Weather Routing

[rgleason]

And immediately a question comes to mind: Which one is the best? The one that requires a short calculation time with TS = 30mn or the one that requires a long duration of calculation with TS = 10mn?

For the data given to WR, it is quite obvious that the 10mn interval is better, but you may be tacking and changing sail sets more frequently.

2 ° Look how the tracks leave the front port of Cherbourg. Two tracks pass through the west exit, and a track goes through the east exit.

Yes, I am not surprised by this Gille. The shorter intervals allow finer calcs all the way across, and it may have been the case that the other route was optimal during the majority of the calculation, but in order to tack up into the destination, the calcs found that the alternative route was better, and switched to that route at the last minute before finishing the optimal routing. Watch as WR calculates to see what is happening. It is only as good as the info and direction that you are providing, it makes the best of what info it gets.

And one more question: Since the conditions are the same (wind and current) with TS = 30 or 20 or 10mn, why does WR make different choices? In these conditions can we trust in WR?

There is another good question. It is always, in the final analysis the Skipper who must make the decision. As stated in the manual, this is just a tool, using the info available. The skipper may know something more, but the WR program gives him the parts that can be figured out based on inputs.

I am even downright skeptical in the choice made by WR in both cases of exit to the west. Indeed, it comes out against the current and with a rather head wind while, towards the east, one is current and wind in the back to arrive at the exit.

Well, yes, you should always analyze a routing. It is the best practice, and you should carry skepticism so you don't blunder. You may know some things or have some experience that WR does not. As we say this is a Tool for you to exercise JUDGEMENT. It is a helpful tool but the skipper has the last word.

Keep thinking about why the program is making the choices, and noticing WHEN it makes those choices during the process of the calculations. That tells you a lot.

WR does not know everything. It does not know that the grib prediction model is often wrong about the amount of veering/backing during certain weather conditions. It just figures out the best route using the gribs given. Nor does it always know the tides or tidal current, in fact the current gribs inshore are often very poor. You will note that we have stated this very clearly in the manual.

[rgleason]

Gilletarom, I have an idea for you. Go virtual racing. Use qtvlm and learn more about routing by racing other good players. Many good virtual players are frenchmen. When virtual racing the Vendee round the world. I was in good position, made a blunder, dropped into 900's, then paid attention to the gribs and tried to anticipate how the grib predictions would change with each update. I got better, but there were still many frenchmen ahead of me. I would get up at night with each grib update, do my routings anticipate and make decisions. I worked my way back to 50th, missed the finish line, ran into the french coast, by over sleeping and poor marksmanship, turned around and finished 125th with many frenchman still ahead of me.
You should do a virtual race, you will learn.

Keep doing routings, that's another way to get the skills.
But the virtual race was addictive, when it was finished, I had to stop.

[did-g]

Originally Posted by Gilletarom

Hello WR team,


2 ° Look how the tracks leave the front port of Cherbourg. Two tracks pass through the west exit, and a track goes through the east exit.

And one more question: Since the conditions are the same (wind and current) with TS = 30 or 20 or 10mn, why does WR make different choices? In these conditions can we trust in WR?

Is my analysis bad?

It's a known issue, currently WR is unable to reduce its time or angular step.

For testing : add a zone closing the Solent's Needles an rerun with the same configuration, it should finish using the east route but time should be longer.

The right way could be:
use default value for time and angular, build a polygon (or a circle ?) from start, intersect with exclusion zones, if not empty halve, rinse repeat. Should be faster in open sea.

[rgleason]

Good point Did-g. Use an ODraw boundary to block & direct.

[seandepagnier]

Originally Posted by Gilletarom

Hello WR team,

And now, a problem around the starting point.

I resume research between Cherbourg and Cowes. In the first tests, I put my starting point completely outside the harbour. Then, I set a starting point in the harbour front, just south of the breakwater: "Cherbourg-inside".

With starting point at "Cherbourg-inside", I did 3 tests with a timestamp of 30Mn, 20Mn and 10Mn. WR does not complete correctly if TS = 30mn. But he finishes for TS = 20mn and TS = 10Mn. So I have two tracks by exporting the virtual tracks of the virtual boat and an incomplete "virtual track". See the screenshot 30mn in pink, 20mn in green, and 10mn in gray.

1 ° Observe that two tracks pass by the west of the Isle of Wight and a track by the east of the Isle of Wight.

And immediately a question comes to mind: Which one is the best? The one that requires a short calculation time with TS = 30mn or the one that requires a long duration of calculation with TS = 10mn?

The shorter time step is more accurate, but takes longer to compute. Look at the end time and duration. The shorter time step should find a "faster" route. At some point reducing the time step further won't make much difference, but you didn't reach that. Try 5mn?

2 ° Look how the tracks leave the front port of Cherbourg. Two tracks pass through the west exit, and a track goes through the east exit.

And one more question: Since the conditions are the same (wind and current) with TS = 30 or 20 or 10mn, why does WR make different choices? In these conditions can we trust in WR?

This is the fun part. Small changes in parameters can give completely different routes.

In this case, it's probably true that west or east the difference is slight. One way is longer, but with wind and current, the shorter route is against both. I would take the more comfortable route.

Changing to a slower boat, you will see, that you won't route against wind and current as often.

I am even downright skeptical in the choice made by WR in both cases of exit to the west. Indeed, it comes out against the current and with a rather head wind while, towards the east, one is current and wind in the back to arrive at the exit.

Is my analysis bad?

What is the overall time? Is the difference just a few minutes?

As for verifying things... you can read the source, and I am confident that it calculates the correct routes based on the parameters (unless you exposed a bug)

You could use an alternative routing program to compare, but then, often the other routing program may also not be verified.

As for safety... it is really not my responsibility, you can have problems either way. The weather route gives interesting "suggestions" that is all that can be asked of it.

I don't suggest program,ing the autopilot to follow the weather route exactly especially in bad weather.

[rgleason]

Gilletarom, your examples have been used quite a lot. I just wanted to clarify something, the attached weatherroutingconfiguration.xml.doc (remove doc) has all the setting which allow the routes to start and fully complete. Please note that I have turned off most of the Boundary's and these routes will all "compute" concurrently (at the same time) by just highlighting them and hitting "compute", and they process quite quickly. I find no problem with successful "completion".

I think I just changed a couple of routes to be Max angle 90 or 100 degrees rather than 120 to speed them up, and more importantly, set the interval on one or two to 20 minutes.

[rgleason]

Sean updated weather_routing-climatology Beta.
For windows See https://github.com/rgleason/weather_routing_pi/releases

[rgleason]

Made PR from Stelian and Did-g of the PR to Sean for
Has the PR to Sean's Repository for:
PR#111 Grib Backport did-g
PR#112 Add missing end did-g
PR#115 Config Report did-g
PR#113 Fic Config Dialog Stelian
https://github.com/rgleason/weather_...tag/1.1108beta

There is a new beta release uploaded, but it does not work with GRIB files from Gilletarom that we have been using.
I believe the problem is from Did-g's PR#111.

Dig-g can you fix this so I can redo this beta? Or should I redo it without the PR#111?

Don't bother using it guys.

[Gilletarom]

Hello,

So, I'm going to wait for the improved version. And this especially since, at the moment, I really have a lot of crash due to OpenCPN (or to WR? Or to Draw?)

[did-g]

Originally Posted by rgleason

Made PR from Stelian and Did-g of the PR to Sean for
Has the PR to Sean's Repository for:
PR#111 Grib Backport did-g
PR#112 Add missing end did-g
PR#115 Config Report did-g
PR#113 Fic Config Dialog Stelian
https://github.com/rgleason/weather_...tag/1.1108beta

There is a new beta release uploaded, but it does not work with GRIB files from Gilletarom that we have been using.
I believe the problem is from Did-g's PR#111.

Dig-g can you fix this so I can redo this beta? Or should I redo it without the PR#111?

Don't bother using it guys.

If you can redo without PR#111, without an Opencpn PR it won't work. It's using an incompatible format for internal grib records.

[Gilletarom]

Hello RGleason,

Originally Posted by rgleason

.....

There is a new beta release uploaded, but it does not work with GRIB files from Gilletarom that we have been using.
...

RGleason, please, do not say "with GRIB files from Gilletarom". What is important is that now, WR can really be used in Europe thanks to the grib files which contain both wind data (speed and direction) and current data (speed and direction).

And we must say thank you to the Greek specialists who created the Skyron website in which they inject free data that we need but also other data.

So, you must say "with Skyron GRIB files".

[rgleason]

Thanks Gilletarom, I saw the currents but did not realize or know how they were created.
Your description helps. Could you post the website here for others?

[Gilletarom]

Hello RGleason,

Originally Posted by rgleason

Thanks Gilletarom, I saw the currents but did not realize or know how they were created.
Your description helps. Could you post the website here for others?

Hum ... But, you already know this website:
OpenSkiron - OpenWRF Gribs

[rgleason]

Gilletarom,

New version without the PR#111 Grib Backport from Dig-g.
https://github.com/rgleason/weather_routing_pi/releases

But now the Routing cannot work when Boundaries is on!!!
Why? This is with ODraw 1.4.47

However it does work with ODraw 1.4.42!!!

[rgleason]

This version of ODraw 1.4.49 works too
http://www.cruisersforum.com/forums/...ml#post2551278

The other alternative is to use Jon Gough's version and download the necessary DLL's to work with a MS VC++ 2017 compiled version.