Course to steer software

[Dockhead]

Originally Posted by marcvilleneuve

A straight line is a straight line; is it over the ground, the water or the air. My approach works well between two marks that are a few miles from each other with the assumption that the direction of the current will not change. You right it means that to maintain a constant COG I CONTANTLY have to adapt my heading according to the strength of the lateral current so COG=Bearing to next mark.

By the way, there a very easy way to achieve that same straight line when a visual on the mark is available. Just keep a constant alignment between the mark and a background feature by modifying the heading. Works great!

Many people discussed crossing a channel here and I think that it is a different and more complex issue better managed by a routing software that will take wind as well as current into account.

Good racing!

In short, your approach works well when you can achieve a constant heading, by following a constant COG! But that works ONLY when the current is absolutely unvarying, either in speed or in direction!


Just remember: constant heading = straight line through water. Constant COG = straight line over ground. These two things are as different from one another as STW vs SOG! So what you are doing is as if you were sailing to SOG, not bothering to clean your speed transducer. It WILL work so long as the water is absolutely still! But I know few racers who would be happy with such an approximation.

[Dockhead]

Originally Posted by Adelie

You can’t do it for waves, I don’t know of any really good prediction software for that.

For wind there are decent predictions for most coastal area and if you have your polars a properly programmed computer could pick a course.

Current predictions in tidal areas are also pretty good and a computer could do this.

Given the computer would have to deal with both wind and current Thera probably not a straightforward analytical solution so the computer would have to simulate a bunch of courses to find the quickest.
Given varying wind and current the computer may find a multiple leg heading that’s the fastest provided the forecasts are accurate.

I bet somebody is working on this but it’s not a trivial task.

Wind and changing currents are two different layers of this problem.


The first job is to find the course to steer to get you to the destination in a straight line through water. Only then do you think about wind and sailing aspects. Yes, it can be compliated and iterative IF there is tacking involved (and if you really want to get into the depths of it, read the old threads we did on the lee bow effect). But if there is no tacking involved, then you sail your CTS.

[s/v Jedi]

Originally Posted by Fbfisher

How can you plot a single “course to steer” with constantly changing variables like wind, waves and current? My ST60 Raymarine (circa 2005) autopilot has a handy track button. Activate that after you’ve entered a waypoint and the XTE reads 0.00 (occasionally 0.01) for the 4.5-5 hours it takes me to reach my waypoint. Or am I just not understanding what the question is?

You are missing the point completely. I recommend a course “coastal navigation” that includes course plotting in tidal waters

[Idlegreg]

Originally Posted by Fbfisher

How can you plot a single “course to steer” with constantly changing variables like wind, waves and current? My ST60 Raymarine (circa 2005) autopilot has a handy track button. Activate that after you’ve entered a waypoint and the XTE reads 0.00 (occasionally 0.01) for the 4.5-5 hours it takes me to reach my waypoint. Or am I just not understanding what the question is?

Well I did make some assumptions – the original question talks about construction of a CTS and mentions tidal information – a later post tells us the OP is an RYA shorebased instructor showing people how to calculate the CTS on a ‘chart’ and he wants to know if it can be done using charting software. So I have assumed he is talking about accounting for tide streams in passage planning as taught to RYA students over many years. I believe a video was posted in the thread showing the technique.

It is true wind and waves will also affect the plan but tide is much more predictable and in many cases more significant – we have tables and diamonds on charts showing the expected tide stream for every hour at multiple locations. Although the stream on a particular day may differ from the predictions a little these tide stream forecasts are accurate for practical purposes. And unlike weather forecasts the prediction is available for years in advance. So it is normal to produce a passage plan taking the tide stream into account and then immediately before the passage make adjustments, if required, based on the forecast weather.

If the weather on the day allows the boat to sail (or motor 😊) with the course and speed the passage plan produced that will often be the best option and certainly the choice of most cruising yachts, unless it might lead to an uncomfortable beat (in which case many would stay in port or go elsewhere).

You can - additionally - do weather routing to find an optimum sailing course but apart from simple rules, like tacking within a cone, there was not much of that taught in RYA courses (maybe they have added it). It is harder to do without software and it requires accurate polar performance tables for the boat plus the wind, even once you have today’s forecast, is very variable especially in coastal waters. I would suspect that few cruising boats have the polar information and a fair few race boats who do have it may not actually have very good information.


An autopilots XTE steering function (where it adjusts to keep XTE as low as possible) can be useful when it is important to remain on a particular ground track but it will normally produce a longer passage time if there are tide streams or currents. Although I am a big fan of electronic navigation I think one a the major things lost by many who have grown up with it is the benefit in planning to account for tide stream. We may have a generation in which many believe that making the GPS track match the straight lines in their route is the best way to go - often that is just wrong.

[marcvilleneuve]

Originally Posted by Dockhead

In short, your approach works well when you can achieve a constant heading, by following a constant COG! But that works ONLY when the current is absolutely unvarying, either in speed or in direction!

Sorry but you got it wrong. Constant heading = constant COG only on a lake without current. When I am sailing on a bearing of 360 toward a mark at 6 kn with a lateral current of 3 kn I may be heading say 330 to compensate for the current. As a result my COG is a straight line on 360. So COG (360) = bearing (360) and then a staight line toward the mark. The problem is that the lateral current is not always the same 3 kn. It could decrease or increase many times on my way to the mark and this is why I keep a close eye on my GPS and constantly adapt my heading so COG stay on my 360 bearing.

It works I can tel you! Usually, most of the fleet get deported down river and they pay the price later to get to the mark. Imagine the advantage I get when the mark is 5 miles away ...in the fog.

Don’t tell them...they speak French any way.

[Idlegreg]

Originally Posted by marcvilleneuve

It works I can tel you! Usually, most of the fleet get deported down river and they pay the price later to get to the mark. Imagine the advantage I get when the mark is 5 miles away ...in the fog.
.

I don't think he got it wrong !


Your method might win in an inexperienced fleet and it might even be the best option IF the variations in the stream, or your boat speed, are not at all predictable.


BUT


If it is possible to make an even approximate prediction of the variations in the stream then you will do better if you sum up the total expected set and drift and then steer a single heading (CTS) that will compensate for that over the time taken for the leg. Given the plan will not be perfect you can change back to your method as you approach the finish to hit the mark but it's simply a fact that constantly altering heading so that you stick to a single straigt line COG in a varying cross stream is not the fastest way.

[Dockhead]

Originally Posted by marcvilleneuve

Sorry but you got it wrong. Constant heading = constant COG only on a lake without current. When I am sailing on a bearing of 360 toward a mark at 6 kn with a lateral current of 3 kn I may be heading say 330 to compensate for the current. As a result my COG is a straight line on 360. So COG (360) = bearing (360) and then a staight line toward the mark. The problem is that the lateral current is not always the same 3 kn. It could decrease or increase many times on my way to the mark and this is why I keep a close eye on my GPS and constantly adapt my heading so COG stay on my 360 bearing. . . .

If you think about it, I'm sure you will realize that this is wrong.


A constant heading will give you a constant COG any time when the water is either still or moving in a constant speed and direction. That doesn't mean that the heading itself is equal to the COG! Heading will equal COG when the water is still or when it is moving parallel to your course line in either direction. In all other cases, heading and COG will be different, but so long as the current doesn't change, they will both be constant.



And your tactic is wrong -- you are changing heading so sailing the rambling long way around through the water, in order to stay on the rhumbline. That is the slow way to the mark. If you tried to cross the English Channel that way, you would not only be slow, but you might not even arrive at your destination, on a light wind day and on spring tides. The current really rips on the French side, so if you are just crabbing along the rhumbline by doing what you propose to do, you can just get swept away to Barfleur rather than ever reaching Cherbourg.



If those currents are predictible enough to calculate CTS, then a racer using the correct tactic would beat you. The fast way is to get a vector sum of the set and drift of the current along your path and calculate a single heading which you hold all the way to the mark.


And no, I'm not the genius who figured this out -- it's taught in any coastal navigation course in this part of the world -- basic knowledge. This is taught at the Day Skipper level in the RYA courses!



Here's a very basic clear explanation of the process: Sailtrain: Navigation and Chartwork,


I also highly recommend reading through several excellent discussions we had on here in past years:


https://www.cruisersforum.com/forums...ad-115361.html

[Dockhead]

Originally Posted by Idlegreg

. . . If it is possible to make an even approximate prediction of the variations in the stream then you will do better if you sum up the total expected set and drift and then steer a single heading (CTS) that will compensate for that over the time taken for the leg. Given the plan will not be perfect you can change back to your method as you approach the finish to hit the mark but it's simply a fact that constantly altering heading so that you stick to a single straigt line COG in a varying cross stream is not the fastest way.

Exactly.



CTS doesn't need to be perfect. The less you change heading, the straighter a line through the water you are following, and the faster you will be getting there, so even a pretty rough approximation will usually give you a significant advantage over crabbing along the rhumbline.


And yes, once you're close enough that you can be sure that there will not be any more significant variations in the current, that's the time to start steering by COG. But not before!!

[Yellowtulip]

I've just had a look at Neptune. It's an impressive piece of software. But a couple of things occur to me.
Firstly, it uses some algorithm to predict tidal flow. The actual tidal flow may be different. On one Channel crossing a few years ago I found myself much further east than I had predicted. I assumed I'd miscalculated but on checking found my vectors were good compared to the published tidal data. Boat speed, too was not far from my predicted speed. I came to the conclusion that the tidal flow was perhaps 0.3-0.5 knots greater than predicted. Neptune will be subject to this also, so any 'optimum' CTS is still a guess.
Secondly, you can easily lose half an hour avoiding ships as you cross the busy part.
So, while I may well buy Neptune out of curiosity, I see no need to change my usual method for a Channel crossing:
Take the last of the ebb to get out past the Needles, work out the CTS from there (which is usually somewhere in the range 179-184) then enjoy the crossing which usually takes us between 11.5 and 12.5 hours.

[Idlegreg]

Originally Posted by Yellowtulip

...Firstly, it uses some algorithm to predict tidal flow. .......

Originally Posted by Yellowtulip

....so any 'optimum' CTS is still a guess.

I don't know for sure but I very much doubt Neptune is calculating the Tide streams directly using an algorithm. I expect that, like other software, it uses a table of data in just the same way you do when plotting the vectors manually. As far as I know some software uses Tidal Diamond data (which I believe is empirical) and others use data calculated using mathematical models but in those cases the modelling is done on much fancier machines than your humble PC and the output predictions are stored in the software in the same way as diamond data would be. Having said that, Neptune probably does use an algorithm to calculate a HW time for the day in question, which enables it to pull the correct stream data out of the table. If it does that using a simplified harmonic method then the HW time and hence the times of the streams may be slightly less well predicted than you could achieve if you have almanac data, but this difference is unlikely to be very significant.

But all this does not make the output a “Guess” any more than someone painstakingly plotting the tidal and vessel vectors on paper is a Guess. Just like our traditional plotting the tide stream in the real World will not exactly match the predictions but in most cases doing the calculation will give a better result than just a Guess - whether using computer plotting software or plotting by hand on a chart. The time of HW on which you base your plan might be slightly better if done manually but the computer will not only do the calculations much faster but it is less likely to make a mistake and it will almost certainly be more accurate in deciding where your boat is at each point in the passage which could well mean a more accurate output.

[Dockhead]

Originally Posted by Yellowtulip

I've just had a look at Neptune. It's an impressive piece of software. But a couple of things occur to me.
Firstly, it uses some algorithm to predict tidal flow. The actual tidal flow may be different. On one Channel crossing a few years ago I found myself much further east than I had predicted. I assumed I'd miscalculated but on checking found my vectors were good compared to the published tidal data. Boat speed, too was not far from my predicted speed. I came to the conclusion that the tidal flow was perhaps 0.3-0.5 knots greater than predicted. Neptune will be subject to this also, so any 'optimum' CTS is still a guess.
Secondly, you can easily lose half an hour avoiding ships as you cross the busy part.
So, while I may well buy Neptune out of curiosity, I see no need to change my usual method for a Channel crossing:
Take the last of the ebb to get out past the Needles, work out the CTS from there (which is usually somewhere in the range 179-184) then enjoy the crossing which usually takes us between 11.5 and 12.5 hours.

Well, your method IS a CTS method, just a crude one (or rather a good and handy one) based on a unique coincidence of factors.


I agree that ALL CTS calculations require guesswork, but I do not agree about guessing the actual tidal currents -- the models are extremely accurate, to a tenth or a couple of tenths of a knot. The currents may vary somewhat if a strong wind up- or down-Channel is moving the surface of the water, but otherwise they are bang on. What's less accurate is using a tidal atlas which just gives hourly tides, and doing the tidal vectors only on an hourly basis. Neptune uses the UKHO digital models which are on 6 minute increments and does a 6-minutely vector calculation so a big leap forward in accuracy.


Where we can't be so accurate is with our passage speed, which is the real guess, so updating the calculations based on plan vs actual passage speed is sensible.


The way I do it is to make a series of calculations based on different passage speeds from the low end of possible to the high end of possible, and that gives a range of CTS. That way I can see the sensitivity of CTS to different average passage speeds so I'll know how to respond as the actual average passage speed starts to be clear. It's usually not more than a few degrees, but can be a lot in some cases. So as the passage develops and I see what the average speed is looking like, then I can change the autopilot to the appropriate heading, then about mid-Channel re-run all the numbers.


This is much more accurate than the RYA hand method, which I used for my first 50 Channel crossings, but the RYA hand method while inherently inaccurate, is still extremely useful, and using that I was rarely more than a mile off at the end, especially if I did a mid-Channel correction. When you're one hour out then you can pretty much just steer by COG the rest of the way in. Works great.


The main point is that even a pretty rough and dirty CTS works better than crabbing along the rhumb line. If you are sailing at 6 knots such that you will get an equal amount of ebb and flood between Needles and C-bourg, then CTS will be very close to 180. I'm usually quite a bit faster than that, so the CTS calculation is more critical. At 9 knots I will have almost double one tide vs the other.


Neptune gives a different calculation which is particularly useful if you're not crossing perpendicular. If you're sailing say Alderny -- Brighton or something like that. It will tell you what departure time is favored by the tides, and tell you exactly what the passage times will be at different departure times, based on a given average speed through water. The differences can be huge and the favored departure time is often non-obvious.

[Idlegreg]

Originally Posted by Yellowtulip

Secondly, you can easily lose half an hour avoiding ships as you cross the busy part…..

Yes you can 😊 plans often get hijacked by life. With a PC it is easy to calculate a new best route to the destination if you find you have been taken well off plan. I still remember learning from Mary Blewitt’s lovely little book that once I am off my plan I must rub out all the lines I so painstakingly drew and create a new plan - from where I am NOW to the finish, rather than trying to get back onto my original route.

Originally Posted by Yellowtulip

….Take the last of the ebb to get out past the Needles, work out the CTS from there (which is usually somewhere in the range 179-184) then enjoy the crossing which usually takes us between 11.5 and 12.5 hours.

It is not clear to me if you are actually plotting the cumulative vectors and correcting for that or not, but of course you are making a passage that you have done before, at a predefined time (with respect to the tide) and it so happens that the crossing works out conveniently to 12 hours, so the tides almost cancel out, meaning the course will be similar whether tides are strong or weak that day. You are clearly doing the right thing and not fighting the tide in an attempt to maintain a straight ground track, so your system works well. But if you were making a passage new to you then software might have more benefit (or where you have been forced some way off track it could make recalculating much quicker).

Despite all this if you like doing it the paper way then I am not trying to suggest you stop – we choose to go sailing and it really is not the fastest, easiest or cheapest way to get about, if you are happy or even prefer calculating tide vectors on paper charts I would not want to discourage you and if a young sailor on a vey tight budget has to choose between not going or going without any electronics I would encourage them to get out and sail.

[Dockhead]

Originally Posted by Idlegreg

. . . Having said that, Neptune probably does use an algorithm to calculate a HW time for the day in question, which enables it to pull the correct stream data out of the table. If it does that using a simplified harmonic method then the HW time and hence the times of the streams may be slightly less well predicted than you could achieve if you have almanac data, but this difference is unlikely to be very significant.. . ..

Neptune uses the data from the UKHO, which is the same data used in Admiralty Total Tides. This is based on a massive database of empirical data run through a harmonic model, and frequently updated. It is the world's best tidal stream (and height) data and is used all around the world. It is typically accurate to a tenth or a couple of tenths of a knot.


I actually have a copy of Admiralty Total Tides on board, a present from a kind commercial skipper. It's out of date but there are ways to make it run This is typically carried on commercial vessels and gives 6-minutely data on streams and tide height.

[Idlegreg]

Originally Posted by Dockhead

Neptune uses the data from the UKHO, which is the same data used in Admiralty Total Tides.

That sounds good - if it is using Total Tide data then the accuracy is proabably as good as one can 'currently' ;-) acheive.

[marcvilleneuve]

Originally Posted by Idlegreg

I don't think he got it wrong !


Your method might win in an inexperienced fleet and it might even be the best option IF the variations in the stream, or your boat speed, are not at all predictable.

You said it: ...are not at all predictable.

That is why my first goal is the straight line. It is easy to get on the GPS. As soon as I rounded the previous mark I go for the straight line and THEN I look around for wind or current opportunities. If there is no clear opportunities I keep the straight line. Note: That situation happens most of the time between ebb and flood and the wind conditions tend to be quite unpredictable and the current conditions VERY unpredictable.

For the sake of this discussion I have to tell you that race in a river here with currents around 3 kn and more. Most of our races are upriver OR downriver and then back; meaning we are aiming to get to the mid-race mark EXACTLY when the current direction is changing. It is very atypical I know even worldwide. To wind a race here you need go fast obviously but also make no mistake with the current. We are stuck with the Saint-Laurence River. No choices!

Thanks for the great discussion guys!