The Grimsvotn Volcanic Eruption

At approximately 17:30 UTC on the 21st May a volcanic eruption started at Grimsvotn in Iceland.  The eruption occurred roughly 30 minutes before Harold Camping, an evangelical preacher from the US, had predicted that a giant earthquake would mark the start of the world’s destruction.  Seismic tremors at the Grimsfjall station did start at 17:30 UTC, reaching a peak at 19 UTC, however the world did not destruct.  Instead a plume of volcanic ash was ejected into the atmosphere reaching altitudes of 20km. The ash plume formed a circular cloud around the eruption site extending ~ 60km in diameter. Heavy ash fallout occurred close to the volcano and visibility was reduced to near zero.  In addition, according to Giles Harrison, the amount of volcanic lightning that Grimsvotn produced was contributed to 20% of all lightning detected on the planet at the time.

Seismic tremor at the Grimsfjall station, Iceland (Icelandic Met Office)

Grimsvotn is Iceland’s most frequently erupting volcano, most recently erupting in 2004.   There have been 12 volcanic eruptions in Iceland since mass air travel began in the 1980’s, equating to one eruption every 1 ½ years.  So, rather than viewing the 2010 Eyjafjallajokull and 2011 Grimsvotn eruptions as unusual events one could say that Europe has been lucky that, for the previous 30 years, volcanic ash emitted from Iceland’s many volcanoes has been dispersed northwards rather than south-eastwards towards Europe’s busy airspace.

Photo of Grimsvotn eruption on the evening 21 May 2011. Photo: Ólafur Sigurjónsson.

Why was disruption to air travel following the eruption of Grimsvotn so much less than for Eyjafjallajokull last year?

  • Firstly the weather patterns were different to last year. In April 2010 a persistent high-pressure system was located over the UK resulting in the ash cloud being transported towards Europe and remaining stagnant in that region. During May of this year the weather situation was much more dynamic. A low pressure system was travelling from the Atlantic into the North Sea varying the direction that the ash was dispersed and hence any disruption was more intermittent.

    UK Met Office synoptic analysis at 00UTC on 23 May 2011.
  • Secondly, a change in policy following the Eyjafjallajokull eruption from  zero ash tolerance to low, medium and high ash concentration thresholds meant that a blanket closure of airspace did not occur.  The London Volcanic Ash Advisory Centre, which is responsible for issuing advice about the dispersion of volcanic ash emitted from Icelandic volcanoes, issued quantitative predictions in line with these new ash thresholds.  These, more detailed, volcanic ash forecasts resulted in partial airspace closures where ash concentrations were high.

    UK Met Office volcanic ash dispersion forecast at 00UTC on 24 May 2011
  • Thirdly, it was widely reported that the ash particles emitted from Grimsvotn were larger than those emitted from Eyjafjallajokull.  Whilst I have seen no evidence of this so far, it is true that ash particles with diameters > 10μm  sediment out of the ash cloud with terminal velocities of several km’s per day compared to smaller particles which have sedimentation velocities of meters per day only.


What have I been doing to contribute to our current understanding of volcanic ash dispersion?

The duration of the Eyjafjallajokull eruption last year meant that the resulting ash cloud was the most intensely observed ash cloud ever. A wide variety of measurements were made using ground based and airborne lidars, in-situ particle measuring devices on aircraft and radiosondes etc.  This means that we can now evaluate volcanic ash dispersion models more quantitatively than before.   I have been assessing the impact of meteorology on the distribution of volcanic ash.  The results allow us to make quantitative predictions of volcanic ash concentrations and also to infer a volcanoes characteristics, such as its plume height and particle size distribution, using a combination of measurements and model simulations downstream.  See Dacre et al. (2011) http://www.agu.org/journals/jd/papersinpress.shtml  for more details.
Helen Dacre

Leave a Reply

Your email address will not be published. Required fields are marked *