If You Don't Trust the NWS You Sure Can't Trust Applied Weather Technology - and The Puzzle of El Faro Capt Davidson

Just read the Michael Lewis book The Fifth Risk. It’s a short book about various functions of the government and how people trust or mistrust.

The NWS collects weather data and companies like AccuWeather then use that data. A lot of people don’t understand that. For example Lewis reports that a member of congress has asked why the NWS even needs taxpayer funding as everyone can just use AccuWeather instead.

The NWS has been trying to determine what weather warnings people heed or ignore. Turns out that people with higher trust in the government tend to heed NWS warning while people with low trust in the government are more likely to heed warnings they hear from a person inside their own social group.

However people with low trust in the government will heed NWS warnings if they get them indirectly, that is from a friend that in turn got it from the NWS.

One of the biggest mysteries of the loss of the El Faro is why were weather warnings from the NHC ignored? Capt Davidson instead relied entirely on Bon Voyage data from the company AWS (Applied Weather Service). The Joaquin forecasts from AWS had two problems, they were delayed because they depend on data received from NWS and the AWS forecast had errors which the NHC data did not have.

It’s always been a puzzle why Davidson rejected the information from the mates regarding Joaquin forecasts. Cognitive dissonance, just not what he wanted to hear seemed the most likely explanation.

However Capt Davidson may have assumed that AWS and the NWS are independent and that AWS was simply a private version of the NWS. If Davidson had strong anti-government views very possibly he simply trusted AWS far more then government sources. According to “The Fifth Risk” it’s a very common belief.


Still, being a ship captain and all the experience that entails should have prevented the captain from making such assumptions.

I think it far more likely he feared being fired and was going to go the short (through a hurricane) way or die trying.

1 Like

I don’t think there’s any puzzle or mystery, or trust in the ’ gul-danged guvmint’ to consider. The evidence, and conversations, and actions all point to and heavily support the idea that the Captain devalued the risk and hazards of hurricane weather by equating with his experience of the ship in Alaskan heavy weather, which is not the same thing, and not a prudent position in any event. Since he had no fear of the weather beyond a bumpy ride, he didn’t maintain a guarded approach or conservatively figure his course. So his ‘trust’ in weather source data isn’t a meaningful inquiry, since he was not using it much. He wasn’t questioning it by monitoring winds, pressure and all that Bowditch would recommend, preferring instead his initial plan to shoot under, and making adjustments based on how the ship felt. Speculating on his biases concerning weather provider products is hypothetical at best and a poor study when the amount and quality of info (his own words and documented actions in real time on the bridge) tells all one needs to know about his management and reliance on information products. Were someone to question how he felt or considered AWS v NWS, I’m sure a marine professional like a ship captain would have described that while the NWS collects and disseminates voluminous quantities of data, AWS collects, collates and presents tailored compilations specific and relevant to a client’s situation out of all of it (consider it like a stock broker advising a client versus clients reviewing open source business media in making investment decisions).

If he were concerned about being fired, his conversations would have betrayed a more craven attitude, e.g. dropping a lot of comments about how the ‘office’ was pressuring him, especially when challenged on his plan (however meekly) by the mates (not the CM). He would have done that to build sympathy, rapport and compliant attitudes from his subordinates, but this is not what is shown, he clearly wasn’t a craven individual, quite the opposite. He was making his own calls from that text message on and executing everything he said he intended. If he was forced into it, the first ill weather effects would have strengthened his nerve to take on the office, over the sea. But he didn’t fear the hurricane (enough), and his manner and statements support this conclusion more readily than alternate theories.


The mystery is not why was a high-risk route picked, the expected pay-off of increased risk was minimum cost (no missed ETA, fuel). The mystery is why was a high-risk route picked that was both high-risk and high cost. There’s no pay-off. Meeting the ETA required wind and sea astern, It was a technical error.

The Captain was 100% using the weather routing software.There is a fairly long section on the VDR transcript where the captain and chief mate do in fact use the AWS forecast and software to lay out a new track-line and the chief mate enters the new waypoints into the GPS which the captain can then be heard ordering the third mate to follow.

KC, the key word in your quoting my post is ‘much’. And as I go on to say, he wasn’t testing the info and evaluating course to anticipated conditions. I recall long discussions in the threads about use of the weather info and failure to follow Bowditch in determining his relative position to the storm center. My point is, he didn’t rate the weather as highly as he should have, based on his trust in the ship’s ability to handle it (and anticipating that whatever would happen would end quickly given the belief that they’d pass each other). We could debate his plan, intentions and actions, all of which I think were pretty clear, but I don’t believe his trust in NWS or AWS data is remotely feasible as a contributing factor or even a bias when compared to the magnitude of other decisive human factors relevant to his role in the incident.

Your post introduces a thesis about some concept of ‘trust’ in data he may have had over the mates - who weren’t arguing particularly forcefully on the NWS info’s part. He was using the AWS products, but his actions and intentions and track were established and the data was of low value, again, because he believed in the ships ability to handle the weather. As contributing factors go, the interpersonal dynamics of a Captain and apparently concurring CM against the mates is so far above the potential bias the Captain’s estimation of the value of AWS over NWS would represent. The numerical representation of any quantifiable contribution of such bias (which would be difficult to impossible to establish he had) would be to the right of the decimal point and have several zeros.

There’s been a lot of focus on the fact that the weather info that contradicted the Bon Voyage (AWS) software came from the watch mates. The more salient fact may have been that the data originated from the government.

There’s another data point so to speak, it was reported that Davidson was a domesday prepper. According to Foy’s book the crew reported that Davidson listened to Alex Jones podcasts, the “Sandy Hook is a government conspiracy” guy.

So good chance he might have had strong anti-government views.

Wow, that’s a hell of leap… so you want to make his actions fit some theme and personality type decision making framework. I can’t think of a less productive inquiry in this incident. Its a fairly safe conclusion that Captain Davidson did not take all the prudent measures one would expect of a ship captain in such a situation, on the planning and execution aspects. It is also fairly conclusive at least in my estimation that he was a confident individual, more like over confident. There is not much point digging further into the psychology once you hit this point. Certainly the doomsday preppers and Alex Jones fans Ive encountered are similarly ‘confident’. So chicken or egg? One more bit of data suggesting his personality would reject non conforming information based on self-assuredness? Throw it on the pile but there’s enough to suggest such confidence was on display. If you’re suggesting that he had a latent mistrust of government (which seems to go hand in hand with an over developed trust in specious information sources), I would think such a personality type would also fit the profile of a type A leader, who certainly wasn’t going to be led around by young ‘uns or change his opinion once set on a course of action. Estimating where a possible distrust in the government product —-like NWS hurricane warnings and track forecasts that literally state that the product couldn’t be trusted completely and with levels of uncertainty— would fall in the spectrum of factors in his discussions with the mates, which do you think would be on top if he was this kind of person on the inside. What kind of analysis would you need to make a conclusive point about the weight of such a factor? Seems a useless and page filling bit of analysis, a salacious detail to what seems the simpler and more obvious answers. He had a plan of action and degraded/devalued contrary indicators. The end.

It’s a useless inquiry because it has no practical value to the point of investigation of the incident—since it would be unique analysis to one person’s mental states, hypothetical at best and not likely to inform any useful future actions to prevent similar incidents. Captain Davidson’s deliberate acts and words are the matters to be analyzed and considered, evaluated and judged for our own actions in the future. And that analysis will produce the same answers that were around before—if you anticipate heavy weather and tropical cyclones, read your Bowditch, read the skies, and stand a taut watch.


No big leap, incremental.

The mystery from the beginning is why sail into the north eye-wall of Joaquin? It never made sense. The answer that it was a high-risk move to save time doesn’t make sense. Why not at minimum steer for the middle of the eye? High-risk but more favorable winds.

Most likely the captain was mistaken as to where the eye was. This is apparent on the VDR transcript as well. By all indications they believed they were south of the eye. At one point, close to the end of the transcript the capt does mention that there is disagreement in the forecasts.

Not sure what you’re saying here. The AWS product is subject to all the same uncertainty as it is based on NWS data. In hindsight it had far more serious errors. I don’t recall any substantive discussion about forecast uncertainty.

It’s a probability thing. That’s the point of the OP. The NWS is not trying to guess one person’s mental state. They did research on what factors cause people to respond or not respond to weather warnings. One of the findings is that for some people sources that they believe are non-government are considered more credible.

The OP used the words ‘mystery’, ‘puzzling’. My point is there is no mystery, no puzzle. The evidence record supports conclusions that can be (were) drawn from the storm track and the AIS track without trying to fit one man’s actions into a spectrum of general public use/trust of data on weather especially for specialized consumers of such info like mariners and I would add pilots. It adds nothing to the case investigation. Similarly it wouldn’t support study thesis as a data point since it would be speculative.

1 Like

Something a man with his background and position would or should have absorbed and internalized long ago.

1 Like

Seemed like to me he just relied completely on the Bon Voyage forecast, and forged ahead till the weather overwhelmed them. He clearly trusted Bon Voyage, and figured the steadily building weather wasn’t a big deal because he had seen worse in Alaska, or so he thought.

Overreliance on what Bon Voyage says is quite common, from what I’ve seen.

I sailed with Davidson in 2012 as 2nd mate (having sailed as Master previously in another program). To the best of my recollection, he WAS NOT anti-government.

On 02 OCT 2015, when I found out that he was the master of El Faro, I was quite surprised, and saddened, for I had not observed any incompetence in his shipboard actions. Indeed, he struck me as a good and conservative shipmaster. And when I say conservative, I don’t mean politically, I mean, if I may borrow an idiom from the aviation community, “There are old pilots, and there are bold pilots, but there are no old and bold pilots”.

1 Like

Just a guess here but since the company paid for its weather routing/forecasting service the captain may be inclined to “trust” it more than something provided free. Even though it is based on NWS data.

1 Like

There is that. Also a obsolete looking terminal spitting out simple text messages with dot matrix printer compared to nice slick software package displaying beautiful interactive color graphics.

But when ECDIS came along almost all the captain I talked to said they didn’t trust it. In spite of the fact that the mates have been steering to the waypoint using the “highway” on the GPS receiver display for the last 15 years or so.

Maybe weather info seems less critical than nav info. Maybe a little of this and a little of that.

1 Like

2 posts were split to a new topic: Use of ECDIS at Sea

From the VDR Transcript, this is the Capt at about 01/0503, about an 1 1/2 before the end of the transcript.

That– that’s fine– but here’s the thing– you got two G-P-Ss– you got
five G-P-Ss– you gunna get five different positions. you got one
weather program (and I use/and use) B-V-S and that’s what I (sent) up
here * we’re gettin’ conflicting reports as to where the center of the
storm is.

B-V-S being the Bon Voyage weather program from AWT (Applied Weather Technology).

The data path is :

NWS > SAT C > SHIP - by text
NWS>AWT> EMAIL>SHIP - by B-V-S computer program

Either way the data source used by B-V-S for the location of the center of the storm originates with NWS, If B-V-S conflicts with NWS the error has to be with B-V-S.

The conflicting reports could be resolved by comparing NWS report with an actual weather observation (wind direction and barometer).


Figure 42 plots the information available to the bridge team around 0500 the morning of
the accident, after the captain downloaded the previous evening’s BVS file that had been available to him at 2304 and just after the new weather information delivered by Sat-C was available on the bridge (at 0446). At that point, the BVS information (blue) was almost 12 hours old (6-hour latency, as described above, plus additional time between receiving the email and downloading the data). The Inmarsat-C information was current as of 0446.

The captain stated on the VDR at 0503, “We’re getting conflicting reports as to where the
center of the storm is.”

The Fifth Risk is an important book and everyone should give it a read. Michael Lewis is good at picking up stories and telling them in a compelling way.

I read through this whole thread with the following thought: The ship wasn’t lost due to the weather exactly. The ship was lost primarily because she lost propulsion. When her engine failed she was no longer able to hold her head up to the wind, she fell into the trough, began rolling heavily and capsized.

Of course the weather was a factor, but I’m not convinced that was the fatal factor. If they hadn’t lost engines I doubt the ship would have been lost due to the weather by itself. I’m sure the captain didn’t include losing engines in his plans. They never talked about the possibility. How many times have we crossed a busy street on foot and allowed for suddenly breaking an ankle?

I went through a hurricane in a 120’ tug towing a barge once and the thought was uppermost in my mind. We were hove to, nose to the winds and I told the chief “if we lose engines we’re fucked, you know that - right?” He nodded.

And the Hindenburg blew up due to random static electricity, not the use of hydrogen… El Faro was lost due to losing propulsion in the weather environment in which she found herself, and since she found herself there by the decisions of the master, well…

Causality, and determination of root cause has to be practical in determining where the ability to intervene in future incidents may be identified. You say because she lost propulsion, perfectly valid opinion, just as perfectly valid to say she didn’t sink due to loss of propulsion, but rather the lube oil tank level was too low to sustain in heavy seas. Or perhaps the quality of the cargo securing, the design of the seawater piping system in the hold, or the scuttle which may have been left open… the point is, with a sufficiently hardened target, the weather would have been less relevant to survivability on a case by case system basis. But the proximity to the storm increased risk, universally, to the multiple systems resulting in the issues noted in the investigation report. So this control of proximity becomes the one independent variable best managed to prevent the incident touching on multiple systems. The alternative is to manage so many other variables, some incapable of adequate control, as to render causality a futile discussion.

Consider it like a lightning bolt. Wherever the lightning hit WAS the path of least resistance to ground for that lightning bolt. And while you could examine and theorize about the various paths available and blame the low resistance home TV antenna or the toilet vent on the roof or the decorative wind vane, the area lightning storm was the overwhelming issue and once enough electrical potential developed, a path would present itself and burn down the house. All the other details are just minutiae. Important details to consider for an investigation sure, because hardening all the systems to vulnerability is a useful and productive endeavour but, the prime factor was undeniably the weather, once the vessel was in the grasp of the storm effects.


I am mindful that whilst peeople argue over the cause of the tragedy, it is still a tragedy and there are grieving families.

Does this debate and the tone of some posts help them in any way?