London VAAC Calling

Ash column over Mount Redoubt. Photograph by R. Clucas, USGS. Used under Wikimedia Commons.

Lately there has been a bit of discussion about Flight Level Ash Advisories and how people interpret them visavi actual columnal height.

Some have been disappointed when it has turned out that the advisories have been up to 3 times as high as the confirmed numbers released by the various agencies at a later stage. So, I thought it was time to clear things up a bit.

So, I sent an email to London VAAC asking them about how they do, since they always tend to be spot on with their advisories in regards of ash height. And they answered in such a perfect manner that I will leave their answer ad verbatim, and then I will talk a bit in the end about difference with other VAAC-agencies.


London VAAC

Pinatubo minutes after the climactic eruption. The image is not related to the work of London VAAC. Photograph by unknown USGS member. Used under Wikimedia Commons.

Dear Carl,
Thank you for your enquiry.

As meteorologists for VAAC London, we have been aware of the Volcanocafé blog for a couple of years, and regularly enjoy the numerous informative articles and discussions on there.

To answer your question, there are a number of different sources that are used to determine the height level of an eruption, however the exact ones used will depend upon the location, and accessibility of equipment and observations at the time. It is also important to note that other volcanic ash centres may have a different approach, and I can only talk for VAAC-London.

Firstly, VAAC-London are fortunate enough to have a close working partnership with the Iceland Met Office (IMO), as this is where the vast majority of the volcanoes within our area of responsibility lie. This means that any eruption that occurs is assessed by them, through a mixture of ground based observation, pilot reports, and local radar. They then report the observed height of the eruption, and this is what we use in the first instance. However, this is a working partnership, and any information either of us have to ensure the correct values are used, is freely shared.

One way we can help guide IMO regarding the observed height of the eruption, is by combining satellite observations, the observed motions of the ash, and also our modelling of the atmospheric winds at different heights. These then have to match up with what we are observing. So for example, an eruption that reaches stratospheric heights could show ash in satellite imagery moving in a very different direction or speed to that at lower levels. This is also likely the best method to use for any remote volcanoes, for example, if we had to assess one from outside our region of responsibility in order to support another centre (e.g. Africa, which is the responsibility of VAAC-Toulouse)

Photograph of the US Naval Station in Subic Bay after Pinatubo. The picture is not related to the work of London VAAC. Photograph by Sgt Paul Bishop, US Navy. Used under Wikimedia Commons.

Once the ash is airborne, we are then able to verify height levels through a mixture of radar and further ground based observations (e.g. LIDAR), as well as continuous monitoring on various satellite channels. Where possible, research flights and pilot reports would also help us here.

However there is no “correction” factor included, as our aim would be to model the ash cloud as observed, and track its progress, and any such factor would affect the accuracy of these forecasts. However, if there is a range of uncertainty, it is possible that we would assume the topmost level in a first instance, and reassess later.

Any safety decisions would then be down to the airlines themselves, as we simply forecast where the ash will be.

Finally, we do have some information on scenarios, especially for the major Icelandic culprits, based upon historical eruptions, and these can be used for “what-if” exercises. However each new eruption would be considered on its own, and assessed from the observations we have, and the information proved to us by IMO.

Kind regards,

London VAAC



The work of all VAAC agencies is to keep the public safe. The airplane was safely on the ground during Pinatubos eruption. Photograph by R.L. Gieger, USGS. Used under Wikimedia Commons.

Not all agencies have the clear advantage of cooperating with the Icelandic Met Office. They have all sorts of equipment to help with discerning ash height, including specialized mobile ash-radars. This makes it possible for London VAAC to be as exact as is possible.

Other agencies might not have this advantage, they instead operate on other less exact data, against language barriers, and have other methodology.

Some seem to operate on the “better safe than sorry”-methodology. So, they go on the worst-case scenario giving ash heights that are initially quite high, and then dial them back as reliable figures come in.

For a VAAC agency safety is the paramount issue, and that is not necessarily the same as satisfying volcano-aficionados with metre-exact columnal heights. London VAAC will be useful, others not so much from our viewpoint.

I guess what I am trying to say is, know thine VAAC, and now we know London VAAC.

I would hereby like to graciously thank London VAAC for the essential work that they do, and their help clearing things up with an official answer.


37 thoughts on “London VAAC Calling

  1. Thanks Carl for passing this on. Many of the same techniques were used out here in California to track the smoke from the recent massive fires we’ve had the last few years. In the case of the CARR fire near Redding last year, the monstrous Pyro Cb that erupted over the fire (and spawned an F4 fire-nado), reached the tropopause. Smoke injected that high circumnavigated the entire northern hemisphere…creating possible health problems for 10’s of millions of people.
    By tracking the smoke accurately, sensitive people were given advance notice of degraded air quality…which I’m sure helped reduce the resultant health impacts.

    • It is the event which was expected, though I think it was expected to be larger. People have been allowed to go back home, this time it seems for good. The debris didn’t go as far as feared, no blocked roads or damaged houses. There’s still a considerable amount of the mountain which is slowly slumping/sliding, but it’s not considered extremely dangerous anymore.
      The shock when the fall happened was felt for miles.

  2. Great reply from Kirk and thank you for sharing! (And thank you for London VAAC for taking time out to talk to us!)

      • It is currently a pond rather than a lake, I think. And it is pretty steep, so it needs to get a lot deeper before it grows big enough to graduate.

        • I don´t know – 50 meter width and 100 meter length sound kind of lake-ish to me! Should be around 10 meters deep now? They expect it to reach between 60 and 70 meter depth!

          • There isn’t a real definition for ‘pond’ and ‘lake’. But the word ‘pond’ comes from ‘pound’ and originally meant an enclosure. The word ‘pool’ is related. This one may qualify for that.

            The largest pond I am aware of is Great Pond in Maine, at 20 square kilometres. Admittedly that is pushing things a bit.

          • I hereby suggest that we split the difference and name it The Great Kilauea Plonk.

    • I take a bit of issue with them holding up Amero as their example. The USGS has done extensive analysis of lahar flow paths for US volcanoes and updated hazard maps accordingly. If local government does not enforce build restrictions I don’t see how that is the USGSs fault. Naturally, this falls under my mantra of “Don’t be there” but it also skirts the boundaries of how much liberty the land owners are willing to give up to support an intrusive government. It is completely foolhardy to build on any sort of flood plain or outflow path, but people do. Entire cities sit in hazard zones for some kind of threat.

      • Similar problem existed when Mt. St. Helens erupted. A study had been done years earlier detailing what the mountain has done before and can do again, but no one paid any attention until after the eruption.

    • I am afraid its absolutely normal for people to object without even knowing what they are objecting to and why. In this case it just needs someone to say ‘they want to fly in and install some measuring instruments destroying part of the volcano’ and everyone piles in saying “no”.

      Trying to explain what you want to do and why is pretty pointless. Can you prove that nothing will be altered? “Well, we have to dig some holes and put a small PV cell in” “aha! So there will be a big PV cell disturbing the wilderness, no way!”.

      End of argument. These same people will complain when earthquakes start and nobody is able to tell them anything useful because there is no information. Then its still your fault.

      Happens to nearly everything all the time.

      • Dunno, I saw a mid sized volcanic field south of Mobile indicated in a paper discussing basement controls for the Gulf of Mexico, purportedly this volcanic field is Jurassic era and is beneath about 10 to 15 km of sediment. I’ve seen no indication of seismic activity for it, and it seems about as dead as Doorpoint volcano over near the mouth of the Mississippi.

        • Got a link to the paper.

          Note that this points out the 14 or so ancient transform faults from when the Gulf of Mexico opened up. The Port St Joe fault is the one running past Panama City Florida to somewhere up in Escambia County Alabama. North of Panama City is karst topology and several sink-hole lakes from where the basement rock has fractured the limestone bedrock. No known sinkholes near Pensacola, though there are a couple of springs either side of the Escambia river, Mystic Springs up north of here and some spring diametrically across the river from it. Further north near Jay Florida is where our tiny quakes were at. I think the largest that system has produced was a Mag 4.8 years ago up in Alabama.

        • Suddenly all sold out.. They are used to ‘extrapolate’ brexit and what happens once in the abyss.

  3. In the words of a relationship status, ‘it’s complicated’. Let me count the ways…

    1) As the article correctly points out, Iceland now has amongst the best monitored volcanoes in the world, following an enormous upgrade effort over the past nine years. Many parts of the world have almost no instrumental monitoring. Cloud also varies considerably over the world and this affects remote sensing – for example, it’s almost always much easier to monitor Italy during an Italian summer than Papua New Guinea during a monsoon.

    2) In general terms, no single instrumentation is sufficient. Radar, for example, is fantastic, but if the eruption is higher than what the radar is set up to look at (geometrically or in software), it’s not so useful. This is why the *two* radars on site during the Pinatubo eruption could not see the top of the eruption (plus it’s hard to measure the height of something that’s raining rocks directly on you).

    3) The scene that you’re looking at can be very confused, particularly in the moist tropics where the atmosphere is primed for deep convection all the time. The target might be an eruption, or a bushfire, or a volcanic cumulonimbus, or a regular cumulonimbus. Consider these log extracts from the Pinatubo radar team:

    21 June 1991 0051UT weakened power is limiting range to 10 miles. Lack of cells past 12 miles provides evidence that this is correct; therefore venting may be impossible to detect. 0202UT No venting noted; picking up 2 echoes around 20 nm out. Echoes weak. No other performance changes observed. 0255UT Unconfirmed report of eruption at 23Z – 00Z, max tops reported to have been at 28,000 ft [8.5 km]. 0314UT No venting noted. 0758UT Minor venting to 20,000 ft [6.1 km]

    18 July 1991 1355UT Possible venting/eruption to 40,000 ft [12.2 km]. No confirmation from Clark. They call it a rain shower – but have been getting ash all evening, getting echoes from up to 58,000 ft [17.7 km] east of volcano – 5 nm [9.3 km]. 1415UT Air Force still report thunderstorm. 1416UT Air Force finally decided it was an eruption. Cloud is heading east. 1430UT Eruption continues to 55,000 ft [16.7 km] – is broken up. Does not seem continuous. Echoes very strong [strong eruption continues, but at 1545UT radar masked by heavy rain until 1630Z]

    15 July 1991 0415UT Venting to 12,000 ft [3.7 km]. Mount appears to be inducing the cells around it. Cells north and east max tops to 40,000 ft [12.2 km]. 0615UT Venting to 30,000 ft [9.1 km]. Ash cloud mixing with Cbs in vicinity moving WSW. Large cell 020 estimated 5 miles [8 km] from Cubi extending toward Mount Pinatubo max tops 60,000+ ft [18.3+ km]. 0650UT Hard to pick out venting height since ash and steam are mixing in with the rainshowers and thunderstorms

    4 Aug. 1992 1149 UT Noted dense haze and large Cb NNE after observation, noted large area of echoes just east and west of Mount Pinatubo. Echoes were in excess of 60,000 ft [18.3 km]. Unable to determine whether echoes were precip, ash, etc. Called Pinatubo Volcano Observatory at Clark, no sightings or seismic activity there.

    These are just a few samples (more in the source paper), but they show how chaotic and rapidly changing a scene can be even with trained military weather observers. From that, a VAAC is going to pluck a single number to represent the height of an eruption.

    4) That height thing… does everybody know that tropical convection can and does regularly reach around 17 km or more, without a volcano being involved? A volcanic eruption in that environment can produce much higher plumes without necessary much ash being involved. The net effect is that you have a field of deep convection, some of which might be volcanic, but all of a similar height. That’s why you so commonly see FL450 – FL550 as maximum heights from the tropical VAACs.

    5) Detrainment height – the height of maximum ash detrainment is not going to be the highest height of the eruption, unless the eruption is a limp thing that’s not penetrating an inversion. It goes up, it falls down a little, then it spreads out. The VAAC is trying to work out how high to make the plume in the knowledge that it’s probably just a little higher. Also, the highest bit is probably ice and SO2 and other gases with a tiny bit of ash – how long are we going to track that so far above cruising levels?

    6) Ground observation bias – there is a strong bias in the historical (pre-satellite era) record, and to a lesser extent in the post-satellite era record, caused by the maximum height of the eruption rarely being observable from the ground (think night, cloud, ashfall). Plus, most volcanologists are not trained weather observers, and many volcano observatories are much too close to the volcano to see the top of a high cloud. So, not only has much of traditional volcanic plume theory has developed in the context of a biased historical record.but apparently simultaneous obserservations of the same event from different disciplines can differ. The same applies for pilot observations – there is a large variety in skill in what the VAACs receive.

    There’s more…. but I should stop! Thanks for the article. Further reading on my particular perspectives below.


    Tupper, Andrew & Oswalt, J. & Rosenfeld, Daniel. Satellite and radar analysis of the volcanic-cumulonimbi at Mount Pinatubo, Philippines, 1991. Journal of Geophysical Research. 110. 10.1029/2004JD005499. 2009

    Tall clouds from small eruptions: the sensitivity of eruption height and fine ash content to tropospheric instability A Tupper, C Textor, M Herzog, HF Graf, MS Richards – Natural Hazards, 2009

    Reducing discrepancies in ground and satellite-observed eruption heights A Tupper, R Wunderman
    Journal of Volcanology and Geothermal Research 186 (1-2), 22-31, 2009

    This comment was held for approval by the system. Hereby released with thanks – admin

    • Many thanks for this response! We are always very happy to hear from professionals. Your points are well made, and show why VAAC’s should err on the side of caution. After all, they exist for the purpose of air safety, and not to keep volcano watchers entertained.

      The point for us is probably not to jump to conclusions. It takes time to determine the true magnitude of an eruption.

      Given this difficulty against monsoon and thunderstorm conditions, it is perhaps not a surprise that eruptions were missed. The 1808 eruption is a case in point: tropical, presumably Pacific, and not a clue which one it was.

    • that was an excellent comment – I hadn’t considered ash entrained in normal weather systems being potentially dragged to high altitudes – while the eruption itself might not have thrown the ash that high on its own – makes a lot of high VAAC reports at small eruptions make sense 🙂

    • The messed up bit is that I understood most of what the log entries were saying. I had an assignment to do “data fusion” for an event years ago and to take different sensor operator notes and place them in a cohesive timeline for the exercise referees/evaluators. They read quite like that.

  4. Thanks to Mots Motsfo I learned that I survived an M5.5 here in Tanzania.
    It was located near Lake Tanganyika at the Great Rift, over att Dodoma 90 something kilometres away it felt like someone was driving a very small truck through the hotel grounds. Nothing much to write home about.

    So, for those with overactive imagination I would be dead in quakemageddon, but for those who are more used to earthquakes it was not something to get up for. To be honest, if I had not been told about it I would not have remembered it.

  5. We are getting closer to the aliens

    Among already 4000+ exoplanets discovered, this is the first exoplanet confirmed to have a water atmosphere located in the habitable orbital region around the red dwarf star (

    The planet has a radius twice the Earth (so probably it’s a rocky planet) and is located about 0.15 AU from the star, so about half way the orbit of Mercury. So it could still mean extreme temperature variations on the planet or extreme radiation, but who knows… Scientists say that the planet may be also covered with a global ocean or an icy surface.

    I would guess volcanoes would place a big role in creating an atmosphere rich in water vapor.
    The rest of the atmosphere is apparently hydrogen and helium. Which is a bit weird as both gases escape smaller planets quite easily.

    The star located 111 light years ago shines at magnitude 13.5 in the constellation of Leo, so a good target for my 10 inch telescope 🙂

    Being potentially a old star, it makes a perfect candidate for life detection.
    The next generation of space telescopes, to be launched in the 2020s, will allow a good chemical analysis of that planet (and many others). Scientists hope to find trace gases, that would be a signature of life (that can only be produced by life – that on itself is another big assumption).

Leave a Reply