Weather Routing

[did-g]

Originally Posted by boat_alexandra

What, if anything does this actually achieve? I read that it predicts how stable or dependable the route is, but is that actually proven?

No I just want to know at what time I have to leave for finding water in harbor after a day of sailing. Rather than choosing a departure interval I want to choose an arrival interval.

Many harbors around here are behind gates, locks, wall or have no water for six hours.

For example on a 12 hours period our berth is only reachable 4 hours: up river, behind a lock in a drying bay and next shelters are 5/10 miles away (themselves behind gates and drying) .

Originally Posted by boat_alexandra

So.. Should it just cut them in half and try again if the routing fails? Maybe cutting angular step down, but not more than to 3 degrees?

Something like that could be worth testing.

By default 'Time steps' (BTW do we need second resolution for it ?) should be auto-magically set, maybe half grib time and reduce on failure? We can sail 12 miles in half an hour, for coastal that's a lot.

There's an issue in grib with interpolation though, on failure it returns an empty grib, I have some current grib like that, U and V aren't at the same lat/lon and interpolation fail. I have some code, need a PR.

[did-g]

Originally Posted by boat_alexandra

if the max deviation is high enough, say 180, then it can explore all directions, but takes a very long time. With a more direct route possible the smaller this can be, and the routing completes quickly.

Now maybe if the optimal route is stopped by this, we could suggest the user to increase it. Also, if an optimal route is found, but the route touches the edge of the route map, this value should be increased.

There's an old saying
"it doesn't mater how fast it doesn't work"

- Either weather routing figures it out.
- Or thanks to ocpn_draw inclusion zone, restricting search is easier for users and IMO it makes more sense than changing non descriptive values IIRC there's four or five of them which could, or not, affect the outcome it not obvious way.

[rgleason]

Thanks for the review, trials and comments.
1. The wiki has the Irma grib example with full documentation, its worth reading.
2. We are providing 4 simple climatology based configurations that work out of the box.
3. I will check the polars. I sent them to Sean earlier for his review. The names of the polars are to help users conceptualize how to use the multiple polars.
4. Since you've done the intilalization once, and the paths and directories are all set up, you could probably skip "cleanup" and just install the plugin and then copy the files over by hand from the program directory to the user/programdata directory.
5. We should probably have a "Begin again" button that copies the plugin program files to the appropriate user/program data folders, for beginners to start over easily.
6. Regarding the Advanced TAB, those parameters affect the amount of calculation required. This pi must work with low powered devices. So perhaps there should be 3 default setting buttons which change the advanced tad values . A. Fewer calcs B. MId calcs C. Many calculations. The button C would have values that were most likely to complete.
7. Thanks for highlighting some of the problems, it always helpful to have a through review, so we know what needs improvement.

[rgleason]

Regarding the effect of changing these values, for example, if your set the max wave height to 10 meters the computation will try to find a route that does not have waves over that height. It will probably be a longer route causing the grib to fail, but if you are also using climatology averages or most likely, there will still be data and the route will complete.

I like the concepts of start and destination planning. What tools are needed for destination planning? I am sure Sean is familiar with the dammed harbors and wicked tides, perhaps he has some programming ideas.

[rgleason]

We are using multple configurations using only climatology (so we don't need a grib file) to give new users a start.

I would like a "restart-start over" button to copy the original files again from c:/programfiles (x86)/opencpn/plugins/weather_routing_pi to c:/programdata/opencpn4.8.0/ or opencpn/plugins/weather_routing and to the correct directory. So a new user can experiment and then "Start over again".

We are providing 4 configurations for beginners. That should be fine. I will check and adjust the polars to what I think is right. We do not want the polars to be too complicated but to show what is possible. I have already sent the polars to Sean for his review earlier.

I have already done a grib example. That uses Imra and is on the wiki with extensive documrntation and can be downloaded there. Read the documentation, it discusses why routings fail.

,
I realize now that these settings below can/should be different for low powered devices and fast machines. Maybe there should be three o "default buttons" that change these settings for speedy routing for low power devices, and a mid button which is slower for low power devices but does more calcs and relaxes the constraints, and a button for fast computers which takes even more calculations but has settings that help to ensure completion?

So if a calc fails, the user would pick the next higher default and try that. If that fails, try the highest setting (but on low powered rpi this setting might be too much for the device. )

Keep in mind that changing settings so they are more likely to work requires more intensive calcs.

Also keep in mind that code to figure out why it failed could also slow the routing down, depending on how it is done.

(QUOTE)
Max Diverted Course 90 deg [120 deg?]
Max True Wind 100 knts [25 knts?]

Max Apparent Wind 100 knts [30 knts?]
Max Swell 20 meters [5 meters?]

Max latitude 90 [ good]
Tacking Time 0 [ 120 seconds?] Is this seconds?
Wind vs Current 0 [What is this value? What does it mean? What does it do? Units?]
Max Course Angle 180 degrees [How is this different than [Diverted course?]
Max Search Angle 120 degrees [Can you explain what this is for specifically? How is it different than diverted course?]

Sean if you can take a moment to explain these, I'll promise to get them into the manual! Also let me know if I should change any of the configuration settings please. (/QUOTE)

[rgleason]

By the way Sean changed the version number to 1.11. and on the release page wr have a version tag with sub numbering.

[Gilletarom]

hello Sean,

In an older version of WeatherRouting, it seems to me that there was some of the software that made it possible to determine the polar of the boat. In a recent post, RGleason says that this part of the software was still in WeatherRouting, but that part of the software was "unplugged".

Indeed, it is a good thing to have disconnected this part of the software.

On the other hand, it is common for users of routing software to search for the polar of their boat. It might be that this part of the software is set apart, in another plugin. At least, if it does not require too much development.

Gilletarom.

[rgleason]

Gilletarom,
Sean may have removed it, but probably did not delete it (I would have to look.), but the point is that this would give you a polar file based upon some of the characteristics of the boat (OA and WLL, draft, beam, displacement, sail area, etc) that could be used for routing, but it would not be as accurate as one that is recorded from actually sailing in constant conditions and recording performance, by saving a VDR file (provided you have all the necessary data in nmea0183 format) and then playing it in Polar_pi to make a polar file.

The other alternative if you have Windows is to use Patrick Teircelin's Polauto program which is quite flexible and will make good polars from available boat data.

These alternatives are all written up in the documentation.

Additionally, we have collected almost 200 different polars for boats in one place, to be used as a starting point. There is no more we can do!

Asking for more programming development when there is a strong ongoing and focused effort to get weather_routing easy to use and with configuration settings that the normal user can understand, is enough right now. IMHO.

I am sure Sean has not lost the code and that it will reappear at a good point. I had much the same reaction when I saw the revisions, but Sean did what was necessary to simplify what is a sophisticated plugin.

Originally Posted by Gilletarom

hello Sean,

In an older version of WeatherRouting, it seems to me that there was some of the software that made it possible to determine the polar of the boat. In a recent post, RGleason says that this part of the software was still in WeatherRouting, but that part of the software was "unplugged".

Indeed, it is a good thing to have disconnected this part of the software.

On the other hand, it is common for users of routing software to search for the polar of their boat. It might be that this part of the software is set apart, in another plugin. At least, if it does not require too much development.

Gilletarom.

[rgleason]

Please see this Issue #94 in github
View Weather Routing Data - Discussion how to improve
https://github.com/seandepagnier/wea...g_pi/issues/94

and add your thoughts and graphics.



Route Indicators - Discussion how to show the route conditions visually.

https://github.com/seandepagnier/wea...g_pi/issues/95

[rgleason]

Yes, I have seen this too. It is a graphic anomaly and other views of the same file look fine and they work.

[Gilletarom]

Hello Sean, Hello RGleason, Hello All,

I just saw Weather-Routing working, as I wish, for the first time.

See the attached screenshots.

The General Plan shows an area of ​​the coasts of France.
This region includes islands and many areas in which it is difficult to navigate.
So I made many draws with Draw in which you do not have to enter the boat. They are recognizable because their border is red (I put an internal grid in light gray and I masked the marks which delimit them.).

Then I used three downloaded grib files here:
OpenSkiron - OpenWRF Gribs.

These grib files contain a lot of data and in particular the two types of data that are important in this area: Wind and currents in the same grib file.

Then I defined several configurations for WeatherRouting, using the 3 grib files and for one of them, I even chose two different start times.
Note that in this area, Climatology is useless.
Also note that I had to choose the COG angle = 150 ° because without this value, it was not always possible to create a route.

I was able to build 4 roads. They are visible in blue, pink, green and gray.

This is good news. Thank you Sean, thank you all.

[Gilletarom]

Hello Sean,

A little bug. Without doubt easy to correct. See the two screenshots attached.
The item "Current configuration" is translated in a strange way in French.

I watched the translation in the Crowdin. The translation that exists in the Crowdin is not the one that currently appears in the screenshot of the French version.

It is probably necessary to update the language files in the next version of WR.

[rgleason]

Note that in this area, Climatology is useless.

Just what do you expect? Climatology is 30 years of NOAA data averaged for each of 12 months. Think of it as a super digital version of Pilot charts but it contains considerably more information.

You are incorrect about Climatology being useless. It is a planning tool and it can provide extended average or most likely data so that a long trip will complete.

It is probably not good for determining the best weather windows for leaving (on a daily/weekly basis) but it is good at determining generally the best times and routes. Grib files are most appropriate to use for departure/arrival planning, if you can get a reliable grib that will go the full distance, otherwise use Climatology, at least it has monthly averages from NOAA which is better than most programs.

To see any difference with climatology in start/finish times you would have to be picking more on the weekly/monthly scale.

[rgleason]

Good, Sometimes it is helpful to read the instructions!

Also note that I had to choose the COG angle = 150 ° because without this value, it was not always possible to create a route.

[fogmachine]

Originally Posted by rgleason

You are incorrect about Climatology being useless. It is a planning tool and it can provide extended average or most likely data so that a long trip will complete.

Exactly. Climatology is great for passages whose duration is longer than weather models are accurate. It doesn't tell you when to leave, but it gives a hint as to when you might arrive.