Weather Routing

[rgleason]

Summary of Sean's recent improvements
Remove bad polars, bundle data, and correct some polars
Handle faulty polars better and remove faulty polar
Add wind barb overlay
First pass at improving terminology
Add wind strength configuration option
Improve handling of boat filename

[Gilletarom]

Hello Sean d'Epagnier,

Originally Posted by rgleason

Gille this i s a close up of a completed routing. The route is magenta. The close lines are isochrones that represent some very close time interval.

Sent from my SCH-I545 using Cruisers Sailing Forum mobile app

Hum ... Sorry, but RGleason reply does not satisfy me. All he said to me, I understood and explained in the French tutorial.

What I want to understand, and I do can not be specific enough in my question, is what is this little circle on the route

Thank by advance.

B.R. Gilletarom.

[rgleason]

Gille.. Play with it to learn.
The O represents the date of the grib file.
If you start 3 days earlier than the grib the O moves forward, etc.
The change in isochrone colors represents a change from using the grib to using climatology data.

See below, look at the grib date that has been set and where the O is. Note that more or less of Grib data is used depending on the start date. 2/7/xxxx, 2/4/xxxx, 2/2/xxxx

[Gilletarom]

@RGleason,

Originally Posted by rgleason

Gille.. Play with it to learn.
The O represents the date of the grib file.
If you start 3 days earlier than the grib the O moves forward, etc.
The change in isochrone colors represents a change from using the grib to using climatology data.

See below, look at the grib date that has been set and where the O is. Note that more or less of Grib data is used depending on the start date. 2/7/xxxx, 2/4/xxxx, 2/2/xxxx

Thank you. Very interresting.

B.R. Gilletarom.

[Gilletarom]

Hello Sean d'Epagnier

I apologize but I come back to the translation problems.

I used since yesterday the compiled version found on the server thingie.
I note that in the "Settings":
- Left, two items are not translated and they are not in the crowdin.
** Display mark when sail plan changes
** Didplay wind barbs
- Right, none of the 21 items is translated except, perhaps, "Start".

Also, if I take the example of the item "Time", it is present both in the "LokBookKonni" plug and leplugin "WeatherRouting." And probably in other plugins.
In the "LogBookKonni" plugin, in French, we have translated as "period".
In pluginWeatherRouting, it must be translated by "Duration".

But now, if the LogBookKonny and WeatherRouting are activated, the translation of LogBookKonni prevails and is displayed in the WeatherRouting.


Similarly, the word "Start" is present in several plugins, and even in the heart. That is why it seems reflected in the plugin "WeatherRouting." But translation in weatherRouting should be "starting point" or "starting Pt". But also the most appropriate translation is "Start". And that the crushes that of WeatherRouting.

Similarly for "End". in weatherRouting should be translated as "arrival point" or "arrival Pt". Another plugin, or heart, crushes the translation and replaces it with "End".

These are only examples. I could give more.

You are not there for anything. In opencpn.shoreline.fr tutorial, I described 21 plugins. Translation problems become insurmountable.

Sean, maybe this post is it useless. Maybe you are, you too, before an insurmountable difficulty. But if it can be used, it will be good.

Best regards. Gilletarom.

[rgleason]

Quote:
Just one thing - if you will be packaging weather_routing, please name the package without the underscore (opencpn-plugin-weatherrouting or opencpn-plugin-weather-routing) - I can't use it in the DEB package name to avoid problems with Launchpad tools and would like to stay consistent between distros...

Put this in the Weather_Routing thread and It might be noticed by SEAN.

Weather Routing

[mgh]

Hello all,

A quick question re OpenCPN v4 with weather routing. Apologies if it has been answered before but, after much searching, I have not been able to find an answer.

I am trying to create a weather route using two independent grib files. The first file is a standard GFS file for wind and surface pressure, the second is an RTOFS file for current.

Based on a hint earlier in this thread (I think it was for v3.3), I was able to create a second grib2_pi.dll in OpenCPN v4 (just copy and rename). This allows viewing of wind and current at the same time from two different grib files. And this works well - very useful by itself.

Using latest weather routing plugin, I can also create a route, so polars and program/plugin are all working fine.

The issue I am having is that the routing seems to only use the wind data. I checked this by trying it with and without the current data. The routes are the same in both instances.

I am pretty stumped - I thought I had this cracked, but no luck so far.

Has anyone been able to create a weather route in v4.0 that accounts for both wind and current? If so, could you share details on how you are doing it?

[rgleason]

mgh
If you look on flyspray tracker for Grib_pi and Weather_Routing you will find that by running two versions of the grib plugin, one with the Wind data and the other with the current data, that Weather_routing will use all the data.

There are techniques to concatenate two grib files which I have described previously in this thread I believe or in the girb_pi thread, to make a single grib file. You will have to search back in these two threads, or in the Tracker to find the description for how to do this. What I wanted was a tool for combining two gribs for this express purpose (to have current data available).



I believe that if the data is there and selected weather_routing will use it.

[mgh]

rgleason,

Thanks for the reply and pointer to the flytracker - very helpful.

I'm afraid I cannot duplicate your success with weather_routing to account for both wind and current from grib files. I have run several tests for a classic Newport-Bermuda route, and the results are as follows:

- With climatology disabled, the routings with/without current in two open grib_pi's, or with a single concatenated (GFS+RTOFS) file, were identical.

- With climatology enabled for current, there is a slightly different routing, but somewhat suspect since the difference was small and counterintuitive.

The only conclusion I can draw from this is that weather-routing will only use the current data from climatology (i.e. not from a separate RTOFS file in a separate grib_pi, or in a concatenated file.)

There is an obvious question of whether the RTOFS file is any good. So I tested the same GFS and RTOFS files in another program. Here the results were as expected - impact of current was clear and more dramatic shift to west.

Would love to find a solution or to be corrected - I would very much like this to work, but out of ideas at this point.

[rgleason]

MGH,
I have not recently tried running dual grib_pi with Wind Grib + RTOFS Current!
It is very useful to know if something has changed, and it appears to have change such that either Grib_pi or Weather_routing_pi no longer can use the current data.

Yes, Climatology does have macro currents such as Gulf stream, and Weather_routing does have a selection for this and it should work as I have tested it more recently.

What I would like to suggest is that you register with Tracker or Flyspray
OpenCPN::Tracker Register new user
and enter your recent experience with both Grib_pi and Weather_routing_pi (enter it into both plugin Projects or Tasks at the dropdown menu as an Additional Feature or as a Bug.
And request that both Grib_pi and Weather_routing_pi try to address using RTOFS Current files!!

Your entry would be very helpful. The more people interested in this the more likely it will happen. I will try to encourage Jean Pierre (grib_pi) and Sean (weather_routing & grib_pi) to help try to figure this improvement / bug fix.

There are several things that need to happen.
1. Grib_pi needs to support RTOFS currents concurrently in real time with Wind data or there should be a software tool which enables the user to easily combine the grib files and run them as one grib file.
2. Grib_pi needs to support RTOFS "current" data-type.
3. Weather_routing needs to support RTOFS "current" data-type and use it in its calculations. (My experience was that this worked earlier, perhaps something has changed.)

We need to reference these problems in both the Grib_pi and Weather_Routing_pi Threads!
http://www.cruisersforum.com/forums/...ml#post1758152


http://willkamp.com/opencpn/flyspray...s&task_id=1709

http://willkamp.com/opencpn/flyspray...=984&project=4

[rgleason]

Polar Files - Ingestion - Cruisers & Sailing Forums

[seandepagnier]

Originally Posted by mgh

rgleason,
re identical.

- With climatology enabled for current, there is a slightly different routing, but somewhat suspect since the difference was small and counterintuitive.

How is it counterintuitive?

The only conclusion I can draw from this is that weather-routing will only use the current data from climatology (i.e. not from a separate RTOFS file in a separate grib_pi, or in a concatenated file.)

I don't think that it can work with two grib plugins running at the same time yet.

How do you make the concatenated file? Does both wind and current appear in a single grib plugin?

There is an obvious question of whether the RTOFS file is any good. So I tested the same GFS and RTOFS files in another program. Here the results were as expected - impact of current was clear and more dramatic shift to west.

What other program? Can you provide screenshots?

Would love to find a solution or to be corrected - I would very much like this to work, but out of ideas at this point.

We can make this work, but I need you to provide the grib files you are working with and it would be helpful to post the weather routing xml files as well.

[rgleason]

Sean, hope MGH is watching... I'd like to help too. Right now sort of booked, but I've written enough way before about it... and I thought it was more of a grib problem so I wrote about joining two gribs in the Tracker for grib_pi.

I had been running two instances of the grib_pi and grib2_pi and loading the two gribs, then weather_routing_pi just used the data successfully.

I think I write about this way earlier in this thread... I don't know if the situation has changed, but it seemed like it would be something that could be improved.

[seandepagnier]

It's good to attach the configuration xml files, polars etc... but if it really can't route past 90 west this is a bug... For efficiency reasons, it cannot route past both 0 and 180 longitude on a single route (it has to choose which one is possible at the start)


I tried with climatology so maybe it's a problem with the grib plugin and how it interpolates data there, not sure. Please attach the grib file used.

[seandepagnier]

I get the same failure, but with a different polar no problem. I think it's because your polar doesn't cover high enough wind speeds to work.