NDFD grib files on OpenCPN

[cagney]

NFDF gribs is a digitized format of the US NWS forecasts, so it is not just a computer model output. This is certainly an interesting alternative to the "normal" GFS gribs if you are in the covered area.



"In principle this should be better than any one model for the region covered, because the human forecasters that created this data have access to all models and they have chosen which they think is best at the time."
( David Burch Navigation Blog)

It should be simple to add NFDF gribs to OpenCPNs grib-plugin as these gribs now are available from saildoc. Available Saildocs Grib Data

Thomas

[sailorF54]

NDFD gribs have been available for some time (2014 I believe) from Airmail-Viewfax (not covering Alaska though)

They display fine in O

[rgleason]

Good idea

[seandepagnier]

is there a way to retrieve historic gribs?

[barnakiel]

Originally Posted by cagney

NFDF gribs is a digitized format of the US NWS forecasts, so it is not just a computer model output. This is certainly an interesting alternative to the "normal" GFS gribs if you are in the covered area.



"In principle this should be better than any one model for the region covered, because the human forecasters that created this data have access to all models and they have chosen which they think is best at the time."
( David Burch Navigation Blog)

It should be simple to add NFDF gribs to OpenCPNs grib-plugin as these gribs now are available from saildoc. Available Saildocs Grib Data

Thomas

bump +1!

I strongly disagree with David. There is no reason whatsoever why a computer/human output could be better when we weigh the human factor heavier. In fact, since we are using computers to get the heavy lifting, the opposite should be actually right (100% computer based / no human intervention models/outputs should be the most accurate ones).

There is no way a single core human forecaster can beat a 4-core Intel.

Cheers,
b.

[jstevens]

"is the"re a way to retrieve historic gribs?"

Try this link: https://nomads.ncdc.noaa.gov/ncep-charts/hires/

It is the FTP directory for pretty much every NOAA produced graphic since the mid 90's. If you want to go back further, send NOAA an email for what you want.
This is a large FTP site, so you may have to wait a bit.

John

Added: Sorry I thought there were GRIBs in that directory, but apparently just the graphics.

[conachair]

Originally Posted by barnakiel

There is no way a single core human forecaster can beat a 4-core Intel..

But he can use experience as a pointer to which particular model might be more likely the various models diverge.

The Opencpn weatherfax plug is great to compare synoptics ontop of gribs ontop of satellite photos.

[bcn]

Originally Posted by barnakiel

bump +1!

I strongly disagree with David. There is no reason whatsoever why a computer/human output could be better when we weigh the human factor heavier. In fact, since we are using computers to get the heavy lifting, the opposite should be actually right (100% computer based / no human intervention models/outputs should be the most accurate ones).

There is no way a single core human forecaster can beat a 4-core Intel.

Cheers,
b.

You want to have fun and get an idea how the forecasters are working with the models?

A daily discussion for Central Europe here:
Wetter und Klima - Deutscher Wetterdienst - Hobbymeteorologen - Synoptische Übersicht Mittelfrist
And there the "Synoptische Mittelfristprognose"

[jstevens]

On my Atlantic crossings I used this document (http://tgftp.nws.noaa.gov/data/raw/a...nm.mim.atn.txt) to help me use the GFS GRIBs. It is a text discussion by the duty NOAA forecaster on how he used the various products, including the models. It is available through SailDocs from SailMail and Winlink. That last passage was over 10 years ago, so there are likely other products available for this purpose. Also the discussion was generally limited to the N. Atlantic west of 40W, NOAA's area of responsibility.

John

[cagney]

Originally Posted by barnakiel

bump +1!

I strongly disagree with David. There is no reason whatsoever why a computer/human output could be better when we weigh the human factor heavier. In fact, since we are using computers to get the heavy lifting, the opposite should be actually right (100% computer based / no human intervention models/outputs should be the most accurate ones).

There is no way a single core human forecaster can beat a 4-core Intel.

Cheers,
b.

The common GFS model, that currently is the base for most of the gribs we use, has some well known weaknesses. The worst one being the inability to forecast tropical systems correctly. The GFS model is regularly updated, and is bound to become more accurate. Right now the European model is better, but not freely available in gribs. For now, I think that a trained human being with access to all facts, including all available computer models and knowledge of the models strength and weaknesses, as well as local knowledge, does the best job.

Thomas

[Gilletarom]

It is also possible to download this kind of grib by Saildocs.

Available Saildocs Grib Data