Title: Modeling Volcanic Ash Transport and Dispersion: Expectations and Reality
1Modeling Volcanic Ash Transport and
DispersionExpectations and Reality
René Servranckx Peter Chen Montréal Volcanic
Ash Centre Canadian Meteorological Centre
2Presentation Topics Ash Transport Models
- Reality 20 years from now
- Expectations for Ash Transport Models (TM)
- Reality today / Limiting factors
- Areas for improvement
3June 22, 2024
4 June 22, 2024
5June 22, 2004!
6- ACCURATE guidance on SPACE / TIME LOCATION / 3D
STRUCTURE of airborne ash - LITTLE (or no) UNCERTAINTY (ash / no ash)
- TIMELY delivery
- Implications for TM?
What area aviation users TM expectations?
7Components Ash Modeling Problem
- Accuracy and timeliness of TM guidance depends
on - Volcanic Ash Source (Source Term / eruption
parameters) - Meteorology
- Transport and Dispersion (TM)
8Despite uncertainties, TM
- Are of great value!
- Especially important for REAL-TIME, operational
response - Sometimes, the ONLY guidance available
- Must be used in conjunction with other tools
(remote sensing, etc.) - Can not be used blindly!
9Limiting Factors VOLCANIC ASH SOURCE
- Eruption parameters largely unknown / poorly
quantified - Detection of eruptions / airborne ash is
problematic - Poor quantitative estimates of atmospheric ash
loading / only 2D
? 3D is needed for TM - Threshold ash concentrations that pose threat to
aviation (?) ? May be
very small (NASA DC-8 Hekla incident) - Visual Ash Cloud criterion on TM guidance
is subjective - DEFAULT SCENARIOS and LOW THRESHOLD values in TM
guidance
10Limiting Factors METEOROLOGY
- HORIZONTAL and VERTICAL resolution of Numerical
Weather Prediction (NWP) Models - Vertical coordinates are not Flight Levels
standard atmosphere - Representation of earths surface (topography)
in NWP models - Mt Mckinley, AK 6194 m
2640 m - Incomplete knowledge of initial conditions of
the atmosphere - Predictability of atmosphere / Accuracy of NWP
vary with flows / patterns
11Limiting Factors TRANSPORT / DISPERSION
- VOLCANIC ASH SOURCE component
- METEOROLOGY component
- Parameterization of dispersal, removal and
deposition of ash - Real time assimilation of airborne ash is not
done - Predictive ability varies with atmospheric
conditions
12Areas for improvement
VOLCANIC ASH SOURCE COMPONENT
- 1998 and 2003 WMO / ICAO volcanic ash meetings
Substantial improvements could be
made in TM guidance if source term estimates were
improved - ICAO (IAVW Ops Group) to IAVCEI
QUANTITATIVE estimates of eruption parameters for
TM? - NASA DC-8 encounter with Hekla diffuse plume
damage from very small ash concentrations (?)
13Areas for improvement
VOLCANIC ASH SOURCE COMPONENT
- If unconditional ash-avoidance is the rule,
small concentrations must be accurately predicted
? Good estimate of Source term is important! - Remote sensing Any technological advancement
that might improve quantitative estimates of the
3D distribution of airborne ash - Assimilation of volcanic ash data in TM
Exploratory work has been done (Siebert et al.
2002 NOAA Air Resources Laboratory) - How much can we achieve? ? Highly dependent on
remote sensing improvements (quantitative 3D
distribution)
14Areas for improvement METEOROLOGY and DISPERSION
/ TRANSPORT Components
- Improvements to NWP Models are ongoing
- Improvements to TM also ongoing
- ENSEMBLE FORECASTING Already done for NWP
Models applicable to TM - Many runs (single or multiple models) using
slightly different initial conditions - BASIC IDEA AVERAGE of many runs BETTER than
single run - Spread among runs is gives an estimate of
uncertainty
15Example Visual ash clouds from 4 TM runs
valid at same time
16Finagles Laws of Information
- The information you have is not what you want
- The information you want is not what you need
- The information you need is not what you can
obtain - The information you can obtain costs more that
you want to pay
17Corollaries
- What you see / interpretation depend on
- Tools / Technology
- How information is presented
- How one looks at information
- What you see may not be what you get !
18 19 20 21Impact of changing visual ash cloud value
- ALL IMAGES TO FOLLOW ARE FROM SAME TRANSPORT
MODEL RUN WITH SAME SOURCE TERM CONDITIONS - 1 hour eruption of Cleveland starting 15 UTC 19
Feb 2001 - Images valid 45 hours after start of eruption
- CANERM (TM) diagnostic average ash concentration
in FL200 - FL350 (micrograms per cubic meter) - ? perception of where ash is or is not present!
22100
80
200
10
1
50
23Summary Transport Models
- Expectations are high
- Despite uncertainties, VALUABLE!
- Must be used with other sources of information
- Can not be used blindly /require careful
interpretation / knowledge of uncertainties - New ways of looking at information and estimating
uncertainties (Ensemble forecasts) - Accuracy can be increased by reducing
uncertainties What can we do to bridge
the gaps?