12/9/11 INTERNATIONAL AIRWAYS VOLCANO WATCH OPERATIONS GROUP (IAVWOPSG) SIXTH MEETING Dakar, Senegal, 19 to 23 September 20111 Agenda Item 5: Operation of the IAVW 5.2: Improvement of the format and content of all volcanic ash related messages INTERCOMPARISON OF DANGER AREA IN VAAs THAT INCLUDE OBSERVATIONAL DATA COMPARED TO CONCENTRATION CHARTS FROM DIRECT MODELLING OUTPUTS (Presented by France and Canada) SUMMARY Taking examples from the eruptions for which both concentrations charts and VAA (volcanicc ash advisory) have been produced and disseminated to users, this paper shows that the direct concentration charts approach is more conservative (i.e. produces large areas) than the one based on VAA benefiting from the expertise of forecasters and satellite observations. 1. INTRODUCTION 1.1 After the eruption of Eyjafjallajökull in April 2010, changes in the contingency plan for air traffic management (ATM) in case of volcanic ash led to a new requirement for the European VAACs. The Toulouse and London VAACs made developments to produce volcanic ash concentrations charts based on two, then three thresholds corresponding to HIGH, MEDIUM and LOW contamination. 1.2 The experience in providing users with pure model concentrations charts has been increasing since then, with the recent volcanicc activity in Europe, the Middle East and in the south Atlantic ocean. 1.3 Toulouse VAAC produced both VAA and concentrations charts for volcanic ash clouds within its area of responsibility and noticed that differences and sometimess apparent inconsistencies could exist between the two products. This is because the VAAs from Toulouse VAAC take into account satellite data, something that direct modelling outputs (i.e. concentration charts) do not. (19 pages) IAVWOPSG.6.IP.012.5.en. docx
- 2-2. DISCUSSION 2.1 During the Cordon Caulle eruption in 2011, and the Nabro eruption in Eritrea, both concentrations charts and VAA/VAG were issued by the Toulouse VAAC. 2.2 The concentrations charts were directly drawn from the METOFRANCE dispersion model MOCAGE, initialised with all information available at that time (volcanologists, discussion groups, etc). 2.3 VAA and VAG were based predominantly on satellite imagery for the observation part and integrated the motion indicated by the dispersion model for the forecast part (+6, +12, +18) as well as the synoptic meteorological features observed on METEOSAT imagery. 2.4 Appendix A shows both the first panel (observed or estimated volcanic ash cloud) of the VAG for a given date positioned in front of the corresponding(s) concentrations charts, reduced to the same scale for a more intuitive comparison. 2.5 Appendix B shows the original files (VAG and concentrations charts as produced and disseminated) that were used for this comparison. 2.6 Appendix C shows another example produced by VAAC Montréal for the Grimsvötn eruption in May 2011. It clearly shows that adjusting direct modelling outputs (specific concentration charts) with observational data (satellite data, AIREP, etc.) produces a better VAA/VAG than using only pre-determined concentration values. 2.7 It is clear from all the examples given in Appendices A, B and C that the danger area (described as the MEDIUM/HIGH contamination of the concentration charts, not to mention the LOW) is often larger than the area described as contaminated (without reference to a numerical value) in the VAAs that integrate observations. This makes the pure model predictions more conservative (i.e. larger areas) which is not good for aviation. 2.8 The group may wish to note that no reports of VA encounter were received by the Toulouse VAAC during the duration of issuance of VAA and concentrations charts. Many AIREP of sulphur smell and yellowish clouds (sulphate aerosols) were received by VAAC Montreal over Canadian airspace during the Grimsvötn event which helped define the location of the volcanic cloud. 2.9 The examples are given in Appendices A, B and C of this paper. 3. ACTION BY THE IAVWOPSG 3.1 The IAVWOPSG is invited to note the information in this paper.
Appendix A APPENDIX A CORDON CAULLE Eruption 14 June 2011 12h00 UTC (both figures show the VA cloud at the same scale) FL200-350 CORDON CAULLE Eruption 16 June 2011 06h00 UTC (both figures show the VA cloud at the same scale) FL200-350
Appendix A A-2 CORDON CAULLE Eruption 16 June 2011 18h00 UTC (both figures show the VA cloud at the same scale) FL200-350 CORDON CAULLE Eruption 20 June 2011 12h00 UTC (both figures show the VA cloud at the same scale) SFC - FL200
A-3 IAVWOPSG/6-IP/12 Appendix A CORDON CAULLE Eruption 21 June 2011 00h00 UTC (both figures show the VA cloud at the same scale) SFC - FL200 NABRO Eruption 15 June 2011 12h00 UTC (both figures show the VA cloud at the same scale) FL200-400 SFC-FL200
CORDON CAULLE Eruption 14 June 2011 12h00 UTC APPENDIX B IAVWOPSG/6-IP/12 Appendix B
Appendix B B-2 CORDON CAULLE Eruption 16 June 2011 06h00 UTC
CORDON CAULLE Eruption 16 June 2011 18h00 UTC B-3 IAVWOPSG/6-IP/12 Appendix B
Appendix B B-4 CORDON CAULLE Eruption 20 June 2011 12h00 UTC
CORDON CAULLE Eruption 21 June 2011 00h00 UTC B-5 IAVWOPSG/6-IP/12 Appendix B
Appendix B B-6 NABRO Eruption 15 June 2011 12h00 UTC
Appendix C APPENDIX C GRIMSVÖTN ERUPTION: A PERSPECTIVE FROM VAAC MONTREAL ON 24-26 MAY 2011 REGARDING THE IMPORTANCE OF ADJUSTING DIRECT MODELLING OUTPUTS WITH SATELLITE DATA AND AIREP TO PRODUCE THE VAA/VAG FIGURE 1: VAG issued by VAAC London on 24 May 2011 based on modelling concentrations of at least 200 micrograms per cubic meter based on modelling. At initial time of 18 UTC on the upper left panel (identified as observed ash in the volcanic ash advisory), the VAG shows a zone of volcanic ash from SFC to FL550 over a large portion of southern Greenland extending over Davis Strait and Baffin Island. Meteorological Watch Offices using only this information would have issued volcanic ash SIGMET covering a very large area. FIGURE 2: The use of satellite imagery in real time paints a very different picture that the one on the VAG. Visible channel satellite imagery captured by the Meteorological Service of Canada s antenna in Resolute, Nunavut (74.72N, 94.94W) on 24 May 2011 by NOAA-19 at 1701 UTC (top image on next page) and NOAA-15 at 1723 UTC (bottom image on next page) shows nearly clear sky conditions based on open waters, ice flows and the coast Greenland with ice inland. Brightness temperature differences (ch4-ch5) for these images did not indicate ash. Based on these and all available information, the diagnostic from VAAC Montreal was that little or no ash was present. A long but thin and fairly high level layer of SO2, and sulphate aerosols with no or very little ash was however identified based again on satellite imagery as shown on figure 3.
C-2 Appendix C NOAA-19 imagery in the visible band valid 24 May at 1701 UTC NOAA-15 imagery in the visible band valid 24 May at 1723 UTC
C-3 IAVWOPSG/6-IP/12 Appendix C FIGURE 3: NOAA-15 imagery in the visible band valid 25 May at 1009Z (top) and 1346 UTC (bottom) showing a high level cloud of sulphate aerosols / SO2 that contained very little or no ash based on ch4- ch5 differences.
Appendix C C-4 FIGURE 4: Atmospheric transport modelling output from VAAC Montreal valid 25 May 2011 at 08 UTC showing the average concentrations in the layer FL200-FL350 (below). An AIREP near Coral Harbour (red circle) at 0853 UTC indicated smell of sulphur in cockpit at FL300. Many other AIREP were received over the next 48 hours (green circles). Note that the modelling also shows an area of high concentrations over Greenland, Davis Strait and Baffin Island. However, since this is not corroborated by satellite imagery, VAAC Montreal would not have show ash in its VAA / VAG in this case had it been the lead VAAC. This shows the importance of considering all sources of information (in this case, modelling guidance, satellite imagery and AIREP) to produce the VAA / VAG. This is where the professional knowhow and expertise of VAAC meteorologists can provide significant added value compared to using direct modelling outputs especially at initial time ( observed as in the VAA). In the present case, using only specific concentrations values such as 200, 2000 and 4000 micrograms per cubic meter to produce the VAA / VAG is completely wrong.
C-5 FIGURE 5: A few AIREP over Canada up to 16 UTC on 25 May 2011 IAVWOPSG/6-IP/12 Appendix C
Appendix C C-6 FIGURE 6: VAAC London VAG valid starting 25 May at 06 UTC (upper left panel of top image) and based directly on modelled ash concentrations (bottom image) from FL000 to FL200
C-7 IAVWOPSG/6-IP/12 Appendix C FIGURE 7: VAAC London VAG valid starting 25 May at 06 UTC (upper left panel of top image) and based directly on modelled ash concentrations (bottom image) from FL200 to FL350
C-8 Appendix C FIGURE 8: VAAC London VAG valid starting 25 May at 06 UTC (upper left panel of top image) and based directly on modelled ash concentrations (bottom image) from FL350 to FL550 END