Eyjafjallajokull eruption

The version of the Unified EMEP model correspond to rv3_5_21 on FORECAST mode. The simulation was set up as chain of 6 day forecasts, driven by ECMWF meteorology (IFS, operational 00 UTC forecast). The simulation domain covers only [25W–35E]x[36N–75N] with a 0.2×0.2 resolution. The full extended model domain extends down to 31N, but the simulation goes only down to 36N.

The fields retrieved for the full extended forecast domain [25W–35E]x[31N–75N], from MARS include:

Emission set-up for Forecasts and other experiments performed during the eruption.

Maximum plume height comes from London VAAC reports.

Emission rates of PM10 and PM2.5 interpreted from PDF and discussion with Ármann Höskuldsson (University of Iceland). Emission rates for the 14-16 are towards the higher end of the observed mean total ash emissions from the 14-16 and the 17-19 are from the low end of the observed mean total ash emissions from the 14-16 because emissions were reported to have decreased after the first three days. For the 20 it is lower because of reports that there is now less ash in the plume.

Emission rates of SO2 for the 15 are based on OMI SO2. The SO2 then decreases through the 18th consistent with reports of weaker plumes and OMI not observing new puffs clearly and then increases on the 19th. The SO2 values except for on the 15th are not based on total OMI observed SO2, rather they are intended to represent trends in the SO2 emissions.

• Volcanic eruption emission are hard-coded into the model.
• The eruption started on 2010-04-14 at 00 UTC and continues trough the simulation.
• Emissions are treated as a vertical line source.
• Emission rates for each day are assumed constant trough the column height.
• Ash emissions are treated in the model as primary PM emissions.
• In order to model only the volcanic plume only ash emissions are taken into account as primary PM emissions, i.e. all other primary PM emissions are neglected.
  • Primary fine PM (PPM25) emission rates are set accordingly to PM2.5 eruption parameters.
  • Primary coearse PM (PPMCO) emission rates are derived from PM10-PM2.5 eruption parameters.
• Volcanic SO2 emissions are added to the standard model emissions.
• Standard model emissions are set accordingly to the TNO-MACC-1 emission dataset.

• Under several assumptions, the aerosol optical depth (AOD) can be estimated from the fine and coarse PM columns. In order to provide a simple comparison to satellite AOD observations, the partial contribution of the volcanic particles was calculated (P-AOD) from the primary fine and coarse columns.

• The simulations are run on titan (UiO) on 8×2 processors.
• The driving bash and perl scripts at:
  • titan:~alvarov/Unify/run/cwf_ext-2010.sh
  • titan:~alvarov/Unify/run/modrun.rv3.CWF-EXT.pl
• Source and meteorology files are located at:
  • titan:~alvarov/Unify/Unimod.rv3_5_21.CWF-EXT/
  • titan:/data3/metno/emep/Data/FORECAST/metdata/
    • 2010/meteo-ext201004??_??.nc (EMEP Format)
    • 2010_ECMWF/cwf-ext_*_f201004??.nc (retrieved from MARS)
• The raw output and images is located at:
  • titan:~alvarov/CWF/output.rv3.CWF-EXT.2010
  • titan:~alvarov/CWF/output.rv3.CWF-EXT.2010/www/images

The latest images (*.png) and animations (*.mpg) for each forecast day can be found in the institute ftp server.

• All images for a given forecast (or hindcast) are collected in one single zip file, e.g. 20100414.zip contains all images for the forecast starting on April 14th.
• Each forecast lasts 6 days, starting at 00 UTC (145 hourly records).
• If emissions estimates improve, past forecast will be re-run (hindcast) and the previous files updated. Zip files for a hindcast include images only for that day (25 hourly records).
• For consistency, zip files are updated when the graphic layout changes or when additional variables are plotted.

• Hourly images are named according to the following convention: var_lev_fsdate_date.png, where:
  • var is the variable name (e.g. ppm25 for fine volcanic particles):
  • lev is the level (e.g. tcol for total column);
  • fsdate is the forecast start date (in YYYYMMDD format);
  • date is the forecasted date and hour (in YYYYMMDDHH format);
• Animations for the whole period are named var_lev_fsdate.mpg.

• GOME-2, SCIAMACHY & OMI SO2 column
• MODIS AOD
• SEVIRI ASH last 12 hours

• SNAP ash dispersion
• MACC SO2 column
• EURAD PM10
• Fall3D ash dispersion
• NERI http://www.dmu.dk/International/Air/Models/volcanicplume/
• Flexpart passive tracer and ash
• Puff http://puff.images.alaska.edu/watch_Eyjafjoll.shtml

• Time-line wikipedia

Emissions are distributed linearly from 1670 m (vent height) to Eruption height in Experiment Tables. Emissions are divided into the size bins as in Size Bins Table.

The mass mean diameter and % of ash in size bin data comes from measurements made using laser by Throstur Thorsteinsson (ThrosturTh@gmail.com), Institute of Earth Sciences, University of Iceland. Measurements were made on 15, 17, 18 April and 6-7 May. We are using the average of these four measurement days. The density of the ash particles is a measurement of the bulk ash made by Throstur. The sigma values are the same as what was using in the model previously, just applied to the new size bins.