SO2 amounts by show that the true background
levels are much smaller than this. Secondly, the seasonal cycles in MLS
SO2 show many features in common with the seasonal cycles in MLS
measurements of ozone and nitric acid; these species have strong spectral
lines in the passband of the 240 GHz radiometer. We conclude that the
seasonal cycles seen in the zonal mean SO2 data are essentially leakage of
information from the O3 and HNO3 measurements. While the MLS data
appear promising for the study of enhanced levels of SO2 caused by
volcanoes, it is not currently possible to average them down in an attempt to
study seasonal variability of the non-volcanic background or other features
with amplitudes below about 2–3 ppbv.
Profiles of SO2 mixing ratio for a day
shortly after the eruption of Soufrière hills, Montserrat, in
2006, and a day shortly after the eruption of Kasatochi in 2008. The
Montserrat profiles are from 0 to 20∘ N; the
Kasatochi profiles are from 40 to 60∘ N. Profiles
with unusual amounts of SO2 as defined in the text are shown in
black, the other profiles in grey. The mean of the grey profiles is
shown in red; the dashed blue line is 7.7 standard deviations above
this mean.
Number of observations of SO2 per day for the entire mission. For
the purposes of this figure an “observation” is a point in the MLS profile
which is more than 7.7 standard deviations above the usual zonal mean value
for that pressure and latitude.
In Fig. we show some profiles for 2 days for latitude bands
affected by volcanic eruptions. Points with mixing ratios greater than
7.7 standard deviations above the mean are identified as for
Fig. , and profiles containing such a point are shown in
black. The two events shown differ from each other in the vertical
distribution of SO2: in one event the SO2 is all at the 68 hPa level,
and in the other it is distributed between 68 and 316 hPa with the largest
amounts at 215 hPa.
Average SO2 mixing ratio for
“observations” of SO2 as defined in
Fig. . Where there are insufficient observations,
the three highest mixing ratios observed that day are used
instead. This choice means that there is little artificially induced
difference in the plotted data between eruption and non-eruption
periods. The data plotted when there is no volcanic SO2 observed
are not typical values or daily average values either for the MLS
data or for the real background SO2.
Table summarising MLS observations of
SO2 during various volcanic eruptions. The eruptions are sorted
roughly in order of impact. For the purposes of this table an
“observation” is a point in the MLS profile which is more than
7.7 standard deviations above the usual zonal mean value for that
pressure and latitude.
LocationDate of firstDaysTotal no. ofHighest volume mixing ratioVolcanolong/latobservationobservedobservations(in ppmv)Sarychev, Kuril Islands153.2∘ W, 48.1∘ N14 Jun 2009314550.55 at 100 hPaKasatochi, Aleutian Islands175.5∘ W, 52.2∘ N8 Aug 2008232680.46 at 215 hPaSoufrière Hills, Montserrat62.2∘ W, 16.7∘ N23 May 200617390.18 at 68 hPaNabro, Eritrea41.7∘ E, 13.4∘ N14 Jun 20119730.29 at 100 hPaKelut, Java112.3∘ E,7.93∘ S14 Feb 20147480.4 at 68 hPaGrímsvötn, Iceland17.3∘ W, 64.4∘ N22 May 20118300.5 at 215 hPaRedoubt, Alaska152.7∘ W, 60.5∘ N23 Mar 200913250.2 at 147 hPaOkmok, Aleutian Islands168.1∘ W, 53.4∘ N13 Jul 200810160.34 at 147 hPaManam, Papua New Guinea145.0∘ E, 4.1∘ S28 Jan 20054220.28 at 68 hPaRabaul (Tavurvur), New Britain152.2∘ E, 4.3∘ S8 Oct 20064180.16 at 100 hPaNyamuragira, Dem. Rep. Congo29.2∘ E, 1.4∘ S28 Nov 2006770.15 at 147 hPaManam, Papua New Guinea145.0∘ E, 4.1∘ S24 Nov 2004670.19 at 100 hPaPuyehue-Cordón Caulle, Chile72.1∘ W, 40.6∘ S5 Jun 2011360.15 at 215 hPaDalaffilla (a.k.a. Gabuli), Ethiopia40.55∘ E, 13.8∘ N5 Nov 2008440.1 at 147 hPaSoufrière Hills, Montserrat62.2∘ W, 16.7∘ N13 Feb 2010230.087 at 68 hPaMerapi, Indonesia110.4∘ E, 7.5∘ S6 Nov 2010470.18 at 147 hPaPacaya, Guatemala90.6∘ W, 14.4∘ N27 May 2010240.086 at 215 hPaPiton de la Fournaise, Réunion55.7∘ E, 21.2∘ S7 Apr 2007120.13 at 215 hPaPaluweh, Indonesia121.7∘ E, 8.3∘ S4 Feb 2013120.11 at 100 hPaAnatahan, Mariana Islands145.7E∘ N, 16.4∘ N7 Apr 2005110.12 at 100 hPaSierra Negra, Galápagos Islands91.2∘ W, 0.8∘ S25 Oct 2005110.048 at 147 hPaJabal al-Tair, Yemen41.8∘ E, 15.6∘ N2 Oct 2007110.06 at 100 hPa
Number of observations of volcanic
SO2 in the days following the eruption of Sarychev. An
observation is defined in the same way as in Fig.
and Table . The eruption is marked as occurring on
day 163 (12 June 2009).
Event detection
We apply the detection procedure used for Figs. and
to the entire mission; the result is shown in
Fig. . The times of volcanic eruptions that are
detected by MLS are marked on the figure. The larger events are very
obvious. Some of the smaller ones are not obvious; detection was
confirmed by a combination of independent reports of an eruption and
co-located observations of enhanced SO2 by the OMI instrument (also
on Aura). The events marked on Fig. are summarised in
Table . Much of the information about names and
locations of volcanoes, and times of eruptions, was provided by the
Smithsonian Global Volcanism Program
(http://www.volcano.si.edu). Figure shows the
typical mixing ratios observed in the volcanic plumes; where there is
no plume the maximum value observed that day is
shown. Figures and show a few
spikes in 2011 and 2013 which on closer inspection appear to be faults
or glitches in the data and do not appear to be associated with any
volcanic activity.
Major events in more detail
In this section we examine in more detail several of the volcanic
eruptions shown in Fig. and Table .
Sarychev, June 2009
The Sarychev Peak volcano is located in the Russian Kuril island chain,
between Japan and the Kamchatka Peninsula. A major eruption of the volcano
occurred between 11 and 16 June 2009; the eruption is described by
and a detailed chronology of the explosion is given by
. The most energetic phases of the eruption were between
12 June and 16 June. We show the time series of MLS observations of elevated
SO2 in Fig. and a map of the locations of these
observations in Fig. .
Figure is produced in the same manner as
Fig. Note that the greatest number of observations occurs at
the 147 hPa level, but that the first observations (days 164 and 165; 13 and
14 June) are at 215 hPa. Figure suggests that most of
the SO2 travelled eastwards away from the volcano, curving northwards over
Canada and Alaska and dividing into two parts: one of which travelled westwards to disperse over
eastern Russia, the other of which travelled eastwards across
northern Canada and Greenland. A portion of the SO2
travelled westwards away from the volcano; closer inspection suggests that
this is the SO2 that was emitted towards the end of the eruption.
Locations where unusual levels of SO2 were recorded in the days
following the eruption of Sarychev which began on 11 June 2009. Colours
represent time in days; day 1 is 10 June 2009. A different symbol is used for
each of the MLS pressure levels.
Kasatochi, August 2008
The Kasatochi volcano in the Aleutian island chain erupted rather
unexpectedly on 7–8 August 2008. The eruption is described by
and the evolution of the SO2 plume as observed by OMI
is described by .
We show the time series of MLS observations of elevated SO2 in
Fig. and a map of the locations of these observations in
Fig. . The SO2 plume appears to be at a slightly lower
altitude than that from Sarychev; there are no detections at 46 hPa and the
largest number occurs at 215 hPa rather than 146 hPa. All of the
observations are to the east of the volcano; the upper level winds were
presumably westerly at all altitudes at which MLS observed any volcanic
SO2. The plume experiences considerable wind shear once over North
America; the SO2 at 215 hPa travelling rapidly eastwards across the North
Atlantic while that at 100 and 68 hPa remains over North America. The
observations persist for a longer time at these three levels than at the
intervening 147 hPa level.
As Fig. but for the eruption of Kasatochi in
2008 which began on 7 August 2008 (day 220).
Montserrat, May 2006
The Soufrière Hills volcano on the island of Montserrat in the West Indies
underwent a long phase of eruptive activity between 1995 and 2010. The
eruption was characterised by the repeated growth and collapse of lava domes
(see , , and references therein). Most
of this activity had little impact on regions of the atmosphere observable
by MLS; the main exception was the dome collapse on 20 May 2006. The collapse
is described by and the SO2 release by
and by .
We show the time series of MLS observations of elevated SO2 in
Fig. and a map of the locations of these observations in
Fig. . Note that the SO2 is confined to a narrow layer,
affecting only the 68 hPa level in the MLS data. This is in agreement with
the modelling results in , which show the SO2 forming a
layer centred at a height of 20 km and full width at half maximum of 2 km. It is some
days after the eruption before MLS observes any volcanic SO2. Inspection
of data from the OMI instrument, which has a higher horizontal resolution,
shows that the plume was of a small size and fell between the MLS orbits
between 20 May and 22 May.
As Fig. but for
the eruption of Kasatochi which began on 7 August 2008. Colours
represent time in days; day 1 is 5 August 2008.
Total SO2 burden
We calculated the total mass of SO2 in a suitable latitude region for each
major eruption. To do this, we first calculated daily zonal means of mixing
ratio. These were then integrated vertically as in Sect. to
give zonal mean column amounts. Finally, the column amounts are weighted by
area and summed for a suitable range of latitudes. Both the latitude range
and the highest pressure used in the vertical integration are chosen
separately for each eruption in order to encompass all of the profiles and
levels showing enhanced SO2 from that eruption.
The results are in most cases dominated by the spurious seasonal cycle
described in Sect. . As this varies smoothly in time we can
remove it by fitting annual and semi-annual sinusoidal cycles to the data for
the time unaffected by the eruption; a typical example is shown in
Fig. .
The excess mass, Mt, above the spurious background can be adequately
fitted as a function of time, t, by a decaying exponential:
Mt=M0exp-(t-t0)τ
or, equivalently
logeMt=logeM0-(t-t0)τ,
where M0 is the total mass injected by the volcano at time t0, and
τ is the e-folding time for the conversion of SO2 to sulfate
aerosol. An example of such a fit is shown in Fig. , and a
summary of the largest events observed by MLS is shown in
Fig. . Table shows the two fitted
parameters M0 and τ for each of these events.
Total injected SO2 mass M0 and decay time τ for a
selection of the volcanic events observed by MLS. Estimates of M0 from
other sources are shown for comparison. also give an
estimate of τ=9days for Kasatochi. For Sarychev we show two
estimates for two choices of highest pressure used (pmax). The
215 hPa result is more use for comparison with nadir-sounder data but is
less satisfactory as it proved difficult to remove the seasonal background
for this case.
pmax/NamehPaτ/daysM0/GgM0/Gg (from other sources)Sarychev14727±2571±42Sarychev21517±31160±1801200±200, 900 Kasatochi21527±11350±381373 , 2200 , 500–2500 (various references cited by )Nabro14720±2543±451500 , 650 (above 10 km) Grímsvötn21517±2108±11350–400 Kelut10034±7144±12200 (http://so2.gsfc.nasa.gov/pix/special/2014/kelut/Kelut_summary_Feb14_2014.html)Rabaul10034±5190±14230 Montserrat68.122±4139±24123–233 Manam (2005)68.120±499±13
The agreement between injected mass estimates from MLS and those found in the
literature is generally good. Clear discrepancies are the much larger values
reported by for Grímsvötn and Nabro. For Grímsvötn
this is unsurprising as most of the SO2 observed by MLS is at 316 and
215 hPa; it seems likely that a fraction of the plume was at rather low
altitudes, where MLS would be unable to observe it. For Nabro, most of the
SO2 observed by MLS is at 100 and 147 hPa, where we would expect the MLS
observations to be reasonably good. However, Nabro is at a rather low
latitude, where the ability of MLS to observe at lower altitudes is more
likely to be adversely affected by high levels of water vapour in the upper
troposphere. A more recent paper gives a mass of 650 Gg
above 10 km for the initial phase of the eruption; this is in somewhat
better agreement with the MLS estimate. It should be noted that all of the
eruptions during the Aura mission to date have been small compared to the
eruption of Mount Pinatubo in 1991; give a total mass of
Mt= 17 000 Gg for that eruption. The e-folding time for Pinatubo
was approximately 33 days, in reasonable agreement with the values in
Table .
Initial validation of the MLS data
The data we have described have not been validated. They appear
reasonable at a first glance in that the values are smaller than the
measurement noise except shortly after the eruption of volcanoes which
have been observed erupting from the ground and from other
satellites. Further validation from in situ measurements is not
straightforward as for most of the time, and at most places there is
not sufficient SO2 for MLS to detect. Volcanic eruptions happen
with little warning so it is not usually possible to plan aircraft or
balloon campaigns to coincide with them. The most tractable approach
is to cross-validate the various satellite SO2 measurements against
each other. While this process can not be considered watertight, we
can feel some increased confidence if several satellite measurements
using different spectral ranges and observation geometries are in
agreement with each other.
In this section we compare the standard MLS product against SO2 estimates
from the 190 and 640 GHz radiometers on the same instrument. We then compare
integrated column values calculated from the MLS data to column values from
the OMI instrument.
Internal consistency
As we noted in Sect. , MLS makes three partly independent
estimates of SO2: from the 190, 240 and 640 GHz radiometers. Some errors
are common to all these estimates: these include errors from the
temperature/pointing retrieval. Other errors, such as those from
spectroscopic data, should be more-or-less uncorrelated between the three
estimates. The field of view of the three radiometers is very similar, to the
extent that we can consider the three measurements to be made at exactly the
same time and place.
As Fig. but for
the eruption of the Soufrière hills volcano, Montserrat, on 20 May 2006.
As Fig. but for
the eruption of the
Soufrière hills volcano, Montserrat, on 20 May 2006. Colours represent
time in days; day 1 is 20 May 2006.
Isolating the volcanic SO2 from the
background value (with its spurious seasonal cycle). The example
shown is for the eruption of Kasatochi.
Mass Mt of SO2 above background
as a function of time after an eruption. The quantity plotted is
logeMt so that the exponential decay becomes a straight line.
The parameters M0 and τ were determined by fitting the
straight lines shown on this figure: the points were weighted with
the inverse square of errors which are constant in Mt and hence
not so in log(Mt). These errors are shown for Kasatochi. Only
the large dots are used in the fit.
Scatter plots of the 240 GHz SO2
against the estimates of the same molecule from the 190 (top row)
and 640 GHz (bottom row) radiometers. We show only the three
pressure levels at which all three radiometers provide usable
data. Data are shown only from periods with enhanced volcanic
SO2; a different colour is used for each such period. See
Table for details of which volcano was the cause of
each period of enhanced SO2.
We compare the 190 and 640 GHz products against the standard 240 GHz
product in Fig. . In order to avoid plotting an excessive
number of points which are effectively zero, we consider only periods of time
after a major eruption; a different colour is used for each such period. We
also show only points from profiles which contain a point at some level which
is more than 5.5 standard deviations above the mean for that level; this
criterion is less restrictive than that used in earlier sections. For each
scatter plot in Fig. we fit a straight line to the data in
two ways: first we assume that the standard product is the independent
variable and that all the errors are in the 640 or 190 GHz product (dashed
line y∼x in the figure). Next, we assume that all the errors are in
the standard product (dot-dash line x∼y). We note that in general the
640 GHz SO2 appears to underestimate the standard 240 GHz SO2; the
slopes of the fit lines tend to be considerably less than 1. In contrast, the
190 GHz SO2 tends to overestimate the standard product. At 46 hPa there
is very little correlation between the different SO2 products, but we note
that the four points with mixing ratios above 0.06 ppmv lie close to the
1 : 1 line. At 68 and 100 hPa the correlation is much stronger and a
larger fraction of the points are responsible for it.
Comparison with OMI
The OMI instrument is described by . OMI is a nadir
UV/visible imaging spectrometer. In its usual operating mode it observes a
2600 km wide swath with 60 image pixels across the width of the swath. The
pixels are 24 km across at nadir, becoming wider towards the edges of the
swath. In the along-track direction the pixels are 13 km across. The
algorithm used to derive total column SO2 from OMI measurements is
described by . It is suitable for most conditions but
underestimates the total column where that column is very large. The error
can be as large as 70 % for a column of 400 DU dropping to 20 % for a
column of 100 DU. The formula used requires that the vertical distribution
of SO2 is specified. The data files contain four separate estimates of
column SO2, each with a different assumption about the vertical profile.
The estimates are called PBL, TRL, TRM and STL, corresponding to
centre-of-mass altitudes of 0.9, 2.5, 7.5 and 17 km respectively. The README
file supplied with the OMI data states that the STL data are to be used when
studying explosive volcanic eruptions where the SO2 is placed in the upper
troposphere or lower stratosphere and that differences in actual
centre of mass from 17 km produce only small errors. As the useful SO2
measurements from MLS are at 10 km or above, we compare them only to the STL
product from OMI.
The OMI instrument has been affected by a somewhat mysterious problem known
as the “row anomaly”. This anomaly first appears in the data on
25 June 2007 and affects an increasing number of pixels over the subsequent
years.
Method
MLS (circles) and OMI (dots) column SO2
in Dobson units for part of an orbit on 11 August 2008. MLS points
are about 160 km apart along the orbit track; two other orbits are
shown. OMI pixels are 13 km apart along the orbit track and 24 km
apart at nadir across the track. The large gap in the OMI swath to
the east of the MLS points is caused by the row anomaly.
As the MLS data are vertical profiles of mixing ratio and the OMI data are
total column amounts, it is necessary to integrate the MLS profiles with
respect to a vertical co-ordinate to form a column amount in order to compare
the two data sets. The MLS column value will always be a partial column as the
instrument can not observe SO2 at altitudes below the 315 hPa level. The
formula used is derived in Appendix A of , assuming
that, between the standard MLS pressure levels, the mixing ratio varies
linearly in log(p). The column amount due only to the SO2 value at the
jth pressure level is
Nj=AMgfjΔ+pΔ+(log(p))-Δ-pΔ-(log(p)),
where p is pressure, A is Avogadro's number, g is the acceleration due
to gravity, M is the average mole mass of air and fj is the retrieved
volume mixing ratio at the jth pressure level. We use the notation that
Δ+x means the change in x between the jth and the j+1th level;
Δ-x is the change in x between the jth and the j-1th level.
The column value for MLS is obtained by summing the Nj values over a
suitable range of j. We use SI units for g and p so that we obtain a
value of Nj in molecules m-2; this is then divided by a factor of
2.687×1020 m-2 in order to convert it to Dobson units. If
the jth pressure level is the highest pressure level used, the resulting
value is essentially an estimate of the column above a pressure halfway in
altitude between level j and level j-1. The error in the resulting column
is a combination of the errors in the mixing ratios at all pressure levels
used. As we note in Sect. , the random errors in the data
appear to be larger than the precision estimates supplied with the data. We
therefore estimate the error in the MLS partial columns by calculating the
standard deviation of the partial columns in a region or day not affected by
volcanic SO2.
The OMI data are provided as column values in Dobson units. As OMI is
a nadir sounder and is on the same platform as MLS, its measurement
at a given location is made at 425±10 s after the MLS
measurement. This 7 min delay is small enough that we do not
attempt to correct for it. As the MLS horizontal field of view is
narrower than an OMI pixel, we only consider those OMI pixels which
are less than 18 km from the line joining successive MLS profile
positions. For each MLS profile we form an average of those OMI
pixels which meet this criterion and which are closer to that MLS
profile than to any other. For most MLS profiles this means that the
single MLS column is compared to a mean value of between 12 and
26 OMI pixels. We also calculate the standard deviation of this set of
pixels. The slight oblateness of the Earth means that the coincident
pixels are not all in the same pixel row of the OMI swath; the
coincident row varies from row 31 (of 60) near the poles to row 39
near the Equator. During the polar summer there may be usable OMI
data on the descending half of the orbit; the coincidences for such
data may be in a pixel row below the 31st. Figure
shows both MLS and OMI column values for a region in the north-east
Pacific. Both instruments show a region of large column values; this
is due to the eruption of Kasatochi which occurred a few days
earlier.
Scatter plot of OMI vs. MLS column
SO2 for the period 20 May 2006 to 9 June 2006 and for the
latitude range 9 to 16∘ N. SO2 above background
levels at this time is attributed to an eruption of the Soufrière
Hills volcano in Montserrat. All points are shown by a small
dot. Where the MLS SO2 profile has a clear peak above the
background noise, the points are shown with an extra symbol
corresponding to the pressure level at which the peak occurs. The
error bars represent the standard deviation of the OMI pixels which
are averaged to give the values plotted. Note that “above
100 hPa” means that we use the MLS data from this pressure level and levels
at higher altitudes. The error in the MLS columns is about 0.25 DU.
The solid line shows the ideal 1 : 1 relationship.
Results
Comparisons as described above can not be performed after January 2009 as the
OMI row anomaly affects pixel rows 28–38 from that time onward. We therefore
perform such comparisons for periods immediately following the eruptions of
Kasatochi on 8 August 2008, Montserrat, on 20 May 2006, Rabaul on
8 October 2006 and Manam on 28 January 2005. The results are shown in
Figs. to .
The agreement between MLS and OMI is reasonable for Rabaul, Manam and
Montserrat, where the SO2 profile peaks at 68 or 100 hPa. In all three
cases, the differences are on the order of 1 DU. This is about 4 times
larger than the 0.25 to 0.35 DU variability of the MLS data in the absence
of volcanic SO2, suggesting that a large part of the differences observed
is caused by the spatial variability of the volcanic SO2. For Kasatochi,
where there are measurable amounts of SO2 down to (and presumably below)
315 hPa, the MLS column tends to underestimate the OMI column by many tens
of DU, or 100–300 %. These differences are very much larger than the
0.8 DU scatter in MLS columns calculated in the same manner in the absence
of volcanic SO2 and are almost certainly due to the presence of SO2 in
the mid-troposphere where OMI is sensitive to it and MLS is not.
As Fig. but
for the period 7–15 October 2006 and for the latitude range 6∘ S
to 8∘ N. SO2 above background
levels at this time is attributed to an eruption of Rabaul in New
Britain, Papua New Guinea. The error in the MLS columns is about 0.35 DU.
Further validation possibilitiesACE-FTS
The ACE-FTS instrument measures a wide range of
chemical species in the upper troposphere and middle atmosphere using
the solar occultation technique at visible and near-infrared
wavelengths. The technique provides great sensitivity at the cost of
limited geographical coverage; data are only available at 15 sunrises
and 15 sunsets per day. The latitude at which these events occur
changes slowly throughout the year. SO2 is not a standard ACE-FTS
data product, but an experimental SO2 data set has been produced
. We have attempted to compare this data set to the
MLS data with no success; the limited coverage of ACE-FTS ensures that
on no occasion do the ACE-FTS measurement latitudes coincide with a
region of SO2 which is concentrated enough to be observable by MLS.
MIPAS
MIPAS was an infrared limb-sounding instrument on
Envisat. The MIPAS spectrometer was a Fourier transform type,
producing spectra with a high spectral resolution. Only a few species
were retrieved from these spectra on an operational basis, but
experimental retrievals have been produced for a number of other
molecules of which SO2 is one.
Monthly zonal means of SO2 have been produced by .
Comparisons of these data to MLS have not proved to be useful as the
non-volcanic background in the MLS data is, as we have noted earlier,
dominated by systematic errors. The small contribution to the monthly zonal
mean from volcanoes affects too small a number of months and latitude bins to
provide any kind of useful statistics.
A single-profile retrieval of SO2 from MIPAS is in development
(M. Höpfner, personal communication, 2014). A forthcoming paper
describing that data set will include a comparison
with MLS.
As Fig. but
for the period 28–31 January 2005 and for the latitude range 12∘ S
to 7∘ N. SO2 above background
levels at this time is attributed to an eruption of Manam in Papua
New Guinea. The error in the MLS columns is about 0.25 DU.
IASI
Several recent papers have
demonstrated the potential of the IASI instrument to provide useful SO2
measurements. In particular, compare the centre of mass of
MLS profiles with plume heights estimated from IASI and find very good
agreement.
Discussion
Although the progress we have made in validating the MLS SO2 data is
limited, the results so far are encouraging in that the measurements agree
well with OMI, and total masses of SO2 estimated from the MLS data are in
line with values previously published by a variety of authors.
As a tool for studying the dispersion of volcanic SO2, MLS has both
advantages and disadvantages over other currently operating satellite
instruments. Its main advantage is that it provides vertical profiles (see
Fig. ) and not just column amounts. The altitude of the SO2
is therefore measured directly rather than requiring to be inferred from the
known meteorology. Its main disadvantage is that the horizontal coverage is
very sparse compared to that of most nadir sounders, as shown in
Fig. . Designs for a future MLS-like instrument
provide for viewing directions at a number of angles to the
orbit track, allowing for global coverage at a resolution of 50 km; such an
instrument would combine the advantages of MLS with those of nadir-sounding
instruments and would be particularly valuable for the study of volcanic
SO2.
As Fig. but for the period 9–25 August 2008
and for the latitude range 29 to 83∘ N. SO2 above background
levels at this time is attributed to an eruption of Kasatochi in the Aleutian
Islands. The error in the MLS columns is about
0.85 DU.
Conclusions
The MLS instrument on Aura has observed enhanced SO2 mixing ratios
following a number of volcanic eruptions of various sizes. Total
injected masses of SO2 calculated from the MLS data agree with
previously published values in most cases.
The total column SO2 calculated from the MLS profiles agrees
well with the total column reported by the OMI instrument
under the right circumstances. The agreement is good for events
where most of the SO2 is clearly in the stratosphere (Montserrat
and Rabaul in 2006, Manam in 2005). Agreement is less good for events
where some of the SO2 is at lower altitudes. This may
be because there are significant amounts at altitudes below 215 hPa,
where the MLS sensitivity to SO2 is reduced or zero.
The MLS V2 data show a seasonal cycle with an amplitude of about
2 ppbv which is thought to be spurious. The seasonal effect is
smaller than the random error in an individual profile but becomes
obvious if sufficient profiles are averaged. This seasonal cycle needs
to be removed with some care if calculating the total mass of SO2
due to a volcanic eruption. Its presence means that the MLS V2 data can
not currently be used to study any seasonal cycle which might exist in the
non-volcanic SO2 background.
Acknowledgements
The authors thank Michael Höpfner and Chris Boone for providing
preliminary SO2 retrievals from MIPAS and ACE-FTS for comparison.
The authors thank the RCUK open access publication fund for paying
publication charges. Work on MLS in the UK has been funded by NERC.
MLS data used in this research were produced by the Jet Propulsion
Laboratory, California Institute of Technology, under contract with
the National Aeronautics and Space Administration. Edited by: G. Stiller
ReferencesBernath, P. F., McElroy, C. T., Abrams, M. C., Boone, C. D., Butler, M.,
Camy-Peyret, C., Carleer, M., Clerbaux, C., Coheur, P.-F., Colin, R., DeCola,
P., DeMaziere, M., Drummond, J. R., Dufour, D., Evans, W. F. J., Fast, H.,
Fussen, D., Gilbert, K., Jennings, D. E., Llewellyn, E. J., Lowe, R. P.,
Mahieu, E., McConnell, J. C., McHugh, M., McLeod, S. D., Michaud, R.,
Midwinter, C., Nassar, R., Nichitiu, F., Nowlan, C., Rinsland, C. P., Rochon,
Y. J., Rowlands, N., Semeniuk, K., Simon, P., Skelton, R., Sloan, J. J.,
Soucy, M.-A., Strong, K., Tremblay, P., Turnbull, D., Walker, K. A., Walkty,
I., Wardle, D. A., Wehrle, V., Zander, R., and Zou, J.: Atmospheric
Chemistry Experiment (ACE): Mission overview, Geophys Res. Lett, 32,
L15S01, 10.1029/2005GL022386, 2005.Brühl, C., Lelieveld, J., Crutzen, P. J., and Tost, H.: The role of
carbonyl sulphide as a source of stratospheric sulphate aerosol and its
impact on climate, Atmos. Chem. Phys., 12, 1239–1253,
10.5194/acp-12-1239-2012, 2012.Carboni, E., Grainger, R., Walker, J., Dudhia, A., and Siddans, R.: A new
scheme for sulphur dioxide retrieval from IASI measurements: application to
the Eyjafjallajökull eruption of April and May 2010, Atmos. Chem. Phys.,
12, 11417–11434, 10.5194/acp-12-11417-2012, 2012.Carn, S. A. and Prata, F. J.: Satellite based constraints on explosive SO2
release from Soufrière Hills Volcano, Montserrat, Geophys. Res. Lett,
37, L00E22, 10.1029/2010GL044971, 2010.Carn, S. A., Krueger, A. J., Krotkov, N. A., Yang, K., and Evans, K.:
Tracking volcanic sulfur dioxide clouds for aviation hazard mitigation, Nat.
Hazards, 51, 325–343, 10.1007/s11069-008-9228-4, 2008.Clarisse, L., Hurtmans, D., Clerbaux, C., Hadji-Lazaro, J., Ngadi, Y., and
Coheur, P.-F.: Retrieval of sulphur dioxide from the infrared atmospheric
sounding interferometer (IASI), Atmos. Meas. Tech., 5, 581–594,
10.5194/amt-5-581-2012, 2012.Clarisse, L., Coheur, P.-F., Theys, N., Hurtmans, D., and Clerbaux, C.: The
2011 Nabro eruption, a SO2 plume height analysis using IASI measurements,
Atmos. Chem. Phys., 14, 3095–3111, 10.5194/acp-14-3095-2014, 2014.D'Amours, R., Malo, A., Servranckx, R., Bensimon, D., Trudel, S., and
Gauthier-Bilodeau, J.-P.: Application of the atmospheric Lagrangian
particle dispersion model MLDP0 to the 2008 eruptions of Okmok and
Kasatochi volcanoes, J. Geophys. Res., 115, D00L11,
10.1029/2009JD013602, 2010.Doeringer, D., Eldering, A., Boone, C. D., González Abad, G., and
Bernath, P. F.: Observation of sulfate aerosols and SO2 from the Sarychev
volcanic eruption using data from the Atmospheric Chemistry Experiment (ACE),
J. Geophys. Res., 117, D03203, 10.1029/2011JD016556, 2012.Fischer, H., Birk, M., Blom, C., Carli, B., Carlotti, M., von Clarmann, T.,
Delbouille, L., Dudhia, A., Ehhalt, D., Endemann, M., Flaud, J. M., Gessner,
R., Kleinert, A., Koopman, R., Langen, J., López-Puertas, M., Mosner, P.,
Nett, H., Oelhaf, H., Perron, G., Remedios, J., Ridolfi, M., Stiller, G., and
Zander, R.: MIPAS: an instrument for atmospheric and climate research, Atmos.
Chem. Phys., 8, 2151–2188, 10.5194/acp-8-2151-2008, 2008.Haywood, J. M., Jones, A., Clarisse, L., Bourassa, A., Barnes, J., Telford,
P., Bellouin, N., Boucher, O., Agnew, P., Clerbaux, C., Coheur, P.,
Degenstein, D., and Braesicke, P.: Observations of the eruption of the
Sarychev volcano and simulations using the HadGEM2 climate model, J. Geophys.
Res., 115, D21212, 10.1029/2010JD014447, 2010.Höpfner, M., Glatthor, N., Grabowski, U., Kellmann, S., Kiefer, M.,
Linden, A., Orphal, J., Stiller, G., von Clarmann, T., Funke, B., and Boone,
C. D.: Sulfur dioxide (SO2) as observed by MIPAS/Envisat: temporal
development and spatial distribution at 15–45 km altitude, Atmos. Chem.
Phys., 13, 10405–10423, 10.5194/acp-13-10405-2013, 2013.Höpfner, M., Boone, C. D., Funke, B., Glatthor,
N., Grabowski, U., Gunther, A., Kellmann, S., Kiefer, M.,
Linden, A., Pumphrey, H. C., Read, W. G., Roiger, A., Stiller, G., Schlager,
H., von Clarmann,
T., and Wissmuller, K.: Sulfur dioxide (SO2) from MIPAS in the upper troposphere and lower
stratosphere, Atmos. Chem.
Phys. Discuss., in preparation, 2015.Krotkov, N. A., Schoeberl, M. R., Morris, G. A., Carn, S., and Yang, K.:
Dispersion and lifetime of the SO2 cloud from the August 2008
Kasatochi eruption, J. Geophys. Res., 115, D00L20,
10.1029/2010JD013984, 2010.Levelt, P. F., van den Oord, G. H. J., Dobber, M. R., Mälkki, A., Visser,
H., de Vries, J., Stammes, P., Lundell, J., and Saari, H.: The Ozone
Monitoring Instrument, IEEE T. Geosci. Remote, 44, 1093–1101,
10.1109/TGRS.2006.872333, 2006.Likens, G. E. and Bormann, F. H.: Acid Rain: A Serious Regional Environmental
Problem, Science, 184, 1176–1179, 10.1126/science.184.4142.1176,
1974.Livesey, N., Santee, M., Stek, P., Waters, J., Levelt, P., Veefkind, P.,
Kumer, J., and Roche, A.: A future “Global Atmospheric Composition Mission”
(GACM) concept, in: 2008 IEEE Aerospace Conference, IEEE,
10.1109/AERO.2008.4526243, 2008.Livesey, N. J. and Snyder, W. V.: EOS MLS Retrieval Processes Algorithm
Theoretical Basis, Tech. Rep. JPL D-16159, JPL, available at:
http://mls.jpl.nasa.gov/data/eos_algorithm_atbd.pdf (last access: 5 January 2015), version 2.0,
2004.
Livesey, N. J., Snyder, W. V., Read, W. G., and Wagner, P. A.: Retrieval
algorithms for the EOS Microwave Limb Sounder (MLS) instrument, IEEE T.
Geosci. Remote, 44, 1144–1155, 2006.Livesey, N. J., Read, W. G., Lambert, A., Cofield, R. E., Cuddy, D. T.,
Froidevaux, L., Fuller, R. A., Jarnot, R. F., Jiang, J. H., Jiang, Y. B.,
Knosp, B. W., Kovalenko, L. J., Pickett, H. M., Pumphrey, H. C., Santee,
M. L., Schwartz, M. J., Stek, P. C., Wagner, P. A., Waters, J. W., and Wu,
D. L.: Earth Observing System (EOS) Aura Microwave Limb Sounder (MLS)
Version 2.2 and 2.3 Level 2 data quality and description document., Tech.
Rep. JPL D-33509, NASA Jet Propulsion Laboratory California Institute of
Technology, Pasadena, California, 91109-8099, available at:
http://mls.jpl.nasa.gov (last access: 5 January 2015), 2007.Loughlin, S. C., Luckett, R., Ryan, G., Christopher, T., Hards, V., Angelis,
S. D., Jones, L., and Strutt, M.: An overview of lava dome evolution, dome
collapse and cyclicity at Soufrière Hills Volcano, Montserrat, 2005–2007,
Geophys. Res. Lett., 37, L00E16, 10.1029/2010GL042547, 2010.
Matoza, R. S., Pichon, A. L., Vergoz, J., Herry, P., Lalande, J.-M., il Lee,
H., Che, I.-Y., and Rybin, A.: Infrasonic observations of the June 2009
Sarychev Peak eruption, Kuril Islands: Implications for infrasonic
monitoring of remote explosive volcanism, J. Volcanol. Geoth. Res.,
200, 35–48, 2011.Prata, A. J., Carn, S. A., Stohl, A., and Kerkmann, J.: Long range transport
and fate of a stratospheric volcanic cloud from Soufrière Hills volcano,
Montserrat, Atmos. Chem. Phys., 7, 5093–5103, 10.5194/acp-7-5093-2007,
2007.Read, W. G., Froidevaux, L., and Waters, J. W.: Microwave Limb Sounder
measurement of stratospheric SO2 from the Mt. Pinatubo volcano, Geophys.
Res. Lett., 20, 1299–1302, 10.1029/93GL00831, 1993.Robock, A.: Volcanic eruptions and climate, Rev. Geophys., 38, 191–219,
10.1029/1998RG000054, 2000.
Robock, A. and Mao, J.: The Volcanic Signal in Surface Temperature
Observations, J. Climate, 8, 1086–1103, 1995.Rybin, A., Chibisova, M., Webley, P., Steensen, T., Izbekov, P., Neal, C.,
and Realmuto, V.: Satellite and ground observations of the June 2009
eruption of Sarychev Peak volcano, Matua Island, Central Kuriles, B.
Volcanol., 73, 1377–1392, 10.1007/s00445-011-0481-0, 2011.Schoeberl, M. R., Douglass, A. R., Hilsenrath, E., Bhartia, P. K., Barnett,
J., Beer, R., Waters, J., Gunson, M., Froidevaux, L., Gille, J., Levelt,
P. F., and DeCola, P.: Overview of the EOS Aura Mission, IEEE T. Geosci.
Remote, 44, 1066–1074, 2006.
Wadge, G., Robertson, R., and Voight, B.: Eruption of Soufriere Hills
Volcano, Montserrat from 2000 to 2010, in: Eruption of Soufriere Hills
Volcano, Montserrat from 2000 to 2010, edited by: Wadge, G., Robertson, R. E.
A., and Voight, B., Vol. 39 of Geological Society Memoirs, 1–501,
10.1144/M39.0, 2014.
Waters, J. W., Froidevaux, L., Harwood, R., Jarnot, R., Pickett, H., Read,
W., Siegel, P., Cofield, R., Filipiak, M., Flower, D., Holden, J., Lau, G.,
Livesey, N., Manney, G., Pumphrey, H., Santee, M., Wu, D., Cuddy, D., Lay,
R., Loo, M., Perun, V., Schwartz, M., Stek, P., Thurstans, R., Boyles, M.,
Chandra, S., Chavez, M., Chen, G.-S., Chudasama, B., Dodge, R., Fuller, R.,
Girard, M., Jiang, J., Jiang, Y., Knosp, B., LaBelle, R., Lam, J., Lee, K.,
Miller, D., Oswald, J., Patel, N., Pukala, D., Quintero, O., Scaff, D.,
Snyder, W., Tope, M., Wagner, P., and Walch, M.: The Earth Observing System
Microwave Limb Sounder (EOS MLS) on the Aura satellite, IEEE T.
Geoscience Remote, 44, 1106–1121, 2006.Waythomas, C. F., Scott, W. E., Prejean, S. G., Schneider, D. J., Izbekov,
P., and Nye, C. J.: The 7–8 August 2008 eruption of Kasatochi Volcano,
central Aleutian Islands, Alaska, J. Geophys. Res., 115, B00B06,
10.1029/2010JB007437, 2010.Yang, K., Krotkov, N. A., Krueger, A. J., Carn, S. A., Bhartia, P. K., and
Levelt, P. F.: Retrieval of large volcanic SO2 columns from the Aura Ozone
Monitoring Instrument: Comparison and limitations, J. Geophys. Res., 112,
D24S43, 10.1029/2007JD008825, 2007.Yang, K., Liu, X., Bhartia, P. K., Krotkov, N. A., Carn, S. A., Hughes,
E. J., Krueger, A. J., Spurr, R. J. D., and Trahan, S. G.: Direct retrieval
of sulfur dioxide amount and altitude from spaceborne hyperspectral UV
measurements: Theory and application, J. Geophys. Res., 115, D00L09,
10.1029/2010JD013982, 2010.