Grib_pi questions

[rgleason]

The grib link is this

https://ftpprd.ncep.noaa.gov/data/nc.../wave/gridded/




I don't know which one Craig is using. Put I tried several.
One on the west coast and one south. I picked
gfswave.t12z.gsouth.0p25.f138.grib2
gfswave.t12z.wcoast.0p16.f351.grib2



I noticed that sometimes the data is shown as NA when the cursor is hovering over the actual graphic arrows. Also sometimes the data shown is NA when between grid points when I think perhaps it could be an interpolated data set.


Other than that I notice no problem. Can you please explain further and be more detailed and specific?



Below are some screenshots



--------------

See Post https://www.cruisersforum.com/forums...ml#post3823322


by RoyalCaribMet

Good evening,

Has anyone noticed the grib file extraction for wave period & direciton is using the incorrect value? The cursor value shows WavePeriod (yellow) and WaveDirection (orange) when in keeping with the significant wave height value, Primary Wave Period (blue) and Primary Wave Direction (purple) should be used. If I could write code I'd offer a fix myself but wanted to bring it to the grib group's attention.

Later:
Hi yes, I understand your plug in has nothing to do with the embedded grib plug in for the app. I couldn't find where to reach the app author(s) so I just stumbled onto this group and thought I'd start a conversation.

To populate the grib fields I use ftpprd.ncep.noaa.gov /pub/data/nccf/com/gfs/v16.3/gfs.20230915/12/wave/gridded with the respective date/time for the most recent run cycle. I use the Atlantic Ocean 0.16 degree resolution.

Thanks for the reply, Craig

[rgleason]

Some screenshots. Unfortunately the cursor does not show up.

[RoyalCaribMet]

Yes, thank you for starting the new thread.


The cursor derived wave period and wave direction is being pulled from the incorrect wave grib datapoint fields IMHO.



As mentioned in a separate post, while the primary wave height is used as the grid point value, the wind wave direction and period are used for the other values. To be consistent, all three should be peak wave height, period, and direction. Most mariners can mostly about those values because those are the wave elements with the most energy and are most important. If we want to get into wave spectra distributions, that's for a different program. This program OpenCPN is really quite amazing in all that it does. I have used it to brief captains for weather routing, and used it in maritime admiralty litigation as well.
Cheers

[rgleason]

@RoyalCaribMet Criag

I am not too well versed in the details of grib_pi and I really don't know how to determine what the underlying data being use is, but others on this forum and certainly the programmers can.

Grib_pi should certainly be corrected if it is determined that there is some problem as it is an essential aspect of weather and weather_routing. However we first have to understand more completely what Craig is asking for specifically and to determine what grib_pi is actually using.

Craig says regarding wind grid data:
1. Location: Primary wave height is used as the grid point value.
2. Height: Primary wave height
3. Period: What is being used now?
4. Direction: What is being used now?

Craig would prefer:
1. Location: Primary wave height or Peak wave height used as the grid point value?
2. Height: Peak wave height
3. Period: Peak wave period
4. Direction: Peak wave direction

I just would like to observe that there are several types of data:
1. Significant wave height, period and direction
2. Peak wave height, period, and direction (or some version of that)

From wikapedia Significant Wave Height

In physical oceanography, the significant wave height is defined traditionally as the mean wave height of the highest third of the waves. It is usually defined as four times the standard deviation of the surface elevation – or equivalently as four times the square root of the zeroth-order moment of the wave spectrum. The symbol Hₘ₀ is usually used for that latter definition

The additional information needed is what NOAA does, because we should align grib_pi with NOAA's standards which are here https://www.weather.gov/dlh/WaveHeightExplanation

1. "Average weight height. This is not displayed in the forecast. The average wave height is estimated to be about 5/8 the value of the significant wave height"
2. "Hs is the significant wave height. This is the average of the highest one third of waves. This is currently what is displayed in the forecast."
3. "H1/10 is the average of the highest 10% of waves observed. This is what is newly being displayed in our forecast. This is indicated by "Waves occasionally to xx feet".

IMHO Craig is asking us to "correct grib_pi" to use "Waves occasionally to xx feet" for grib wave data.

Is this appropriate, given that NOAA uses "significant wave height" too?
Should we add this data "Waves occassionally to xx feet"?

Also what does grib_pi current use? Is it correctly shown?

[RoyalCaribMet]

Hi rgleason et al,
There are many different terms and definitions used to describe waves. The GFS-Wave (formerly WW3) model output is constrained to only a few variables/fields (probably due to time/size/computational limitations).
As mariners, we are mainly concerned with Primary Wave (PW) energy, which is often described as combined wave energy but includes the dominate energy direction and period of the PW. And while HTSGW is the model value we commonly use, it is not the Hs (significant wave height) by definition, although it could be considered that from a ship observer's visual observation.
Therefore, if we are using and displaying HTSGW, "significant height of combined wind waves and swell", in our graphical platforms, then keeping apples to apples, we should associate and display with it, DIRPW "direction of combined wind waves and swell, and PERPW "mean period of wind waves and swell".
Just my suggestion, I can find and extract it in other ways. But it looked odd to me there were 6 meter swells, a thousand miles away from Hurricane Lee, with a 3 second wave period and no remarkable wind wave to speak of.
Cheers,
Craig

[rgleason]

Craig, to simplify this discussion and since you have experience with NOAA waves can you determine what grib_pi is using now and what should be changed and why?


Please note that we do not necessarily have access to all wave types when using GFS, RTOFS, ECMWF etc. We are restricted to what data types are provided!

[RoyalCaribMet]

Yes, yes, thanks. Let me find some time this week and get back to you.
Sorry for taking the long way with this.
Cheers!