1. WRF Modelling of Volcanic Ash Dispersion
Ralph Burton1, Stephen Mobbs2, Alan Gadian1
1 NCAS Weather 2 NCAS Director
The model can be initialised with a single column (for large-scale runs) or with a point source plus large surface thermal perturbation (for high-resolution runs.)
The plume height and emission rates are a function of time
Three ash bins are included: 0.1μm, 1 μm and 10 μm (with an assumed density of 2500 kg/m3)
Additionally, the tracer can be removed at the surface or when precipitation occurs – although these representations of deposition are not switched on for the results shown.
The WRF model has been used to simulate the dispersion of volcanic
ash using several passive tracers. The tracers has been given
settling velocities to mimic the effect of sedimentation.
The settling velocities are a function of height AMSL (see
e.g. Sparks et al. 1997). WRF is run with its usual full range
of microphysical schemes, etc. This allows a dispersion model to be
embedded within a full NWP model.
Eyjafjallajökull,
May 2010
Grimsvötn, May 2011
The plots below show isosurfaces of ash concentration at several times after eruption. The isosurface is coloured by hydrostatic pressure to show the height of the ash cloud.
The top-left image is from MODIS
taken on the morning of the
6th May 2011 (from NASA Earth
Observatory, 2010).
The ash cloud is clearly
seen.
The lower left image shows
modelled isosurface of total ash
concentration (brown colour)
for 6th May 12Z. Also shown is an
isosurface of cloud water mixing
ratio.
Note the model has the cloud in
the correct location. N.B. both images
use the same domain.
T+18
T+36
Both the Grimsvötn and
Eyjafjallajökull runs are initialised
with operational ECMWF analyses.
The WRF model resolution is 15km
with 121 vertical levels.
T+54
T+72
High-resolution runs
Here, the model is initialised with a sounding from Keflavikurflugvollur (24th April 2010 , see left) and is run over a limited domain at 100m resolution (with boundary layer parametrisation switched off.)
Note the fine structure visible in the ash cloud, and the layering due to the stable layers in the profile.
Side view:
Aspect ratio 1:1
Ash from above
25 km
Further work:
Progress is currently underway at Leeds to implement a multiphase version of WRF, whereby ash particles will have mass and there is a two-way coupling between the fluid and particle phases.
Acknowledgements
Thanks to Wei Wang, NCAR, for details of the tracer code; thanks to Sue Laughlin for plume heights from Grimsvotn and to Steve Sparks for communications regarding settling velocities.
References
Sparks, R. S. J. et al. Volcanic Plumes, Wiley, 1997
MODIS image retrieved from NASA Earth
Observatory, 27th June 2011: