Vertical distribution of ash at source

1
How sensitive are NAME ash plume forecasts
to input source
characteristics? Helen Dacre, Robin Hogan,
Stephen Belcher, University of Reading, UK
(h.f.dacre_at_reading.ac.uk)

• Introduction
• The Eyjafjöll volcano erupted on 14/04/10 causing
  widespread disruption. The Met Office dispersion
  model, NAME, is used to predict the evolution of
  the ash plume. The accuracy of these predictions
  depends heavily on the input of accurate source
  characteristics.
• The aim of this work is to analyse the
  sensitivity of the NAME model results to
  uncertainty in the input source characteristics.
• Case Study 16th April 2010
• A high-pressure system was located over the UK
  and the north Atlantic and a low-pressure system
  was located over northern Europe.
• The ash plume was observed by a ground-based
  lidars at Chilbolton.

Overview
Chilbolton
Mass concentration at 00UTC on 16/04/10. (Left)
average concentration from 0-12km . (Right)
vertical cross-section taken from 45-55o N and at
2oW. Note the contours are factors of 10
concentrations in arbitrary units.
Synoptic Analysis 00UTC 16/04/10
Lidar backscatter 10-18UTC 16/04/10

• Initially the ash plume is advected SE from
  Iceland towards Europe.
• After 24 hours the ash plume diverges one
  branch is advected around the high-pressure
  system whilst another branch is advected around
  the low-pressure system.
• During the16/04/10 the anti-cyclonic branch of
  the ash plume was advected over the UK.
• The vertically slanted structure observed by the
  lidar is captured by the NAME model and is a
  result of vertical wind shear.

• Vertical distribution of ash at source
• Time-height plots of mass concentration at
  Chilbolton. The height above the summit into
  which ash particles are emitted at the source is
  (a) 3-3.5km, (b) 4-4.5km, (c) 5-5.5km and (d)
  6-6.5km.
• The height at which ash is emitted into
  atmosphere influences the plume evolution as wind
  speed and direction vary with height.
• The operational model assumes a uniform
  distribution of ash from the summit to the plume
  height to account for a fluctuating plume.
• Emitting ash in layers at decreasing heights
  above volcano decreases the plume depth and width
  and delays the arrival time over Chilbolton.

• Particle size distribution at source
• Mass concentration from 0-12km at 00UTC
  on16/04/10 for (a) 0-30µm (b) 30-10µm (c)
  100-300µm diameter particles

• Emission rate of source
• Time-height plots of mass concentration at
  Chilbolton for varying mass flux in the sub 100µm
  range, (a)106 kg/s (operational), (b) 105 kg/s
  and (c)104 kg/s.
• The mass flux attributed to the sub 100µm range
  determines the magnitude of the mass
  concentration contours.
• An empirical relationship is used to determine
  the emission rate H 0365M0225, where H
  (km) is the maximum plume height above the summit
  an M is the total emission rate (kg/s).
• For an initial plume height of 8.5km above the
  summit, this implies a total emission rate of
  106 kg/s.
• The mass concentration over Chilbolton scales
  linearly with the percentage of the total mass
  flux in the sub 100µm range.

• Observed Mass Concentration
• Time-height plot of mass concentration determined
  from a combination of lidar retrievals and sun
  photometer measurements at Chilbolton from
  10-18UTC on 16/04/10.
• The lidar observes the ash plume with a depth of
  3km, width of 500m and arrival time at 12UTC on
  16/04/10.
• Peak concentrations of 800µg m-3 are observed at
  a height of 1.8km
• Conclusions
• Vertical and horizontal structure of ash plume
  predicted by NAME model is very sensitive to
  height at which ash is emitted above volcano and
  is also to sensitive to particle size
  distribution.
• Magnitude of mass concentration is determined by
  the of mass in sub 100µm range. Comparison with
  observations suggests between 1 and 10 of total
  emitted mass is contained in this range.