Atlantic crossing

[franksingleton]

Is the data complete for 1900-1920 though? I'd be surprised if coverage was sufficient to spot and report every storm over the atlantic. Hurricanes probably fine, of course.

A good question. They are as provided by the US NHC and published online at North Atlantic Ocean Statistics compared with climatology. Off season storms are listed by Wiki at List of off-season Atlantic hurricanes - Wikipedia. I never fail to be surprised by the numbers reported in the mid 18th century before routine observing was coordinated internationally after the first international meeting of maritime nations to discuss weather.


I can imagine that on any one day the data would be incomplete, however, these are big enough to have lifetimes long enough for them to be reported. By 1900many ships were producing weather logs that were sent to their national weather service. Information from logbooks was shared between all maritime nations. Most will have been keyed to computer some years ago. At least, they were for the UK to my direct knowledge an would have been to most other countries.

So, in answer to your query, I think that missing data effects would be small.

 

[franksingleton]

I have kept my boat in the Azores since summer 2014 and have seen two hurricanes in the area. One out of season, Jan 2016 Alex, downgraded to tropical storm about 100 miles south of Terceira. Oct 2019 Lorenzo passed near to the western islands inflicting considerable damage to Ilha Corvo and its port.
Despite an active interest in met (a subject I taught to MN officer candidates) I had never picked up on the Azores being threatened by Hurricanes.
This page lists Tropical Storms and Hurricanes near the Azores by decade from the 1990's:
List of Azores hurricanes - Wikipedia
1990's = 6
2000's = 4
2010's = 11
It seems that potential storm tracks may be extending over a wider area than older records would suggest.

I cannot remember where, but I saw a paper saying just that, ie tropical storm tracks were extending further north than previously.

PS. Added later.
Dredging the deeper recesses of my mind, I do recollect in about 1977 one of my senior forecaster colleagues being concerned about a tropical storm approaching the Azores from the south. I had to convince him not to start issuing warnings for the UK. It was obviously going to fizzle out.

 

[lustyd]

I'm surprised it was that comprehensive given the dates since the logs would have been useless by the time anyone got to see them. I guess for record keeping and weather prediction it would have been helpful though so it makes sense that they were putting effort in to record in as much depth as possible. It does rely on a ship being present though, whereas we now cover the whole Atlantic regardless of occupancy so it's entirely likely that some were missed or unreported.
Given the low numbers to start with I will disagree with you and say I think it's quite a significant gap and I wouldn't draw conclusions based on it. I will agree that it's probably changed, and probably in the way we think, but that's intuition rather than good data. Unfortunately the first step in any data science is qualifying the data, some of my customers spend millions quality assuring data every month to avoid being misled.

 

[capnsensible]

Digging around on this site is interesting....if its a subject that interests you!

Hurricane Archive | Weather Underground

One of the guys who did/does? Write for them spent several years flying on the hurricane hunter aircraft that flies through trs to gather data.

 

[franksingleton]

I'm surprised it was that comprehensive given the dates since the logs would have been useless by the time anyone got to see them. I guess for record keeping and weather prediction it would have been helpful though so it makes sense that they were putting effort in to record in as much depth as possible. It does rely on a ship being present though, whereas we now cover the whole Atlantic regardless of occupancy so it's entirely likely that some were missed or unreported.
Given the low numbers to start with I will disagree with you and say I think it's quite a significant gap and I wouldn't draw conclusions based on it. I will agree that it's probably changed, and probably in the way we think, but that's intuition rather than good data. Unfortunately the first step in any data science is qualifying the data, some of my customers spend millions quality assuring data every month to avoid being misled.

The routine collecting of ship’s logs began after the 1855 conference but many logs have been acquired from before then. The intention was not daily prediction but the compilation of a marine climatology. A great deal of effort has gone into data analysis and re-analysis of historical weather data. I would expect data coverage to be pretty good by 1900, which is why I looked at the period 1900-1920. Much before then, there must be some doubts although NOAA seems quite happy to use them. Few storms would have been missed although details would be increasingly uncertain the further back you go.As I said, the lifetimes of these systems is sufficiently long for analysis purposes. No doubt positions and depths of many hurricanes would have been deduced using conventional chart analysis.
I started my career in meteorology before satellites and data buoys. Don’t underestimate the analysis skills of those who did the job in those days. Remember they were thinking in space and time. A chart analysed at a particular time would have included short term predictions from the previous chart. We used to draw a chart using a light table so that you could see the previous analysis and data. I doubt that many of today’s breed of forecasters have those skills. But, of course, they do not need them. The computer is king and weather analysis techniques are highly complex as are the prediction models.

 

[lustyd]

although NOAA seems quite happy to use them

It's fine to use them, as long as you know the limitations. The light box method you describe is similar to some data science techniques, although not using paper and pen, the locations add weight digitally over time. The difference is that your method was 2 dimensional while the modern computer based ones might have hundreds of dimensions which precludes the use of pens to sensibly calculate.
Although there may be gaps, the data that is there is perfectly valid to add to the weighting of whether a storm is likely in January. The difference is that a lack of data doesn't confirm that a storm is less likely in February since we have low certainty that missing events did not happen rather than weren't recorded. Knowing what is useful for what purpose is a whole career, as is designing the systems to allow the data to be used or not used, more so now that computer models are so complex that humans wouldn't spot the problems. In fact these days, stopping people misusing data is a bigger industry than analysing data, a side effect of lowering the bar for entry through ubiquitous compute power. Back in the day when compute was expensive us normals weren't able to see the data let alone do stupid things with it

 

[Roberto]

I want to leave when the weather is right not because a rally says we leave on this date regardless of the weather. On an Atlantic crossing, the only weather you can choose is the weather you leave in. Far nicer to have perfect weather for the first few days to bed into the way of life before anything nasty pops up.

Yep, it goes together with the November "early" departure date; one does not even need a storm: on one edition they left with fresh SW, very uncomfortable sailing beating to windward for days since the very start, a number of boats came back a couple of days later...
I think no one in his right mind would leave for a transatlantic having to immediately beat for days knowing that if they waited a week or so conditions would have been perfect to start with.

 

[geem]

Yep, it goes together with the November "early" departure date; one does not even need a storm: on one edition they left with fresh SW, very uncomfortable sailing beating to windward for days since the very start, a number of boats came back a couple of days later...
I think no one in his right mind would leave for a transatlantic having to immediately beat for days knowing that if they waited a week or so conditions would have been perfect to start with.

We have left a couple of times in November. It's not the month of departures that's the issue, it's the forecast on the day and for the next 4 or 5. By then you can be at the Cape Verdes where you get steady trades usually.
We left late December last time and had light winds for the whole trip. Friends were a week ahead of us and had pretty consistent strong trades nearly all the way. Like I said in an earlier post choose a nice forecast, it's the only bit of the trip where you can choose the weather

 

[franksingleton]

It's fine to use them, as long as you know the limitations. The light box method you describe is similar to some data science techniques, although not using paper and pen, the locations add weight digitally over time. The difference is that your method was 2 dimensional while the modern computer based ones might have hundreds of dimensions which precludes the use of pens to sensibly calculate.
Although there may be gaps, the data that is there is perfectly valid to add to the weighting of whether a storm is likely in January. The difference is that a lack of data doesn't confirm that a storm is less likely in February since we have low certainty that missing events did not happen rather than weren't recorded. Knowing what is useful for what purpose is a whole career, as is designing the systems to allow the data to be used or not used, more so now that computer models are so complex that humans wouldn't spot the problems. In fact these days, stopping people misusing data is a bigger industry than analysing data, a side effect of lowering the bar for entry through ubiquitous compute power. Back in the day when compute was expensive us normals weren't able to see the data let alone do stupid things with it

Prior to WW2, yes, analysis was only at the surface so, I suppose was 3D, X, Y, T. After WW2, we had radio sonde plus aircraft, so it became X, Y, Z, T although the z component was very crude. Modern computer analysis is genuinely 4D. The big problem is that satellites do not measure temperature and humidity as some would have you believe. They measure the effects of temperature and humidity on absorption of infrared, on microwave emission and on radio propagation. The equations cannot be inverted, therefore, as ever, there is always uncertainty in the data analyses as well as in the calculation of all the physics that drive the atmosphere. In fact, it is amazing how well models do and, equally, how few major errors there are even a week ahead.

 

[geem]

ARC or NARC I’m amazed (bemused?) by how ill prepared some boats are that sail across oceans. From the trivial incidents of boats arriving in Gran Canaria and discovering they should have thought to have onboard a plank or something to use as a passerelle to the boat we saw arrive in Barbados without a Q flag or any pilot book and didn’t know whether they were in a Port of Entry or not.

Yep, they are the ones that normally join the ARC
They join them all together with bungee cord

 

[mattonthesea]

Yep, it goes together with the November "early" departure date; one does not even need a storm: on one edition they left with fresh SW, very uncomfortable sailing beating to windward for days since the very start, a number of boats came back a couple of days later...
I think no one in his right mind would leave for a transatlantic having to immediately beat for days knowing that if they waited a week or so conditions would have been perfect to start with.

2010: AFAIR two SW gales before the butter melted and then becalmed. Several returnees. Correct me if I'm wrong.
I left 22nd Dec; 24 day crossing to Grenada.

 

[capnsensible]

Isn't ocean sailing fun?

 

[zoidberg]

It seems that potential storm tracks may be extending over a wider area than older records would suggest.

Tell that to the New Zealanders.....

 

[geem]

Tell that to the New Zealanders.....

Friends there with boat on the hard. Just gone down to check if it's still there. Power been out all night

 

[zoidberg]

You know you're in trouble when the Council's 'No Parking' signs are washed away.

 

[Mudisox]

Am I alone in enjoying the "hassle" of dealing with officials in ports, ensuring that the boat is properly prepared for the trip and having an enjoyable and safe sail.

Rather the whole point of being a 'skipper', not with gaining bits of paper, and paying a lot to some organisation to do it for you.

 

[dunedin]

Am I alone in enjoying the "hassle" of dealing with officials in ports, ensuring that the boat is properly prepared for the trip and having an enjoyable and safe sail.

Rather the whole point of being a skipper, not with gaining bits of paper, and paying a lot to some organisation to do it for you.

Perhaps you are unique in enjoying the "hassle" of dealing with officials in port.

But not in terms of being thorough in preparing boats and crew for voyages, and having an enjoyable and safe sail. Anybody who has been on an ARC would know that the skippers and crew spend a LOT of time preparing their vessels and crew for the trip, and benefit from extensive documents giving great advice on preparation, together with detailed inspections prior to harbour. Arguably these are some of the best briefed and prepared boats crossing the pond. But this wont cut much ice with the people (generally who have not participated) who look down on the ARC.

It won't make much difference anyway. I believe already all places are fully booked for at least a year if not two years ahead, in spite of adding two extra start dates (ARC Plus and ARC January), so plenty seem to welcome the service they provide and the hugely positive feedback from most previous participants encourages more to participate.

 

[geem]

Perhaps you are unique in enjoying the "hassle" of dealing with officials in port.

what hassle? Dealing with officials in the Canaries or Mindelo before an Atlantic crossing is insignificant compared to the officialdom you will need to endure in the ex British Islands in the Caribbean. Trinidad, Grenada, SVG, Antigua are all a chore. You get used to it as part of the cruising life here. Head to Curacao in the ABC Islands and you would think they really don't want you there. A cruising permit needed for each anchorage with a 3 day maximum stay. No customs and immigration a the main anchorage of Spanish Water. You have to go in to the main port of Willemstadt on the bus then walk customs. Then cross the river and walk another mile to immigration and the harbourmaster. Allow a day for clearing in and another for clearing out.

 

[Kelpie]

One of the reasons we picked a French island to make our landfall...

 

[geem]

One of the reasons we picked a French island to make our landfall...

Unless you are planning to stay in French Islands you will have to deal,with it sometime. Its just part of the process of cruising the islands. The first time we came across with a dog we went to a .French island for the same reason. Now we understand how it all works we just arrive in Antigua since its our favourite island. We have lots of friends here. The kitesurfing is great. We have a crew spot for the Classics race. We will watch the start of the Caribbean 600 on Monday. There is racing going on this week and the round island race. Lots to do here but we love the variety that the islands offer. If want to hike, we go to Dominica. If you want to dive, go to Bonaire. There is an island for everybody