ncdf plugin for reading NetCDF tidal currents

[Rasbats]

ncdf has been updated.

The release section of Github is now being used for the binary files like the plugin, netcdf, example file etc.

Now works for OpenGL ON and OFF.

Forward/back buttons implemented.

Work has started on areas for display but these do not do anything yet.

The problem of the density of information (arrows) was considered. If they are reduced, even with scale change they will not portray streams accurately in many areas.

Intend making a context menu to "Drop ODAS buoy". This virtual buoy will show the tidal stream at that location. Several of these could be dropped by the user. The main current display can be turned off to reduce clutter.

Mike

[rgleason]

Mike, dropping a tide buoy sounds like a good idea.

I have tried to build this new version using VC++, as is required, with the settings as shown in the attached ncdf-VC++instructions-4.txt and it builds with 16 LNK errors to src/ncdOverlayFactory.cpp these errors are in the attached ncdf-errors.txt.

I'm going to try your dll next.

"The release section of Github is now being used for the binary files like the plugin, netcdf, example file etc."

I look in the "build/release" directory and don't find anything, I guess because it has been overwritten by my attempt to build. If there was a dll there it is gone now, and there is no way to git fetch and get it back. So I guess I have to get on my account at github got to your repos and just download it. Looked in your repos and there appears to be no build directory thus no release directory and no dll to test.

[Rasbats]

Rick,

On my ncdf GitHub page you will find a link called 'release' (actually '1 release'). Click on this and you get another page where the binary files can be downloaded and also a zip and gzip of the source code.

Mike

[rgleason]

Thanks Mike, this is a new feature to me.
https://github.com/Rasbats/ncdf_pi/releases

Should make things much easier for me!!!

[Rasbats]

ncdf v0.3 is now on GitHub.

https://github.com/Rasbats/ncdf_pi

Areas are now implemented and a context menu to 'Drop Tidal Arrow'.
************ IMPORTANT***************************
With the areas it is important to select the area you want to use from the drop-down before finding the file and forecast times in the tree. If the two do not match you get no arrows.
************************************************
TODO ... give the user a warning about this.

The Download panel of the plugin shows the minimum/maximum lat/lon for the Copernicus download of the area forecasts. If the area from Copernicus is too small ... no arrows.

The context menu is activated by right-click 'Drop Tidal Arrow'. You can drop as many arrows as you like. With these showing the normal plentiful arrows disappear.

To return to 'normal arrows' you right-click 'Drop Tidal Arrow' and 'Cancel'. All the dropped arrows disappear and you can tick the checkbox for the normal arrows, if this is not already selected.

For compiling on VS2013 and enabling OpenGL you will need to compile with
glu32.lib
opengl32.lib
in the additional linker libraries section, as well as opencpn.lib.

(Rick ... I think this may be the reason for your linker problems)

A Windows compiled plugin dll is in the GitHub release section (v0.3).

Mike

[rgleason]

Mike I added those libs

For compiling on VS2013 and enabling OpenGL you will need to compile with
glu32.lib
opengl32.lib
in the additional linker libraries section, as well as opencpn.lib.

It compiled easily and I tried it in v4.2.0 and v4.2.1 and had trouble seeing any current files. Am I missing some files here that need to be put somewhere? DLL,s whatever? It would be helpful to get the full instruction.

[Rasbats]

Rick,

I have placed two recent Copernicus files in the 0.3 release section on GitHub.

ec.nc = English Channel
sb.nc = Southern Brittany

Download these to a suitable folder on your machine. Select this folder with the plugin 'Folder' button.

Select English Channel from the drop-down. Go down the tree until you find ec.nc. Expand this section and you should find the forecast times.

Click on a forecast time to show the arrows. Once you have selected a forecast time the next/previous buttons let you scroll through the forecast times.

Right-click gives you the option to drop a tidal arrow (See my previous post).

If you want to look at Southern Brittany arrows you need to change the drop-down and renavigate the tree.

Mike

[rgleason]

Mike, I started out by using SB yesterday and that threw me off, because there is no current showing!! I do not know why, but I tried all different kinds of things to make it appear and it did not. Wasted too much time.

Then tonight I tried the EC an found that that worked!

Several things:
1. Is there some way to make the applied arrows show when zoomed out more?
2. Can the menu box for the selection of tides be larger or made so it is adjustable?
3. Can the final settings of the menu be saved so that the menu opens in the same way next time Opencpn is opened?
4. Also why was there not any current in SC

This is working better now. Any idea how I can be able to compile it without those errors?

PS: Later, It is compiling fine, and it is working! I was mislead by the way SC worked, and I was thinking of Hydrographic Survey, sorry.

[Rasbats]

Rick,

1. Is there some way to make the applied arrows show when zoomed out more?

Yes.

2. Can the menu box for the selection of tides be larger or made so it is adjustable?

No problem.

3. Can the final settings of the menu be saved so that the menu opens in the same way next time Opencpn is opened?

You should find that the area is retained in the drop-down.
I cannot predict what files the user has in the tree so I can't take the user to the point in the tree he/she was using before.

Having said all this ... the priority at the moment is building the plugin for Linux. This is causing a lot of head/Google scratching.

Mike

[rgleason]

I cannot predict what files the user has in the tree so I can't take the user to the point in the tree he/she was using before.

Not a problem.

Hope you can figure out the linux issue.

[sailorF54]

The French Hydro office has recently started to offer free .nc files for Channel / Biscay and the Med (se pic)
They are available here
http://data.shom.fr/#donnees/oceanographie
Select 'hydrodynamique surface horaire'
hit 'télécharger données sources' and double click on the area
They are about 65 Mo per day

I have uploaded an example here
DL.FREE.FR
(valide 1 month)

I join the included xml file (under .pdf guise)

Your ec and sb samples display fine, but not the SHOM files

Thanks Rasbats for your hard work (including OT)

[rgleason]

Thanks. Getting the left menu to show is a little trickly. Try clicking or right clicking in the upper bar.

It would be nice to know which files are usable with NCDF currents.

[Rasbats]

sailorF54,

Thanks for the pointer to the SHOM.FR data.

COPERNICUS:
:julian_day_unit = "Hours since 2011-03-31" ;

short v(time, latitude, longitude) ;
short u(time, latitude, longitude) ;

SHOM.FR
time:units = "days since 1950-01-01 00:00:00" ;

short u(time, depth, Y, X) ;
short v(time, depth, Y, X) ;

Even though depth is zero the extra dimension for u and v would need quite a bit of extra coding. Also the forecast is only 24 hours at present: time = UNLIMITED ; // (24 currently)

File download size is about 10 times that of Copericus for specific areas, because of salinity and temperature data (not needed).

Interesting but these files would need too much work to extract the information.

Mike

[rgleason]

Mike, Is there a reason ncdf.dll would not work with v4.2.0? I had been using it with v4.2.1 but have changed to my bigger laptop. ncdf does not appear in the v4.2.0 plugin list when the dll is in plugins directory. Is Pavel limiting plugin names somehow?

I am going to just compile 4.2.1 anyway, but I wonder.

[sailorF54]

Originally Posted by Rasbats

sailorF54,


Interesting but these files would need too much work to extract the information.

Mike

Mike

All your points are well taken and your conclusion makes perfect sense.
I am just sad that such good data will be wasted, but the culprit is SHOM, not you