An increase in the frequency of clear air turbulence doesn't necessarily entail an increase in reportable accidents and incidents. The NTSB is only notified when a specific set of criteria are met. See 49 CFR Part 830 for details. If the increase in turbulence is all light to moderate turbulence with no serious injuries, there's nothing to report to the NTSB.
This deserves to be the top comment. Turbulence accident statistics are only going to ever reflect clear air turbulence if the aircraft sustained detectable damage or passenger(s) sustain serious enough injuries.
In that case what’s the explanation for clear air turbulence events having an exactly proportionally lower rate of causing damage and injury such that the rate has remained flat despite the increase in events? What’s making turbulence safer?
Edit: this isn’t a rhetorical question. I’m very interested in any proposed actual mechanism. As someone who is very alarmed by turbulence I’d love a reason to believe it’s getting safer.
As I understand it modern US airlines (Delta in specific) employ a meteorology team tasked to predict turbulence and they also run complex turbulence tracking systems that allow one plane that detects turbulence to communicate to a centralized system that allows other planes to change altitude or heading to avoid turbulent areas when possible.
Therefore both can be true at the same time: turbulence events are increasing, but we are also getting better at predicting, avoiding, and dealing with these events.
Flight plans are now algorithmically generated to shave fuel usage.
Because weather forecasts are more accurate, the algorithms write flight plans that take the aircraft closer to storms. This saves fuel while slightly increasing the risk of severe turbulence.
Couple that with the lightweight materials used in modern aircraft, and passengers are likely to experience more frequent moderate to severe turbulence.
There's no reason there couldn't be a statistical increase in turbulence without a statistical increase in accidents if the intensity hasn't crossed a threshold for the amount of turbulence todays planes can safely sustain.
Also, they just looked at 2 years, so there could be cherry picking. Jet stream is affected by el nino. 1979 was weak el nino, 2020 was moderate la nina.
Why do you say that? Just looking at the plots in Prosser et al (2023) it is obvious they had a lot more years than just two.
If you actually read the short paper you'll find they actually used reanalysis data sampled at a rate of every three hours across 42 years to compute their statistics:
> Global ERA5 reanalysis data (Hersbach et al., 2020) [...] were extracted on the 197 hPa pressure level with 0.25° horizontal resolution at three hourly intervals from 1 January 1979 to 31 December 2020. To allow the computation of CAT diagnostics that require vertical derivatives, fields on the 188 and 206 hPa levels were also extracted. The 21 turbulence diagnostics were then calculated from the extracted reanalysis fields every three hours.
How you go from that to "they just looked at 2 years" is beyond me.
Right, and those two years are generationally distant. Another question I have is if pilots have a stable mechanism (that is, an unchanged objective sensor or something) that records the CAT or if it’s recorded by pilots, whose sensitivity to CAT might differ over time. Didn’t mention in the article how it’s measured.
CATs are recorded in pilot reports using terms like "light", "moderate", "severe", and "extreme" which each have a definition. I.e. severe is "Occupants are forced violently against seat belts or shoulder straps. Unsecured objects are tossed about. Food service and walking are impossible." I think only severe and extreme turbulence need mandatory reports and the lower two levels are a bit more subjective ("Food service and walking are difficult")
Newer planes have sensors to measure eddy dissipation rates which are an objective measure of turbulence but I don't know how widespread those systems are and whether they get reported anywhere. They're mostly used for long distance transoceanic flights.
Unsecured objects being tossed around has no real meaning, it’s about controllability of the aircraft.
I’m a pilot and it’s been a while since I went over PIREPS but generally severe is rarely used, severe means the turbulence is so bad you can no longer control the aircraft. What most passengers imagine as severe is probably light turbulence. Most of the time it’s not even reported. As a side note if you’re ever on an aircraft and not secured at all times, you’re making a huge mistake.
Have there been any reports of true severe clear air turbulence (where the pilot cannot control the plane) or are all of these cases not technically severe because the pilots were in control the entire time and it was just a bumpy ride?
Another question I have is what do you do in that scenario if you can't control it? Just ride it out and hope for the best?
You don't have to live in fear of turbulence when flying, just keep your seatbelt on when you're seated. Turbulence is fairly rare but it's still a numbers game. The probability that you experience it the 99% of the time you're seated is much higher than the probability of experiencing turbulence while standing, especially since pilots proactively turn on the seatbelt sign when turbulence is expected.
Certainly no detectable trend in that data. But the accident frequency is so low that the random variation dominates and makes it impossible to distinguish any trend.
What is demonstrably increasing is CAT, due to climate change. But considering how infrequent these incidents are we might not see a clear increase for several decades.
Do we really not record turbulence sensor data off the airplanes and download it when they’re on the ground?
I’m also surprised that these airplanes have on demand satellite TV streaming to these airplanes but airlines claim that it costs 100k to add that to existing planes. There’s just no way it’s 100k per plane - there must be a cheap way to retrofit the data without having it be reliable since it’s opportunistic. And heck, France is doing it every 4 minutes for their planes so why can’t Americans figure out how to do it.
> Do we really not record turbulence sensor data off the airplanes and download it when they’re on the ground?
No, accelerometer data is only recorded to the FDR. Which has a limited storage window (1-24 hours depending on the aircraft) and is slow to download requiring moderately specialized equipment and a technician to carry out the task. Aircraft downtime and technician hours are both expensive and in short supply.
> I’m also surprised that these airplanes have on demand satellite TV streaming to these airplanes but airlines claim that it costs 100k to add that to existing planes. There’s just no way it’s 100k per plane - there must be a cheap way to retrofit the data without having it be reliable since it’s opportunistic. And heck, France is doing it every 4 minutes for their planes so why can’t Americans figure out how to do it.
Everything on airplanes is expensive. Even cabin amenities. You have to prove it won't start a fire, was installed correctly, won't interfere with other equipment, won't interfere with the aircrafts structure, and again requires technician hours and aircraft downtime.
> No, accelerometer data is only recorded to the FDR. Which has a limited storage window
Apple and Google could fix this my streaming accelerometer data to the ground when people are connected to in-flight wifi. It is fairly easy to identify which phones out of a set are the stationary ones.
I find it hard to believe that the headsets they are using for software and hardware meet that level of criteria. They’re clearly using off-the-shelf parts. Some amount of care is called for sure, but 100k to apply a software patch or tweak the tech in FDRs which are swappable and upgradable? A flight recorder is 10k. You can’t tell me it costs 90k to install a new one capable of sending data over the satellite link in bursts. Clearly other countries and airlines with a similar safety record and cost of living and salaries are able to accomplish the feat.
You’re solving a human problem with technology. Pilots are resistant to data collection because of the proven track record of airlines using it against them. Unintended consequences are fatal in aviation. Saying “it ought to be easy” is an immediately disqualifying statement. You should ask instead why it is so expensive. Then decide if there’s a margin worth eroding.
What human problem am I solving with technology? I’m just saying we should have public data about turbulence so that we can understand changes to the jet stream. I’m not talking about making planes safer or solving human errors. I’m not sure how this data could possibly be used against pilots. I’m not talking about recording the cockpit; just the sensor data about what the plane is doing and experiencing.
> You should ask instead why it is so expensive.
That is literally my question. I’m highlighting that 100k seems really high to make a system that opportunistically transmits data we are already capturing locally. Rather than a flippant “airplanes should be expensive”, why not ask what is the cheapest retrofit we can do that doesn’t change the safety profile. As I said, this system should not be in the critical path and shouldn’t be a required other than the airplanes should generally be maintaining it to be functional (i.e. the SLA can be 75-90% and still provide tremendous value instead of the 100% SLA target for flight critical components which is what that 100k price tag sounds like).
That this is something highlighted by crash investigators as something that would help in corner cases like incidents over the ocean is just gravy.
Meanwhile an accelerometer, microcontroller and a years worth of storage for the data probably all could be made with $5 worth of parts and a summer intern...
Oh come on. It's not like raspberrypi is going to force plane down.
Also there's likely 200 accelerometers on board already. Onboard wifi is becoming ubiquitous. Perhaps an app that trades in wifi time for accelerometer time would be good trade-off. And wouldn't require tons of certification.
Well, you have to have paperwork claiming it was installed correctly. You don't _have_ to install it correctly. See e.g. doors installed so correctly that they blow out in flight etc.
What's the point of your comment? Should we just install a bunch of accelerometers on planes with faked paperwork because some people faked some paperwork somtime?
Sounds like you're upset at Boeing and figured you would tell us you're upset on an unrelated thread. Note that it doesn't really matter if you are right to be upset at Boeing or not. It's still unrelated.
I think his point is that the proof is expensive, not the act itself. Reminds me of rivets in composites joined by adhesives. The benefit is inspectability. The cost is diminished strength.
Some aircraft are equipped with a system that records parameters in flight, and sends it via 4G when on the ground. This is used for preventive maintenance mostly. It’s a service that’s more and more common.
Yeah I feel like the FAA should require the airlines to share some of the data so we get research into the jet stream that we can cross correlate with other data sources.
There's essentially no cheap way to add anything to an airplane. Most changes will require extensive testing, verification, and sometimes certification when it comes to planes.
But anyway, where are you seeing a claim that it costs $100k to record and save turbulence sensor data? I don't see anyone upthread claiming that, and the article doesn't touch on it at all.
I read it in another article on the topic researching before commenting. Just didn’t bother to cite it but that was the official reason given as to why America doesn’t do this.
If satelite TV is down, some passengers are mildly annoyed. If a regulatory required part of the aircraft is non-functional, you're not going flying today.
You’re saying it’s impossible to have an optionally required feature? If the satellite TV is down some sensor data isn’t sent. Why would that be cause for grounding the plane? You could easily make the regulation an SLA like all routes flown must be sending data for 90% of the flights on that route for the month and failures to meet the SLA are investigated.
Also I’ve flown a bunch and I’ve rarely seen the Internet link go out except where there’s technical limitations like crossing the ocean where they can’t maintain an internet and have to rely on preprogrammed content. Given how much money they make from cabin internet, the airlines are clearly incentivized to apply pressure to keep those things running. I doubt I’ve seen anyone be really annoyed when there’s technical difficulties. Most people who fall into that category would have made other arrangements for entertainment anyway.
There is a big factor out there that is 'masking a rise in incidence of CAT from accident stats'. It's Aerospace Engineers and the aviation engineering and safety community.
--
Most CAT events (seen so far) are survivable by current aircraft designs, so you can have an increase in CAT without a spike in crashes and other accidents.
Could be an inverse relationship, too - the more frequently they occur the more experience, training and guidance the pilot, and the other crew members get to manage it: how to control the the airplane, urge passengers to wear seatbelts more, etc.
The article seems to say that other than direct passenger injury, the issue is premature airframe fatigue, which I guess that if remarked on inspections, does not end up in the incident category.
Yeah, you should have more concern over human error and Boeing than this. But, boy, the more I have flown and the older I am, the more I get anxious during turbulence when I fly.
Thank you! That's definitely a great way to look at it. Complacency and habit breed accidents. Although not the the Boeing/mechanical issues. Still worried about the ghost in the machine or gremlins!
> Several booked passengers cancelled their tickets at the last moment to see a ninja demonstration. These passengers, Albert R. Broccoli, Harry Saltzman, Ken Adam, Lewis Gilbert, and Freddie Young, were in Japan scouting locations for the fifth James Bond film, You Only Live Twice (1967).
The James Bond franchise would likely have been quite a bit different had Broccoli died in 1966. Crazy to think he and his colleagues cheated death because they wanted to play tourist a little bit longer than originally scheduled.
>The Boeing 707 jetliner involved disintegrated mid-air
I'm guessing their use of "disintegrated" there is supposed to be taken literally as dis-integrated, but upon first read, I took it for its more colloquial meaning (which to me is closer to pulverized, turn to dust, dissolve etc).
Well, assuming that there is indeed more turbulence, that could also mean more vigilance against accidents, so that could even itself out as safety regulations get stricter than in the past. Also, this is mentioned in the conclusions:
> The report includes an important discussion of the risk to unrestrained occupants onboard aircraft, including flight attendants – who account for nearly 80% of those seriously injured in turbulence-related accidents. Key recommendations in the report are intended to help ensure better protections
for flight attendants
... which makes me think of two more possibilities:
1 - I suspect any careless flight attendant involved in a turbulence-related accident would learn their lesson after the first time, and take better safety precautions. Perhaps that is a stabilizing factor on the number of accidents, since the number of flight attendants who need to learn that lesson the hard way is probably more a function of how many new flight attendants enter the field than it is a product of how much turbulence there is.
2 - Flight attendants under-report minor accidents so they don't get into trouble for not respecting safety rules
Of course, this is pure speculation (and assuming that the premise of there being more CAT incidents holds up), I'm sure the actual document goes into this kind of thing in more detail but I don't have the time to dig through 115 papers.
> I suspect any careless flight attendant involved in a turbulence-related accident would learn their lesson after the first time, and take better safety precautions.
They are briefed, no need to learn their lessons after the first time. However, it's part of their job to walk around the plane (eg to serve food), and so they're less likely to be seated than pax. That is the (rather obvious) explanation for the fact that they constitute a very high proportion of victims, not "careless"ness.