New Generation GRIB Viewer?

[PjotrC]

Immensely impressed by earth wind map ...

The looks and readability of this might perhaps be repeated in a new kind of GRIB viewer.

It should be possible to generate randomly a number of "wind-rays" and trace them through a GRIB grid, interpolating, smoothing and colour-coding between the grid cells.

Would be very nice for small-scale, long distance route evaluation.

(A static snapshot looks good, but is not so useful, so I did not reproject Orthographic into Mercator. it's just a sample view. The real thing should be dynamic).

[bdbcat]

PjotrC....

Wow....Very Cool.

Lets build a GRIB viewer like this, Mercator of course. Everyone loves a good movie

Dave

[PjotrC]

poor connection... hope now the attachement sticks...

[europaflyer]

Ask the people who wrote this for their code and hack it in to Ugrib or OpenCPN or whatever. This is just an alternative way of displaying what I suspect is exactly the same source data, using animated lines with colour and speed corresponding to windspeed. The density of the lines possibly corresponds to isobar distribution, or maybe this is just a function of wind data as well. Without being overly disparaging, for people who understand standard synoptic charts the value of this is a little dubious, as there is no way of finding the magnitude of the wind and pressure, or tracking fronts. It is very beautiful though!

[bdbcat]

europaflyer....

Most people cannot read synoptic charts.
But most people would be able to look at this graphic and predict what the wind direction will be in 12-24 hours at their location. Or, at least as in this case, when there is a sharp cold front on the US seaboard.

And that is useful information.

I can imagine a mouse rollover popup with detailed predictions at a particular location, and overlayed cloud cover/precip, etc.

And then a time lapse view, and then yada/yada....

To me, much more approachable than dry old wind barbs and isobars.

Dave

[conachair]

Originally Posted by europaflyer

Ask the people who wrote this for their code and hack it in to Ugrib or OpenCPN or whatever. This is just an alternative way of displaying what I suspect is exactly the same source data, using animated lines with colour and speed corresponding to windspeed. The density of the lines possibly corresponds to isobar distribution, or maybe this is just a function of wind data as well. Without being overly disparaging, for people who understand standard synoptic charts the value of this is a little dubious, as there is no way of finding the magnitude of the wind and pressure, or tracking fronts. It is very beautiful though!

No need to ask. It's open source by the looks of it.

https://github.com/cambecc/earth

Nothing to do with isobars, a pretty grib2 viewer. But a nice one.

As it's a weather thread, for a look into synoptics check out 11....

http://www.metoffice.gov.uk/learning...ons/factsheets

[Littlechay]

Yes, it's been going viral on FB. Not a forecast, it is using actual wind values apparently, I suspect from one of the SCAT satellites.

Useful for now-casting maybe - but so is looking at the barometer, truewind and sticking your head outside.

Not sure how big the files would be.. anything over about 10 - 20k forget it offshore.

I will admit it is pretty.. but I would rather see more effort put into dealing with actual grib files and getting some better history functionality (TW, Temp, Pressure) in the dash.

my 2 cents
Chris

[europaflyer]

Originally Posted by Littlechay

Yes, it's been going viral on FB. Not a forecast, it is using actual wind values apparently, I suspect from one of the SCAT satellites.

Not sure how big the files would be.. anything over about 10 - 20k forget it offshore.

I will admit it is pretty.. but I would rather see more effort put into dealing with actual grib files and getting some better history functionality (TW, Temp, Pressure) in the dash.

No, this is just a fancy GRIB viewer in effect. And yes, I agree that the utility of it is limited.

As far as I can tell from the website https://github.com/cambecc/earth the essence of it is that standard GRIB2 files are converted to JSON format, then the actual rendering (what we want) is done by a browser app (HTML5 I think). I am pretty sure that to do something similar in a program like Ugrib or OpenCPN the code would have to be completely rewritten. Perhaps it could be done locally, just using their HTML code in the browser to process a GRIB file rather than downloading the GRIB and the HTML code to process it each time.

Regardless, I'm still not sure how valuable it would be to the majority of users. I think most people don't really have too much trouble with synoptic charts, which do present more data than this does. Yes, it is an extremely pretty and very 21st century way of presenting wind data, but it is exactly the same data we get in GRIB files. More importantly, with the relatively small density of data points in a typical sailor's GRIB file, essential to make it downloadable, this presentation format wouldn't look nearly as pretty I think. Also, if you zoom in to the sort of scale we are interested in as sailors, the whole thing just becomes a big swirling nonsense. Pretty, but useless. Their code is geared towards a global view, and I doubt the more subtle variation in larger scales would be well suited to this presentation method anyway. It also makes it difficult to see any detail of the ground, so relating the green swirlyness to a practical situation would be tricky.

An interesting diversion for a layman, but I would hope that doesn't apply to too many of us... We need to know what the fronts are doing, and what pressure systems are behind the wind. I can't think how this could be adapted to a format useful to most sailors. Sorry!

[conachair]

Though a handy place to see what the 500mB and 250mB heights are up to. So not useless.

[rgleason]

We would like to see grib2 support and support for dual files or a join utility first

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

[PjotrC]

Interesting intput about Earth Wind Map.

However, as Dave has observed, the main point here is just using colour coded "wind rays" in place of well known wind barbs.

Indeed, I feel a bit tired by looking at masses of digits and symbol-encoded data on so many displays... Longing for a more intuitive visualisation wherever possible.

I have no doubt that such presentation is feasible and useful for a single point-in-time snapshot. Not sure though, if a smooth transition between subsequent record sets in a GRIB file is practical enough to result in a nice movie.

[rgleason]

I agree about addition of color to the standard feathers. It would be an improvement I believe. Perhaps I his color feather feature could become an option? Maybe someone should put it in tracker?

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

[Littlechay]

Originally Posted by rgleason

I agree about addition of color to the standard feathers. It would be an improvement I believe. Perhaps I his color feather feature could become an option? Maybe someone should put it in tracker?

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

I don't agree with colouring the grib arrows. A few other programmes have tried that and it adds very little or nothing to the display.

We already have the shaded colour overlay, a few issues but it almost works; in my opinion the shaded overlay is much more useful as a visual method of displaying wind strength. ZyGrib and Maxsea TZ do it very well.

You can't use a coloured graduated overlay with coloured arrows as some of the arrows will disappear, therefore my preference is to stay with the coloured overlay with black arrows.

Chris

[rgleason]

Have not experinced the problem you mention. Do you remember what program..? The mercator projection colored seem to work don't you think? I do agree the colored overlay is working well now and arrows cannot be colored for that. I was thinking colred arrows could be a diff display option ---icing on the cake, certainly.



.

[Littlechay]

Originally Posted by rgleason

Have not experinced the problem you mention. Do you remember what program..?

The other programmes are not that memorable sorry. Polar Navy is one that uses coloured wind arrows; not pretty.

The problems have been discussed before, disappearing shading when zooming being the main one. Selection of shading for a particular data type requires going into settings rather than being able to switch between the layers from the viewing screen... yes it works but needs tweaks.