Volcanic SO2 emissions

On this page we describe the volcanic emissions inventories that are used by GEOS-Chem.

OMI-based volcanic emissions

Implementation in GEOS-Chem 12.3.2

Volcano (v7) emissions for 1978-2019 were obtained from NASA/GMAO. See the README file for details on how these files were created.

The emission inventory is available as daily text (.rc) files in the format:

  ###  LAT (-90,90), LON (-180,180), SULFUR [kg S/s], ELEVATION [m], CLOUD_COLUMN_HEIGHT [m]
  ### If elevation=cloud_column_height, emit in layer of elevation
  ### else, emit in top 1/3 of cloud_column_height
  volcano::
  50.170 6.850 3.587963e-03 600. 600.
  45.780 2.970 3.587963e-03 1464. 1464.
  42.170 2.530 3.587963e-03 893. 893.
  38.870 -4.020 3.587963e-03 1117. 1117.
  ::

The text files can be read directly into HEMCO via the Volcano (formerly AeroCom_Volcano) extension. It can be activated in the HEMCO extension section:

  117     Volcano           : on    SO2
      --> Volcano_Source    :       AeroCom
      --> Volcano_Table     :       $ROOT/VOLCANO/v2019-04/$YYYY/$MM/so2_volcanic_emissions_Carns.$YYYY$MM$DD.rc

The Volcano_Source option was added to specify OMI or AeroCom emissions. If OMI-based emissions are used then the closest available year will be used if the simulation year is outside of the available time period.

--Melissa Sulprizio (talk) 19:05, 1 March 2019 (UTC)

Original implementation

Volcanic emissions of SO2 for 2005-2012 from Ge et al. (2016) are based on satellite measurements of SO2 from OMI (Ozone Monitoring Instrument) and ancillary information from the Global Volcanism Program. These data can be found in ExtData/HEMCO/VOLCANO/v2018-03/.

Cui Ge wrote:

Our degassing data are from Carn’s OMI retrieved data.

The difference for eruptive emissions is the 'cloud column height'. The emissions amount directly come from Carn’s, and the 'cloud column height' is from Smithsonian Institution Global Volcanism Program (GVP) report. Usually there are several records including radar, satellites and pilot reports for each eruption, we intended to get a better estimation by using the average of all the reasonable observations for each eruptive events. In the supplementary material of our paper, we did some sensitive tests about the uncertainties caused by different eruptive height and different eruptive activity duration.

Our data is only from 2005-2012, including:

• 48 eruptive volcanoes;
• 8 tropical degassing volcanoes.

If the elevation is the same as the cloud column height, the volcano is considered degassing, otherwise it's eruptive. Following the recommendation of Thomas Diehl, erruptive emissions are emitted into the top 1/3 of the emission plume. Most entries are degassing and there are only a few eruptive volcanoes each year.

--Melissa Sulprizio (talk) 17:17, 28 March 2018 (UTC)

References for OMI volcano emissions

1. Carn, S.A., K. Yang, A.J. Prata, and N.A. Krotkov, Extending the long-term record of volcanic SO2 emissions with the Ozone Mapping and Profiler Suite (OMPS) Nadir Mapper, Geophys. Res. Lett., 42, 925-932, doi:10.1002/2014GL062437, 2015.
2. Carn, S., Multi-Satellite Volcanic Sulfur Dioxide L4 Long-Term Global Database V3, Greenbelt, MD, USA, Goddard Earth Science Data and Information Services Center (GES DISC), Accessed: 28 Feb 2019, 10.5067/MEASURES/SO2/DATA404, 2019.
3. Ge, C., J. Wang, S. Carn, K. Yang, P. Ginoux, and N. Krotkov, Satellite-based global volcanic SO2 emissions and sulfate direct radiative forcing during 2005-2012, J. Geophys. Res. Atmos., 121(7), 3446-3464, doi:10.1002/2015JD023134, 2016. (PDF)

--Bob Yantosca (talk) 14:16, 22 March 2019 (UTC)

Aerocom-based volcanic emissions

Data for GEOS-Chem v10-01 and higher versions

For GEOS-Chem v10-01 and higher versions, the volcanic emissions of SO2 were taken from AEROCOM point-source data (release date 01 Mar 2010, cf. Thomas Diehl). Christoph Keller then used the Python script volc2grid.py to convert these point-source data files to regularly gridded data files.

# volc2grid.py
#
# DESCRIPTION:
# Converts volcano point source data into a gridded netCDF file.
# The volcano data is assumed to be in the form provided by Thomas
# Diehl (thomas.diehl@nasa.gov), available from
# http://aerocom.met.no/download/emissions/AEROCOM_HC/volc/
# 
# These data contain daily volcano point source emissions, including
# the location of the volcano (lon,lat) as well as the crater elevation
# and the emission plume height (both in m asl).
# The vertical information is mapped onto the hybrid-pressure sigma
# grid provided in a separate input file. The barometric equation
# is used to convert between elevation in m and pressure.
#
# Following the recommendation of Thomas Diehl, erruptive emissions
# are emitted into the top 1/3 of the emission plume.

Christoph Keller wrote:

I have also updated the volcano data by splitting it into eruptive and degassing. The diagnostics are now also split up again. These updates are included in the updated configuration file and HEMCO source code.

The volc2grid.py script created new Aerocom data files (in COARDS-compliant netCDF format) for use with HEMCO. These new data files are contained in the HEMCO data directory tree. For detailed instructions on how to download these data files to your disk server, please see our Downloading the HEMCO data directories wiki post.

--Bob Y. 13:20, 3 March 2015 (EST)

UPDATE: As of July 8, 2015, we have AEROCOM volcano emissions (separated into eruptive and degassing) for years 1979 through 2009. These are stored in the HEMCO data path HEMCO/VOLCANO/v2015-02.

--Bob Yantosca (talk) 14:56, 11 March 2019 (UTC)

Original Aerocom documentation

NOTE: This data (dated 06 Mar 2009) was used in versions of GEOS-Chem used in versions of GEOS-Chem prior to v10-01.

Volcanic SO2 emissions
Contact: Thomas Diehl
Release Date: 06 Mar 2009
http://aerocom.met.no/download/emissions/AEROCOM_HC/readme_volc

Description of the emission file

The file volc_so2.nc contains volcanic SO2 emissions and other variables for all days from January 1st 1979 to December 31st 2007 for all volcanoes with historic eruptions listed in the Global Volcanism Program's database provided by the Smithsonian Institution. Subglacial and submarine volcanoes are excluded. Each element of a variable is dubbed an "event", where an event corresponds to the emission of SO2 by a specific volcano on a specific day. The order of events is as follows:

ED1Vm ... ED1Vn   ED2Vp ... ED2Vq ...... EDkVr ... EDkVs 

(with E indicating events, V volcanoes, and D days). The dimension nevents contains the total number of events. Currently, nevents=12,403,236. The following variables are provided:

vid

the standardized 8-character volcano number, which can be used in data processing for identification purposes.

vname

the volcano name as provided by the GVP database; this variable is for informational purposes only and should not be used in data processing, since it contains non-ASCII characters.

jdn (integer)

the Julian Day Number

so2 (float)

the amount of SO2 emitted by the given volcano on the given day in kt

cloud_column_height (integer)

the height of the volcanic plume in m above sea level

elevation (integer)

the elevation of the volcano in m above sea level

lon (float)

the longitude of the volcano

lat

the latitude of the volcano

The following four variables contain the grid indices of the volcanoes; they are specific to the grid of the GOCART model. Other models will typically have to compute their own indices from the provided longitude and latitude.

ic

the longitudinal grid index for a resolution of 2.5° x2.0°

jc

the latitudinal grid index for a resolution of 2.5° x2.0°

if

the longitudinal grid index for a resolution of 1.25° x1.0°

jf

the latitudinal grid index for a resolution of 1.25° x1.0°

Since the inventory is constructed such that cloud_column_height = elevation for non-eruptive degassing, the SO2 emission should be placed only in the model level which contains the crater elevation for the case cloud_column_height = elevation. For all other cases, the emission should be injected into some fraction of the levels located within the plume. In the GOCART model, the SO2 emission is evenly distributed among the levels located within the top third of the plume.

Methodology used for compiling the inventory

Volcanoes are assigned posteruptive, extraeruptive, and pre-eruptive degassing rates depending on their eruptive state. In addition, we use the SO2 emission rates provided by Andres & Kasgnoc (1998) for 49 quasi-continuously erupting volcanoes. Additional results from observations (from the TOMS and OMI instruments and COSPEC measurements) were retrieved from the literature in a number of cases. For those days of eruptive periods without data from instruments, we compute an averaged daily emission rate from either the mass of the ejected magma (if available) and intraeruptive degassing, or from the VSI corresponding to the listed VEI and intraeruptive degassing. The plume height is derived from the VEI (for days with SO2 data from instruments) or from the mass weighted average of the VEI-based plume height and the emission height for intraeruptive degassing (assumed to be at the elevation of the crater). In some cases, the plume height is taken from measurements provided in the literature.

A more detailed description of the procedure, which assigns the SO2 emissions to each volcano for each day, is given below:

1. Volcanoes with a timeframe code of D1 or D2 in the GVP database (i.e. those which had an eruption since 1900) are assigned a posteruptive degassing rate of 7.0x10-2 kt/d. All other volcanoes with historic eruptions are assigned an extraeruptive degassing rate of 6.2x10-4 kt/d.
2. These values are overwritten with the SO2 emissions listed in Andres & Kasgnoc (1998) for 49 quasi-continuously erupting volcanoes.
3. On days when the TOMS instrument provides SO2 data the entries from 1. and 2. are overwritten by the TOMS detected SO2 amount.
4. Additional results from observations (from the OMI instrument or COSPEC measurements) were retrieved from the literature for a number of cases. These values overwrite entries from 1.-3. on the available days.
5. For each eruption listed in the GVP table, the 7 days preceding the start of an eruption are assigned a pre-eruptive degassing rate of 7.5x10-1 kt/d. This step is not applied if a value was assigned for this volcano on this day in step 2, 3, or 4.
6. For the days of each eruption listed in the GVP table, an SO2 value is assigned as follows:
   1. No value is assigned if an SO2 value was already assigned in steps 2-4 for a specific day and volcano.
   2. If none of the days within the eruption period was assigned an SO2 value in 2-4, the SO2 emitted during this period is derived from the mass of the ejected magma if this quantity is available. Otherwise, the SO2 amount is derived from the VSI corresponding to the VEI listed in the GVP table for this eruption. Also, an intraeruptive degassing rate is applied to ndays-1 days in both cases (where ndays is the length of the eruption period in days), and an average daily emission rate for the whole period is calculated.
   3. If one or more day(s) within the eruption period were assigned SO2 values in steps 2-4, we approximate the SO2 emitted during the remainder of the eruption period by applying an intraeruptive degassing rate to ndays-odays days, where odays are the number of days within the eruptive period with assigned values from 2-4. If the mass of the ejected magma is available and if SO2magma > SO22-4, the residual SO2magma - SO22-4 is added and an average daily emission rate for the remaining days of the eruption period is calculated.

The plume height is derived from the VEI for days with SO2 data from instruments. For other days, it is derived from the mass weighted average of the VEI-based plume height and the emission height for intraeruptive degassing, which is assumed to be located at the elevation of the crater. In some cases, the plume height is taken from measurements provided in the literature.

Non-eruptive degassing

Our rates for preeruptive and intraeruptive degassing are simply the average of the range given by Berresheim & Jaeschke (1983) for these classes of degassing: degaspre = 7.5x10-1 kt/d and degasintra = 7.5x10-1 kt/d. Preeruptive degassing is applied to the 7 days preceding the start of an eruption.

We modified the rates provided by Berresheim & Jaescke for posteruptive and extraeruptive degassing, given as 1.05x10-1 kt/d and 5x10-3 kt/d, respectively. These rates were inferred (like the rates for intraeruptive and preeruptive degassing) from a small sample of volcanoes. Applying this rate to all posteruptive days of all volcanoes would probably overestimate the amount of SO2, since not all volcanoes are exhibiting this type of posteruptive degassing (or at least not on all days of their posteruptive period). Since we generally do not know the individual volcanoes which display this type of degassing, we reduce the degassing rate by the factor actualpostv/totpostv (where actualpostv is the actual (estimated) number of volcanoes displaying posteruptive degassing and totpostv is the total number of volcanoes which erupted during the past 100 years) and apply this effective rate to the posteruptive days of all volcanoes which erupted during the past 100 years. The number of 365 volcanoes given by Berresheim & Jaeschke exhibiting posteruptive degassing is most likely an overestimate, while the number of 102 provided by Stoiber et al. (1987) is probably an underestimate since they did not account for degassing without plumes, i.e. they omitted the fumarolic part of the posteruptively degassing volcanoes (see the discussion in Graf et al., 1997). In the current (as of September 2008) GVP database, 365 volcanoes had a subaerial eruption after 1900. On average, 60 volcanoes have an eruption per year(number from volcano.si.edu). Not counting the erupting volcanoes to a first approximation, we get totpostv = 305 volcanoes with potentially posteruptive degassing. Approximating actualpostv with (305+102)/2 yields 0.105x0.667 = 7.0E-2 kt/d.

For the extraeruptive case, we do not have a lower boundary from observations (like the 102 volcanoes in the posteruptive case) to derive a scaling factor. Thus, we simply evenly distribute the global 0.5 kt/d from B&J onto all 807 volcanoes in the GVP database whose last eruption occurred before 1900 (excluding volcanoes for which it is uncertain whether the last eruption was in the Holocene). This yields 6.2x10-4 kt/d per volcano. Thus, the rate is only about 10% of the one in B&J (5.0x10-3 kt/d). Alternatively, this can also be interpreted as only 10% of the volcanoes, which had eruptions before 1900 in the GVP database to display extraeruptive degassing (with the unmodified rate from B&J) in our inventory.

Estimation of SO2 from ejected magma

In some cases, the GVP database provides volume estimates for the erupted lava and/or tephra. In these cases, we use the formula from Blake (2003) to estimate the SO2 mass released during the eruption: mSO2 [Mt] = 1.77 mmagma[Gt] 0.64

Since the provided lava and tephra volumes are bulk volumes, they must be converted to dense rock equivalent (DRE) volumes. Averaging a range of values found in the literature yields ftephra = 0.5 and flava = 0.85. For (DRE, we use 2.7 g/cm3.

References for AeroCom emissions

1. AeroCom hindcast emisions and READMEs: http://www-lscedods.cea.fr/aerocom/AEROCOM_HC/
2. M. Chin, R.B. Rood, S.-J. Lin, J.-F. Muller, and A. M. Thompson, Atmospheric sulfur cycle simulated in the global model GOCART: Model description and global properties, J. Geophys. Res., 105, D20, 4671-24687, 2000.
3. Andres, R. J.. and A. D. Kasgnoc, A time-averaged inventory of subaerial volcanic sulfur emissions, J. Geophys. Res., 103, 25251-25261, 1998.
4. Berresheim, H., and W. Jaeschke, The Contribution of Volcanoes to the Global Atmospheric Sulfur Budget, J. Geophys. Res., 88, 3732-3740, 1983.
5. Blake, S., Correlations between eruption magnitude, SO2 yield, and surface cooling, in Volcanic Degassing, Special Publication of the Geological Society of London No. 213, edited by C. Oppenheimer, D. M. Pyle, and J. Barclay, 177-202, Geological Society, London, UK, 2003.
6. Carn, S. A., A. J. Krueger, G. J. S. Bluth, S. J. Schaefer, N. A. Krotkov, I. M. Watson, and S. Datta, Volcanic eruption detection by the Ozone Mapping Spectrometer (TOMS) instruments: a 22-year record of sulphur dioxide and ash emissions, in Volcanic Degassing, Special Publication of the Geological Society of London No. 213, edited by C. Oppenheimer, D. M. Pyle, and J. Barclay, 177-202, Geological Society, London, UK, 2003.
7. Graf, H.-F., J. Feichter, and B. Langmann, Volcanic sulfur emissions: Estimates of source strength and its contribution to the global sulfate distribution, J. Geophys. Res., 102, 10727-10738, 1997.
8. Newhall, C. G., and S. Self, The Volcanic Explosivity Index (VEI): An Estimate of Explosive Magnitude for Historical Volcanism, J. Geophys.Res., 87, 1231-1238, 1982.
9. Schnetzler, C. C., G. J. S. Bluth, A. J. Krueger, and L. S. Walter, A proposed volcanic sulfur dioxide index (VSI), J. Geophys. Res., 102, 20087-20091, 1997.
10. Simkin T, Siebert L (2002-). Global Volcanism FAQs. Smithsonian Institution, Global Volcanism Program Digital Information Series, GVP-5 (http://www.volcano.si.edu/faq/), 2008.
11. Stoiber, R. E., S. N. Williams, and B. Huebert, Annual Contribution of Sulfur Dioxide to the Atmosphere by Volcanoes, J. Volcanol. Geotherm. Res., 33, 1-8, 1987.

--Bob Y. 10:13, 16 February 2010 (EST)

Previous issues that are now resolved

Bug fix for eruptive volcanic emissions

This update was included in GEOS-Chem 12.3.0, which was released on 01 Apr 2019.

Barron Henderson wrote:

I had previously brought up that I think there is a problem in the VOLCANO emissions. At the time a new volcano option was expected to supersede this option, but I think that VOLCANO is still the default.

The problem is with the eruptions being treated as cycling. The VOLCANO emissions use hemco's cycling option for both degassing and for eruptions. The last year in our dataset is 2009, which had a particularly active June. For example, 2009 June SO2 eruption mass is 4 times the average of the 2000-2009 June. The cycling option takes 2009 and applies it to *all* future years.

In pre-HEMCO emissions, the degassing was treated as cycling and eruptions were limited to available years. I think the old approach is preferable.

Obviously, an aware user can disable the cycling by replacing "C" with "R" for eruptions in HEMCO_Config.rc. I suspect, however, that many people are not thinking about this.

--Melissa Sulprizio (talk) 16:26, 28 March 2019 (UTC)