GRIB 2 file format - current status?

[Moo]

What is the current status of OCPN's ability to display GRIB 2 format files?

I did a search and the last I came up with some old discussions on including better/more lat lon projection handling into 4.8

I ask because I came across this page with GRIB data of the ASCAT satellites (actual wind data measured by satellites) but OCPN reports no valid data when I try to open the files.

Index of /data/ascat_ab

It would be great if this data could be accessed via the GRIB plugin too.

[did-g]

Hi,


GRIB2 should be ok, mostly. There's a couple of issues with time intervals and units.


Index of /data/ascat_ab is not a grib2 vs grib1 issue, Openpcpn doesn't know what to do with 'categorical rain'.

[sailorF54]

OpenCpn is not alone...
These gribs fail to open with Xygrib, QtVlm and Squid...

[Moo]

Hmmm

Seems odd, but it is what it is. Pity

[Moo]

Further research shows that there are several apps out there that read the files I referenced above just fine though. So standards compliant they may not be but readable they are.

Here's a screenshot I just grabbed from Expedition.

[did-g]

Originally Posted by Moo

Further research shows that there are several apps out there that read the files I referenced above just fine though. So standards compliant they may not be but readable they are.

Here's a screenshot I just grabbed from Expedition.

Hi,


These files are ' standards compliant', they use 'Reserved for Local Use' parameter values.


IMO it's not about standards but about what to do with data and how to display them in a way which is understood and without misleading whoever is looking at them.

OpenCPN is not a generic GRIB viewer, in the standard there's thousand of tuples (discipline, category, parameter) * by x levels * by centers each with their own quirks.
cf for an example:
LuckGrib - Sea level pressure in GFS


OpenCPN only displays a subset of them, of course (x,y,z) or (a,b,c) can be added.

[rgleason]

I just added these two lines to the Grib_pi wiki.

To learn more about the various grib formats and models available, see this Luckgrib (Apple App) page. See this Introduction to NBM Oceanic Winds which introduces probability to the results. This is a new resource for the weather community to consider.

[Moo]

Originally Posted by did-g

IMO it's not about standards but about what to do with data and how to display them in a way which is understood and without misleading whoever is looking at them.

OpenCPN is not a generic GRIB viewer, in the standard there's thousand of tuples (discipline, category, parameter) * by x levels * by centers each with their own quirks.

OpenCPN only displays a subset of them, of course (x,y,z) or (a,b,c) can be added.

OK I'm with you now. I misunderstood and thought you and others were saying that the files were incorrectly formatted.

I see being able to view the ASCAT files overlayed on a forecast GRIB, as can be done in Expedition for example, as a useful check of the Forecast's initialisation values and/or forecast accuracy for the area of concern.