Use of RTOFS for Current.Gribs - Gulf of Maine

[rgleason]

I am lost in a sea of data on the RTOFS site trying figure out how to get a Current Grib to use with Weather_Routing_pi for a race from Marblehead to Halifax in July. Weather_Routing_pi has a checkbox for Currents that will use these files.

If someone has some experience with this I'd appreciate a little guidance.
Do I use "Full Volume" or "Surface" RTOFS?
Here is the beginning of the DATA down load page
WAVE Products

These are the type of RTOFS files I think we need but I can't figure out where to get them.

ofs_atl.t00z.n024.gma.grb.std - interpolated to standard lat-lon grid For Nowcast

ofs_atl.t00z.f024.gma.grb.std - interpolated to standard lat-lon grid For Forecast

Thanks
Rick

PS: Has some good general information about Gulf of Maine Tides and Currents Tides and tidal currents on the Maine Coast
but I am specifically looking for Current.Gribs!

[.Paul.]

Rick,

The I use the "NOAA Operational Model Archive and Distribution System" (NOMADS) as a source of both wind and surface current GRIB files for the Weather_Routing_pi. Here is the link

NOMADS-NOAA Operational Model Archive and Distribution System

I use the Grib Filter feature on that web page. The wind GRIBs can come from from one of the GFS data sets, and the surface current can come from an RTOFS data set (near the bottom of the NOMADS page).

These RTOFS parameters must be selected for the surface current:

UOGRD
VOGRD

Check the box next to "subregion" to download a smaller region than a whole ocean and fill in

leftlon=
rightlon=
toplat=
bottomlat=

Longitude and latitude numbers are specified in degrees. West longitude or south latitude require a minus sign.

You will find that each time you select the parameters you can download a forecast for a single time. Try it a few times, but you will find it easier to ask for the URL that gets the download. A batch file can be built using the URL as a template with each line in the batch file calling out a different forecast time.


Paul

[rgleason]

Thankyou Paul!
Thats hugely helpful. Stay tuned I may have a question or two.

[rgleason]

1. Get the GFS wind and RTOFS current grib files by sending an email request saildocs, using the walking grib method - area of 4 degrees x 5 degrees (starting with 44N,41N,72W,68W assuming 8.2 knots at a 74 degree course) , to reduce file size. The GFS is at .25 resolution, interval 3 hrs for 5 days. The RTOFS is at .25 resolution, interval 12 hrs (minimum available) for 5 days. The two emails take this form.

To: query@saildocs.com
Subject: gribauto
Message:
send GFS:44N,41N,72W,69W|0.25,0.25|0,3,6..120|=
WIND,PRESS|8,74

To: query@saildocs.com
Subject: gribauto
Message:
send RTOFS:44N,41N,72W,69W|0.05,0.05|0,12,24..120|=
CUR,WTMP|8,74

2. Wait, and then download the two saildoc response emails and attachments.

3. Save the attached files to the grib directory. RTOFS (17kb) GFS (35kb)

4. Load the RTOFS & GFS grib files. See the instructions Load two Grib Files Concurrently: Combine Saildocs wind + Current file

a. In Grib_pi open the grib file. Select both Gfs & Rtofs files, right click with mouse and pick select.
b. Note that Rtofs gribs show only every 12 hours, but there is interpolation done between.

5. Then in the weather routing configuration the following is completed.

a. Start and Finish WP Positions are selected.
b. Grib start date is selected, alternatively select "Grib Time" but make sure the grib is showing the start time needed.
c. The Constraints section is generally ok.
d. Set the time step for isochrone calculations. (Grib interval for GFS is 3hrs & no need to interpolate.)
e. Wind strength is set to 90% of grib prediction.
f. Detect land and detect boundary are checked (no boundaries are defined yet) .
g. Grib is checked. Climatology is disabled.
h. Last valid if Data Deficient is checked, so interpolation occurs.
i. Boat Edit is selected and "Alden-52-copy.pol" Polar file is selected.
j. Pick "Save as" Boat File "Summerwind.xls" whiich saves all the polar files used for Summerwind.

k. The Alden-52-copy.pol file looks like this:
twa/tws;6;8;10;12;14;16;20
52;5.5;6.6;7.3;7.6;7.8;7.9;8
60;5.8;7;7.6;7.8;8;8.1;8.2
75;6.2;7.3;7.8;8.1;8.3;8.4;8.6
90;6.2;7.4;7.9;8.2;8.4;8.6;8.9
110;6.2;7.4;7.9;8.3;8.5;8.7;9
120;5.9;7.2;7.8;8.2;8.6;8.8;9.3
135;5.2;6.4;7.4;7.9;8.3;8.7;9.3
150;4.3;5.4;6.5;7.3;7.8;8.2;8.9

Questions:

I have several questions and observations.

The manual states that Current Rtof files are available for 6 hr intervals, however the grib_pi drop down only allows 12 hr minimum.

Also does Grib_pi or Weather_Routing_pi interpolate Rtofs "currents" to 3 hour intervals (ie: weather_routing allows setting 2 hr interval calcs, when the Gfs Wind file has an interval of 3 hours, for example. I believe grib_pi is directed to perform those calcs by weather_routing_pi. Do those calcs also occur to the RTOFS file? or can they made to?

If as you page through the the "current" Rtofs and the "current" gribs only appear at 12 hr intervals, and not at the other intervals, does that mean that there is no "interpretation being done?

See attached photos, please.

[rgleason]

In trying to download GFS & Rtofs files that have the exact same Interval that is short enough to be meaningful (3 hours) I tried sending the following emails to query@saildocs.com

send RTOFS:44N,41N,72W,68W|0.25,0.25|0,3,6..120|=
CUR,WTMP|8,74

and

send GFS:44N,41N,72W,68W|0.25,0.25|0,3,6..120|=
WIND,PRESS,SEATMP|8,74

and the two resultant files are attached (take off the ".pdf"!) and the Rtofs file seems to act like a 12 hour interval instead of the requested 3hr. (6hr request does the same thing) So it appears the 3hr interval is not available? Is there another good source for RTOFS current in the gulf of maine?

Additionally, I don't believe the Zygrib email service works anymore. Can anyone confirm? If so, I think the grib_pi email utility should be updated!

When these files are loaded together, using the load dual grib file technique, it appears that the Rtofs only shows 12hr interval.

[rgleason]

Additionally it appears, that a 12hr Rtofs when downloaded, appears to actually be a 24 hour Rtofs from the data that actually shows in Grib_pi. Why?

[rgleason]

Terry [yachtvalhalla] has a page about Oscar currents accessed at query@svsarana.com

OCEAN SURFACE CURRENTS
How to Get World-Wide Ocean Current Data via Email in GRIB Format-- OSCAR -- NOAA

These current files might work, but they are at
"..resolution is in 1/3 degree increments. The time-frame is a 5 day resolution.
Once a day, usually at midnight Pacific Standard Time, the Seattle office updates their simulation of ocean surface currents. This snapshot represents their best estimation of what the ocean surface current is doing. They use a number of techniques and data sources for estimating ocean current flow"

Whereas the GFS model offers .25 resolution, so there is a problem running them together I think unless there is interpolation done on both files.

They seem to be for a single time. Perhaps I need to specify more than

oscar:44N,41N,72W,68W

[rgleason]

The other approach to get current data is to use Paul's initial information, which has changed because NOAA has reorganized somewhat.

This page About the Atlantic RTOFS GRiB is also helpful.

Atlantic RTOFS GRiB file name conventions

Format
ofs_atl.t00z.[lnnn].[3d].[region1].grb - hybrid curvilinear coord. system grid
ofs_atl.t00z.[lnnn].[region2].grb.std - interpolated to standard lat-lon grid

Where
l = nowcast/forecast and surface/volume indicator:

• f - forecast surface fields only (I think this is what we need)

nnn = the forecast hour. For nowcasts, these numbers start at -23 and go forward to 000 (n-23, n-22, n-21, ... n-02, n-01, n000).

• 3d = if present then the grib file contains all the fields and on the native grid.
• region1 = the files are subsetted to a smaller region (if no region is specified in the filename then it the grib is the full Atlantic basin). The subdomain abbreviations are as follows:
• region2 = gma

We want region2 gma - Gulf of Maine

---
Grib File Name Examples:

ofs_atl.t00z.f001.gma.grb.std
1 hour forecast, lat-lon grid, Gulf of Maine region

Atlantic RTOFS GRiB Parameter Names and Definitions
The parameters Paul suggests are:
UOGRDu of current [m/s] - 40 Z-levels
VOGRDv of current [m/s] - 40 Z-levels

[rgleason]

Now in re-reading Paul's post, I will need to build a batch file to request the data at each time period... so that explains why I was only getting one time with the svsarana page, I think.

[.Paul.]

Originally Posted by rgleason

Now in re-reading Paul's post, I will need to build a batch file to request the data at each time period... so that explains why I was only getting one time with the svsarana page, I think.

Rick,

I had been using the "grib filter" on the NOMADS web page to extract forecasts of surface current from the "RTOFS Atlantic" and from the "RTOFS Atlantic HIRES" ocean models. This morning I saw your new posts on CF, and I started to download from the latest model run. I was surprised to see the latest run to be "ofs.20170321"--just over a month ago! It turns out that the Atlantic model was discontinued last month in favor of the "RTOFS global" model, but I don't see the "grib filter" for this.

Paul

[rgleason]

Thank you Paul, for checking and advising.

I think I need the second region2 format:
ofs_atl.t00z.[lnnn].[region2].grb.std - interpolated to standard lat-lon grid

To make it for the Gulf of Main
ofs_atl.t00z.[lnnn].[gma].grb.std

I think I need to have l = nowcast/forecast and surface/volume indicator be f - forecast surface
ofs_atl.t00z.[fnnn].[gma].grb.std

Next I think we need to figure out what forecast hours to ask for. I guess I will start with 000 and then increment by 3 hours until 120 (5 days x 24 hr/day=120)
nnn = the forecast hour.

ofs_atl.t00z.[f000].[gma].grb.std
ofs_atl.t00z.[f003].[gma].grb.std
ofs_atl.t00z.[f006].[gma].grb.std
ofs_atl.t00z.[f009].[gma].grb.std
ofs_atl.t00z.[f012].[gma].grb.std
ofs_atl.t00z.[f015].[gma].grb.std
ofs_atl.t00z.[f018].[gma].grb.std
etc to ...
ofs_atl.t00z.[f120].[gma].grb.std

If I obtain the 40 files and concatenate them, will that work as a 3 hour interval RTOFS grib?

Also what is the resolution of these files? Can we set it?

[rgleason]

I have a grib subdirectory with "Currents" which is left over from testing etc.
It contains data from June 13, 2013 - June 15, 2013 with 6 hr interval currents which is named rtofs-atl20120609182304999.grb (Sun 09-Jun-2013 00:00 UTC) and was downloaded using Grib_pi.

I believe we can no longer download rtofs with 6hr intervals.

[rgleason]

Ok Paul, so I guess I need to use the RTOFS Global Model somehow. There are scripts to download, but it looks pretty involved. See this page
Global RTOFS Data Access

Information about the Global Model
A Real Time Operational Global Ocean Forecast System
"Results include daily volume and 3 hourly surface fields in netCDF format with CF conventions. Some surface fields in GRIB format are also generated for internal use at NWS. "
--So I think 3 hr interval should be available.

Here is an NCEP "Z-Levels" visual loop "Latest" for the North Atlantic Ocean Horizontal Current, at level 0 meters, in 12 hour increments (f012,f024,f036,f048,f064..f0192)
Global RTOFS Z-Level Nowcasts/Forecasts

What we need is a "Currents" grib file like this with 3 hr interval, I believe.

[rgleason]

This page explains the changes best Environmental Modeling Center / Marine Modeling and Analysis Branch
"Global RTOFS
A hybrid coordinate, 1/12° global ocean model, run once a day. Each run starts with 48 hours of hindcasting and produces forecasts every 3 hours (surface values only) and daily full-volume forecasts from the initial time (0Z) out to 144 hours (6 days). "

NCEP Descriptors
Global RTOFS Product Description
We want Surface files :rtofs_glo_2ds_[n,f]XXX_BBBBB_CCCC.nc
where:
n=nowcast
f=forecast
BBBBB=[3hrly, daily]
XXX=hours (000-048 for nowcast and 001-144 for forecast)
CCCC=[diag, prog]

So
rtofs_glo_2ds_f003_3hrly.nc (f006,f009,etc through 144)

--Some of this looks like duplication. How do I get 3hr interval through 144 hrs?

I think we want "prognostic" so we add _prog to the file name.
rtofs_glo_2ds_f003_3hrly_prog.nc

"prog stands for prognostic variables: u_velocity
v_velocity
sst
sss
layer_density"We don't need sst, sss, layer_density. How to eliminate?
We just want a certain area, actually I'd like to have a walking grib.


Similar to this queried from saildocs

send GFS:44N,41N,72W,68W|0.25,0.25|0,3,6..120|=
WIND,PRESS|8,74


From the examples at the bottom
rtofs_glo_2ds_f003_3hrly_prog.nc
rtofs_glo_2ds_f006_3hrly_prog.nc
etc

[rgleason]

Actually I think XXX defines the max hours wanted so the file name would be
rtofs_glo_2ds_f144_3hrly_prog.nc

Just need surface currents and a lat/long box next.

Now going to
ftp://ftpprd.ncep.noaa.gov/pub/data/...tofs.20170524/

I see that these files are huge. More later.