Why Thunderstorms Are So Dangerous for Airliners

airplanethunderstormAs I write this, AirAsia flight 8501 has been missing for less than 24 hours, and in the absence of wreckage its too early to speculate on what happened. But the flight, which took off from Surabaya bound for Singapore, appears to have been traveling through an area of intense thunderstorm activity, so it may be instructive to look at the kind of danger this sort of weather can present to aircraft.

The region around the equator is known to meteorologists as the Intertropical Convergence Zone, or ITC. Here, the heat and moisture of warm ocean waters provides the energy to power tremendous updrafts that produce clusters of thunderstorms called a Mesoscale Convenction Complex. These storms can punch up through the stratosphere up to 50,000 feet, far above the crusing altitude of commercial airliners. From Smartcockpit.com:

A thunderstorm brings together in one place just about every known weather hazard to aviation. A single thunderstorm cell can hold 500 000 tons of water in the form of liquid droplets and ice
crystals. The total amount of heat energy released when that much water is condensed amounts to approximately 3 x 1014 calories. Equated with known energy sources, this falls just below an entrylevel hydrogen bomb. Even a small thunderstorm would have the caloric equivalent of a Hiroshimatypeatomic weapon… The thunderstorm occupies a unique place in the pantheon of aviation meteorology because it is the one weather event that should always be avoided. Why always? Because thunderstorms are killers.

Some of the deadly forces include lighning, airframe icing, large hailstones, extreme turbulence, and downdrafts that can reach speeds in excess of 100 mph. Perhaps the greatest hazard facing a modern airliner, however, is the sheer volume of precipitation that a thunderstorm can put out.

On May 24, 1988, a TACA 737 en route from Belize to New Orleans was descending towards its destination when it blundered through a thunderstorm. At an altitude of just 2000 feet, a deluge of rain and hail doused the flames of its twin turbofans. Unable to regain power, the captain managed through superb airmanship to put the stricken plane down undamaged atop a mile-long levee. Notes superb aviation writer Peter Garrison:

The event was not unique. Nine months earlier, an Air Europe 737 descending through rain and hail over Thessaloniki, Greece, had suffered a double flameout. In that case, the crew managed to restart the engines and land without trouble. In 2002, a Garuda Indonesia 737, also descending among thunderstorms, suffered a double flameout over Java. Its crew ditched the airplane in a river; one person died, and there were a dozen serious injuries.

According to preliminary reports, the pilot of QZ8501 had asked air traffic control for permission to ascend from 32,000 to 38,000 feet in order to evade the weather. Historically, however, attempting to fly over a thunderstorm has proven a dangerous strategy. In 2009, Air France 447 was flying through the upper reaches of a thunderstorm when it hit turbulence and its pitot tubes froze, leading to loss of airspeed indication; in the ensuing confusion the pilot flying lost situational awareness and flew the plane into the ocean.

107 thoughts on “Why Thunderstorms Are So Dangerous for Airliners”

  1. From pprune forum earlier today:
    ” 30th Dec 2014, 03:12 #475 (permalink)

    Join Date: Mar 2014
    Location: N. California
    Age: 70
    Posts: 123
    Re: timeline
    A regular poster in another forum who works in some aspect of Indonesian aviation safety has stated that from his inside information, the plane went “from 32000 to 36300 and down again to 24000ft or so (in) just over a minute.”

    He hints that the people who have looked at the radar data believe that the plane must have broken up or lost control surfaces, quote: “it ain’t a pretty picture”.

  2. @Chittagong – fascinating, except this is SITA coverage, no? It’s probably not much different than radar coverage, in any case.

    As Victor suggested, we’re really going to need the real radar data released.

    @nihonmama – I’ve always been suspicious when somebody says they can make a profit flying an airplane. In the aggregate, it’s yet to be done, and certainly not on low fares. Southwest made their money on fuel bets, not fares. Sure, there have been good years, but overall it’s harder than it looks to walk in with low fares and make any money.

  3. @JS @nihonmama

    Between different days there seems quite some variation in fr24 coverage.

    For the particular flight, by comparing with nearby flights I tend to agree that 23:18 coincides with limit of coverage of fr24

    Question is if fr24 coverage is sensitive for weather conditions (I think yes). So maybe the same thunderstorm system that caused all the trouble could have influenced signal transmission.

    I’m very curious to know the exact time connected to the leaked ATC radar screen shot where it shows the a/c climbing and on westerly track.

  4. @Chittagong:

    Thanks, it is interesting.


    You would be shocked (but maybe not) at the razor thin margins for the non-low cost (read: big) airlines. Why they now charge for everything (as the service gets even worse). Falling fuel prices can only mean they’re thrilled.


    “I’m very curious to know the exact time connected to the leaked ATC radar screen shot where it shows the a/c climbing and on westerly track.”

    Great point. Here’s the original tweet from Gerry Soejatman, with a load of comments:


    And this is as close as he gets on the time:

    “all I can say is that this was given by someone in the ATC this morning, the relative position to Emirates is right”

  5. My heart goes out to loved ones of ALL these tragedies – including the latest, on which news may be breaking this hour.


    Fugro Discovery (finally) this week completed passes 7 and 8 across the IG’s best-estimate impact point for MH370 – apparently without detecting anything. This extends the width of the area searched from 7 to 9NM.

    I dusted off the impact distribution model I posted Nov.14 to Jeff’s “Sophisticated Hijackers” forum. Its calibration was either tacitly approved, deemed unworthy even of a response, or (most likely) simply overlooked by the IG’s members (@airlandseaman did make one comment, but it couldn’t have been directed at my stochastic model, as it criticized attempts to model a SINGLE path). I re-invite any and all peer review:


    I tried to calibrate it to @airlandseaman’s flight sim work, who summarized his results thus:

    “when the sim observations are combined with the Inmarsat BFO observations, the case for impact within 5NM becomes quite compelling…I shared all this with ATSB and their response was that our sim experience was essentially the same as theirs.”

    IF my model is reasonably constructed and calibrated, and IF the ATSB has searched the “centre of the distribution”, THEN the search has now covered over 80% of the probability density under the “immediate corkscrew @ 7th arc” hypothesis.

    (…assuming impact east of E88. As discussed previously, the E84-E88 region of the 7th arc appears to have been cleverly misrepresented as INfeasible by the ATSB, and as such remains 100% UNsearched.)

    I offer this as another reason to suspend the search, fire the JIT (get 5 new nations), & resume the search.

  6. The Straits Times (11:15 pm 12.29.14)

    BREAKING #AirAsia #QZ8501 Navy spokesman says crew on choppers had visuals of people on the sea surface, not far from debris. (1/2)

    BREAKING #AirAsia #QZ8501 “We are checking whether they people are still alive” says Navy spokesman Manahan Simorangkir to TVOne (2/2)

  7. @Brock

    It is not possible to assign a probability density to the 7th arc. That would require making assumptions relative to the aircraft speed and heading (AP mode). I’ve pointed out many times that these assumptions are baseless in the context of the MH370 scenario.

  8. Brock McEwen:

    As requested, I looked at your spreadsheet. I did not try to dig into the logic, but the radius distribution appears to be consistent with my impressions gained from the simulator experience.

    I hope by now readers understand that we are dealing with probability functions that are not precisely defined. When it was suggested that the post fuel exhaustion turn radius was probably ~5-10NM, and that the priority search width should be ~30NM, this does not mean we know 370 is within that arc width. It means that is the most likely place, thus the preferred place to start searching. It is also true that there is a smaller chance that 370 will be found a little further from the arc, but it makes sense to start looking where the probability is highest and adjust as necessary.

  9. Brock: I rather like the idea of suspending the search (after the search of the high probability areas) to pressure JIT/Malaysia to better inform the search effort and mount a proper investigation. Perhaps this agenda could be pursued even somewhat politely and with some visibility, so that the Malaysians had but no choice but to cooperate and display real interest in locating the remains of the aircraft.

    Chittagong: the Shan and Kachin States along the border with Yunnan see very few Hui, which are actually ethnic Han Chinese. These areas are, rather, populated with former KMT exile families directing the heroin and jade trades, partnered up with various Shan, Wa and Kachin ethnic insurgent groups. The insurgencies have mostly been neutralized, but their guns and the protection they provide for the labs and poppy trade routes have not. These places are indeed rather lawless and formerly counted amongst my favorite in which to run amok, back when I was younger and considerably dumber (thus revealing my inner Dugain by way of rather shamelessly putting myself within the story). I can assure you that there aren’t many Hui running around and that they are rather dull and conservative farmer and trader types as opposed to their more radicalized (and rightfully pissed off) Uighur Muslim brothers. Nope, no Islamists in reasonable numbers in Yunnan, as far as I know. But if you are looking for exiled KMT living disguised as Lisu tribesmen, or perhaps some old fart, puttering around his retirement garden in a faded and tattered uniform of Khun Sa’s Mong Tai Army (RIP), this is the place.

  10. Could the perceived two-hour advance departure be attributed to the consolidation of two flights, one departing at 5:20 and the other previously scheduled at 7:20?

  11. @Rand — You certainly know more than I do about that world. But I was trying to throw a mix of data points in order to quickly (and perhaps hyperbolically) draw attention to the types of…ahhhh…special people who run around these parts and how nefarious the general environment is there.

    And Yunan isn’t the key, necessarily…but the whole north/ northeastern corner of Burma is. Wild, rough and tough.

    Couple that with the radar-less coverage “corridor” up the Bay of Bengal and into that very region and voila — an interesting perspective. And right near the 7th arc. Where bad people do bad things and are internationally intertwined with nasty groups…like triads. Or like radical Sunni Muslim groups (like from Chittagong). Seems interesting enough.

    Anyway, regardless of any specifics, the general perspective was built around the radar-less corridor and the lawless environment. The hypothesis is that this may be a northern route which might mirror some of the more contentious southern routes (Xmas island, for ex)…with varying speeds, altitudes and directions.

    Or maybe not.

  12. Rand:

    According to FR24, all flights for several days before and after Dec28 were scheduled to depart at 05:20am local (22:20 UTC). Reports that the time was changed seem to be confused. I suspect that the time was changed, but changed weeks before, not hours before. Jeff Wise is looking into this posibility.

  13. Sadly enough….I think finding flight QZ8501 on 12-30 with debris field and….bodies, proves that “The System” search & rescue / mobilization efforts all work just fine. Which otherwise proves that, MH370 was no 8-hour “accident”, but a well planned & coordinated effort with many months & scenarios practiced on a simulator, conspired for whatever personal or political reasons, conceived one of, if not the greatest aviation mysteries of all time.

  14. @Mike: re: imprecisely defined distributions: couldn’t agree more. That’s why my post contained clear caveats and limits on its interpretation.

    @Chris: well said. Unless part of a grand-scale campaign of industrial sabotage on Malaysia(n Airlines) – which I’m starting to look at more closely, now – I think the primary “connection” between MH370 and QZ8501 is that the latter will serve as a foil to the former. The clear images of floating lifejackets and slides will strike MH370 passengers’ families as a particularly bitter blow.

  15. Mike, Okay, I’ll step up to the plate here. Based on this Washington Post article:
    … it’s pretty clear that the flight scheduled had been changed at least several weeks before, and only one family failed to get the message and so missed the flight. The impression I’d gotten was that the flight time had been moved up that morning, which of course seemed logistically impossible.
    Personally, if someoneone changed my flight from 7.20am to 5.20am I would not be a happy camper!

  16. @Brock: There are lots of reasons why a plane may fall out of the sky, including failures, weather, operator inputs (errors or deliberate), sabotage, or missiles, since all have caused planes to crash in recent years. At this point, those officially investigating should consider all possibilities, even if not publicly discussed. There may or may not be a link between QZ8501, MH370 and MH17. I can think of some possibilities which I prefer to not post at this point.

  17. Thanks Brock…Yeah, the missing debris field is, IMO..the most vexing aspects regarding MH370, which lends it’s self at times, to the Kazakhstan reversal super spy hijack scenario. I can’t bring myself to believe that. I can’t turn my attention away from Shah along with thoughts of Silk Air & Egypt Air tragedies, but on a highly sophisticated level.

  18. @Brock: I should add that based on what we know today, the weather is the initiating event most likely to have caused the crash.

  19. So far, we have:

    1. A storm several other planes safely navigated – both before, during and after QZ8501’s crash

    2. An oddly explicit and plaintive set of posts from a Chinese blogger TWO WEEKS BEFORE the crash, warning against an imminent attack on AirAsia by the same group that brought down MH370 and MH17. (The first reports I’ve seen of expert analysis suggest these posts were NOT edited after the fact.)

    Don’t get me wrong: it could easily have been JUST the weather, and the Chinese blogger could easily have been JUST lucky.

    But we CLEARLY need to investigate the investigators, here.


    (tianya is apparently akin to reddit in China; “black hand” apparently refers to a powerful, behind-the-scenes force)

  20. Thanks, PhilD. Jeff’s post implied that AF447 made the mistake of trying to fly over a thunderstorm, and I was just trying to point that trying to fly over would be better than trying to fly through, but in fact AF447 tried to fly around.

  21. Bruce Lamon:

    It is not possible to ply over a Cu_Nim T-storm. They are much taller than the service ceiling of anything commercial. Always go around. Climb to avoid clear air turbulence, but never a T-storm.

  22. Everything about this incident is so frustrating.

    @VictorI – Weather definitely a variable. But I still believe the cause of this crash (ruling out the nefarious) is greed. Whoever put this flight at this schedule at this time of storm, tucked in with, what was it, 6 other flights(?) caused this crash. The pilot didn’t even have enough room for tower to authorize maneuvers. Incredible.

    @nihonmama – Saw your link of twitter link of three holding hands – Couldn’t help wondering the possible that if they are floating on water, holding hands, they may have been alive after the crash.

  23. It looks like the debris was all found quite some distance behind the plane on its flight path. Are there any theories as to why?

    My own interpretation from the maps is that a breakup had started before the request for a higher altitude. Does anyone have a clearer picture of this?

  24. Scenario 1.
    Following an encounter with very strong lift and turbulence, the pilot made a climbing left turn to slow up and get out of Dodge. At some point shorly thereafter, there was a partial structure failure and normal flight ended at altitude with some contents/passengers spilling out. Debris will be wide spread. The point of impact for the main debris filed should be 10-20 NM west of the final radar point. Surface debris probably drifted ~50-75NM east of the impact point.

  25. Jeffwise, All:

    FYI, couple of people informed me that airline pilot Karlene Petitt was on CNN (don’t know which program) and apparently said that QZ8501 was moved up two hours because of the storm, not because of a reported schedule change, weeks or months ago.

    I didn’t see this but perhaps others here did. She is on Twitter, so I just queried her.

    If true, an early departure due to weather would be highly irregular from an airline ops perspective, to put it mildly.


    “Couldn’t help wondering the possible that if they are floating on water, holding hands, they may have been alive after the crash.”

    You may be right. The mere thought of that image is beyond heartbreaking.

  26. Mike, as far as I can tell, the height of a thunderstorm varies, and in part depends on what stage it is at, e.g., thunderstorm do not extend as high when developing as when mature.

    According to Wikipedia, cumulonimbus cloud “[p]eaks typically reach to as much as 20,000 ft (6,000 m), with extreme instances as high as 75,000 ft (23,000 m).”

    But regardless of whether it’s possible to fly over a thunderstorm, Jeff’s suggestion that AF447 made a dangerous attempt to fly over a thunderstorm is incorrect.

  27. My frustration is moving to fury!

    From day one I was perplexed that the golden hour passed before ‘missing’ was announced (apparently 1 hr 38 min, in this case).

    My first thoughts were, while you’re ‘wondering’ send up some jets damnit, or a transport and fire off a few drones spiralling outwards or something).

    But what followed next irked more – they stopped SAR at dark first night?! I thought weather had improved.

    But OK, if not, then by the next morning they could go before dawn? Nope, they restarted around 7:30 am. Just a question: Why was it OK they to fly another commercial jet at ~5:30 am when SAR couldn’t fly?

    And finally, at this time, four more bodies reported recovered. The second three bodies reported fully clothed, first three reported holding hands, and the last reported wearing a life jacket.


    Regardless, how about the early “likely at bottom of the ocean” comment before anything was found.

    Whatever follows, I hope hard questions are asked.

  28. Correction – my last post said four more, but three more was confirmed. I read ‘fourth’ body, likely was being referred to as 4/6.

  29. @airlandseaman

    Scenario 1 is the best call imo.

    Only nefarious characters here are hail & wind-sheer.

    My condolences to all as recovery efforts begin.

  30. Chris: To be clear, scenario #1 is just a “best guess” based on scant information and my pilot and met experience. The wind and current data, plus various ADS-B, Radar, debris, wreckage, etc. position reports that have been published over night don’t add up, so we need to wait for the smoke to clear.

  31. I read in one of the Asian rags that one body was recovered with a life jacket on. If that turns out to be true, it has huge implications relative to the sequence of events leading to the plane being in the ocean.

  32. So, I can conjure three questions, with three concomitant potential violations of the public trust re that animal referred to as Commercial Aviation:

    1. Was there Indonesian ADS primary radar data on QZ8501 made available to investigators? If so, why not on MH370? Will we ever learn of it, or will we simply be told that ‘the aircraft was not detected by primary radar,’ or ‘it is irrelevant’ given more informative secondary/ATC radar?

    2. What were the full circumstances associated with advancing the departure time for the flight two hours? For example, were there weather or purely commercial considerations inherent to the change? Is there any causal relation to the crash?

    Most importantly, will the general public ever be informed of such elements found largely in the subjective domain, or will the crash simply be reviewed for technical and human system failures? In other words, if we want to know the totality of he event, will we ever be ‘allowed’ to know?

    3. There wasn’t any distress call while now they are apparently recovering victims wearing life jackets. Was there truly no distress call that can be rolled up into the public aviation rubric of ‘aviate, navigate and then communicate,’ or is there, rather, a bit of reluctance to make public information associated with a distress call? As with MH370, is the general approach to avoid offering information as part of a risk management approach?

    And then ask yourself: have you ever previously been concerned with such matters,or have you generally left the totality of air incidents to others? Do you now see beyond mechanical failure and human error and the location of debris as is so often reported as the sum total of the ‘facts?’

    Shortly after the downing of MH17, I can (barely) recall scribbling something to the effect that the loss of a third commercial airliner in the immediate timeframe could potentially crack the facade of public trust in commercial air travel.

    We are now here.

    At a deep unconscious level, most of us are not aviators. The fear of flying is perhaps a phobia that can be statistically identified (DSM-V), but then perhaps it could also be viewed as a quite natural response to what amounts to flying in a carpeted beer can at barely subsonic speeds at 35,000 feet. This fear, we could say, has been suspended/rendered out of our awareness, and then perhaps even neurotically, in a quid pro quo that enables us to basically show up anywhere on the planet at will. It’s truly amazing that the relative democratization of travel and then jet travel has provided us with what may be a unique historical opportunity to enable materializing virtually anywhere within 24 hours. Who knows how long the present paradigm will last; it could prove unaffordable for most within a decade or so. I personally continue to relish this powerful technology and, as often as possible, playing it to maximum effect, going so far as to seat myself covertly in a friend’s or family member’s dining room, say, just prior to dinner, waiting in the dark until my host, unmet for a year or more, arrives home.

    I am intentionally invoking a bit the woo-woo of romanticism here (I am hopeless in this regard) in an attempt to illustrate the subtle transcendent qualities inherent to commercial air travel. It does operate at this level, and then it comes with a price tag: you must come to deeply trust the glorious safety statistics of commercial air travel if you are going to strap yourself into a beer can with anything less than a cold sweat.

    I am no nervous Nancy, and I have even gone through the exercise on several occasions in the past of terrorizing myself by coming into full awareness of hurtling through space at a high rate of speed. These exercises led to a permanent love of air travel, to the point that I even book extra, unnecessary segments, just for the sheer thrill of it. Call me a freak, call me Ismael for all I care; my childhood aspiration was to be an astronaut, and this is as close as I have ever come to realizing my pre-pubescent dream job.

    And now? I can sense a crack of doubt. On my next flight from Tokyo to Shanghai on Jan 9, I suspect that the old exercise of self-terrorization will have transformed into a compulsion shadowed by wondering what, actually, I have been allowed to know.

    I hope that Jeff covers this subtle existential ground linking three otherwise unrelated air incidents in a future piece, as his book speaks enough to him perhaps being uniquely suited to the task.

    Nihonmama: just passed the New Year hour with family amongst the died-in-the-indigo community of Japanese snobbery that is Kyoto. I love it here. So many bring the practice of art into the everyday aspects of their lives, from arranging pickled daikon on a hand-thrown plate to sweeping out the genkan.

  33. @Dennis, re: “AP Mode”: I’m trying to prove you RIGHT.

    IF my model is well-calibrated, and IF the ATSB is searching the appropriate band, THEN the [AP Mode (fit to minimize BFO error) to fuel exhaustion, no pilot control thereafter] theory is now over 80% likely to be WRONG.

    Of course, zero surface debris had already told us that, but we still LOVE that (faked?) signal data.

    Time to force the JIT to turn ALL its models inside out, so we can verify the authenticity of the entire SEARCH.

  34. @ALSM – your scenario 1 makes sense. I was thinking along similar lines.

    Presumably, the transponder can operate past certain levels of destruction, such as an explosive decompression in the rear of the aircraft. A decompression would probably also keep the crew off the radio for a few minutes, IMO.

    This may be a silly question, but how susceptible is the altimeter to changes in atmospheric pressure in a storm? Could certain fluctuations be attributed to entering a low pressure system? If so, how much? A few feet? A few hundred feet?

  35. @Brock

    Also the fact that nothing has been found in the primary search area does not mean nothing is there. If I recall correctly the 447 search strategy used a Bayesian model to reduce the “probability” that an area searched contained the wreckage, but it (the probability) did not go to zero.

  36. JS: The altimeter is just a pressure sensor calibrated to read altitude. It is connected to the static pressure port(s). The lower the staic pressure, the higher the indicated altitude. I do not believe a local T-storm would have any significant effect on static pressure.

  37. @Rand:

    Happy New Year from over here, 17 hours behind you!

    Love Kyoto too and hey, being the original (and eastern) capital of Japan, a little snobbery comes with the territory. It’s like D.C., but far more spiritual — and cultured.

    Winter in Japan brings to mind bags of hot, freshly roasted chestnuts piled high on vendor’s carts outside the train stations, warm, boisterous izakayas and those delicious, long-legged crab caught up north in the waters near Russia and available in abundance. Hope you’re enjoying the old capital.

    Here’s a Kyoto test: you visit someone’s home and while you’re there, your host asks if you’d like something to eat. What does your host mean?

    And on the subject of aviation (and doubt)—

    Recall way back, I asked what would be the odds that two Malaysian jetliners (the latter carrying the step-grandmother of the Prime and Defense Ministers of Malaysia) were lost/hit, months apart. I got no takers.
    And now we have a third airplane, also Malaysian-owned, and someone appears to have foreseen a tragedy befalling THIS airline.

    Perhaps the blogger in China is psychic.

    As you rightly note, the “subtle transcendent qualities inherent to commercial air travel… come “with a price tag: you must come to deeply trust the glorious safety statistics of commercial air travel”

    What would happen to the commercial aviation industry (particularly international) if people were allowed to know more rather than less?

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.