[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Fwd: Re: Fwd: netCDF and IDV for the Deepwater Horizon]



Hi Ansley:

The followwing changes to your file would make it "CF 1.5 compliant" :

<?xml version="1.0" encoding="UTF-8"?>
<netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2"; location="E:/work/deepwater/drifter_example.nc">
<attribute name="Conventions" value="CF-1.5" />
<attribute name="CF:featureType" value="trajectory" />
<variable name="LON">
<attribute name="units" value="degrees_east" />
</variable>
<variable name="LAT">
<attribute name="units" value="degrees_north" />
</variable>
<variable name="SST">
<attribute name="coordinates" value="LAT LON TAXIS" />
</variable>
<variable name="SPEED">
<attribute name="coordinates" value="LAT LON TAXIS" />
</variable>
<variable name="DIR">
<attribute name="coordinates" value="LAT LON TAXIS" />
</variable>
</netcdf>

On 5/25/2010 5:08 PM, Ansley Manke wrote:
hi
Adding CC to Ben Domenico-

Hi John, CJ,


On 5/25/2010 1:45 PM, CJ Beegle-Krause wrote:
Hello Sam and Ansley,

Below please find a note from John Caron at Unidata and a request to
see how well their discrete sampling data proposal will work for us.
Please take a look to see if this will work for us.  Ansley, could you
make a test file and send it back to John, and cc me an Ben?  My read
is that he would like a "trajectory" which would be a glider path",
not what we think of as a trajectory forecast.

Ansley, Rutgers glider data is here I thinkl.
http://rucool.marine.rutgers.edu/deepwater/

Sam, if Ansley needs connection to get data to use for an example,
could you help her?  If you have comments on the standard, please send
to Ansley.  I'll take a look shortly.

Thanks,
CJ


We're still working on getting another example or two of data from
gliders but here is a netCDF file with a surface trajectory of drifter
measurements. I used Ferret to read data from a spreadsheet-style ascii
file and write out this netCDF file.

I've followed the "trajectory" style, as in 9.4.2, Single Trajectory in
the CF proposal.  A glider would be similar but would have Z as well as
lon and lat. The whole file is attached, but here's the ncdump header.


> ncdump -h drifter_example.nc
netcdf drifter_example {
dimensions:
         TAXIS = UNLIMITED ; // (162 currently)
         bnds = 2 ;
         STRING1_3 = 3 ;
variables:
         double TAXIS(TAXIS) ;
                 TAXIS:units = "DAYS since 1990-01-01 00:00:00" ;
                 TAXIS:axis = "T" ;
                 TAXIS:bounds = "TAXIS_bnds" ;
                 TAXIS:time_origin = "1-JAN-1990" ;
         double TAXIS_bnds(TAXIS, bnds) ;
         float BUOY(TAXIS) ;
                 BUOY:missing_value = -1.e+34f ;
                 BUOY:_FillValue = -1.e+34f ;
                 BUOY:long_name = "Buoy Number" ;
                 BUOY:history = "From drifter_example.dat" ;
         float LON(TAXIS) ;
                 LON:missing_value = -1.e+34f ;
                 LON:_FillValue = -1.e+34f ;
                 LON:long_name = "Longitude" ;
                 LON:units = "Degrees E" ;
                 LON:history = "From drifter_example.dat" ;
         float LAT(TAXIS) ;
                 LAT:missing_value = -1.e+34f ;
                 LAT:_FillValue = -1.e+34f ;
                 LAT:long_name = "Latitude" ;
                 LAT:units = "Degrees N" ;
                 LAT:history = "From drifter_example.dat" ;
         float SST(TAXIS) ;
                 SST:missing_value = -1.e+34f ;
                 SST:_FillValue = -1.e+34f ;
                 SST:long_name = "Sea Surface Temperature" ;
                 SST:units = "Deg C" ;
                 SST:history = "From drifter_example.dat" ;
         float DIR(TAXIS) ;
                 DIR:missing_value = -1.e+34f ;
                 DIR:_FillValue = -1.e+34f ;
                 DIR:long_name = "Direction" ;
                 DIR:units = "Deg" ;
                 DIR:history = "From drifter_example.dat" ;
         float SPEED(TAXIS) ;
                 SPEED:missing_value = -1.e+34f ;
                 SPEED:_FillValue = -1.e+34f ;
                 SPEED:long_name = "Speed" ;
                 SPEED:units = "m/s" ;
                 SPEED:history = "From drifter_example.dat" ;
         char SOURCE(TAXIS, STRING1_3) ;
                 SOURCE:long_name = "Source" ;
                 SOURCE:history = "From drifter_example.dat" ;

// global attributes:
                 :history = "FERRET V6.61   25-May-10" ;
                 :Conventions = "CF-1.0" ;
                 :title = "drifter example from Coast Guard SLDMB data,
Gulf of Mexico 5/2010" ;
}



-------- Original Message --------
Subject:     Re: Fwd: netCDF and IDV for the Deepwater Horizon
Date:     Tue, 25 May 2010 14:10:07 -0600
From:     John Caron <address@hidden>
To:     address@hidden
CC:     Ben Domenico <address@hidden>, Unidata netCDF Java
Support <address@hidden>
References: <address@hidden>
<address@hidden>



Hi CJ:

We are recommending that new netCDF data follow the CF "discrete
sampling" data proposal at:

 https://cf-pcmdi.llnl.gov/trac/wiki/PointObservationConventions
<https://cf-pcmdi.llnl.gov/trac/wiki/PointObservationConventions>

Unfortunately, it is not a final standard yet, and has changed a bit
in the last week. However, Im hopeful its very close. See section 9.4
for trajectory data. Any comments on that proposal would be
appreciated, in particular if it meets your needs.  If you'd like to
try putting it into that format and sending us a sample file, we may
be able to the process .

You can use any software to write netCDF, Java is not needed. The IDV
will be able to read that format (not the current release, but soon
after).

Regards,
John



Hello Ben,

I've been talking with Tiffany Vance and Steve Hankin about our data
and visualization needs for the incident.  I'm one of the trajectory
modelers with NOAA Office of Response and Restoration / Emergency
Response Division, and working on making new kinds of displays for
information.  We have the lead for modeling for the incident, and are
putting out the daily trajectory forecasts.  We are moving quickly to
3D forecasting. I have questions along two lines.

netCDF - we are pushing for using netCDF for all the information
management for the modeling team.  The two data streams are to move
things into GIS for the spatial folks, and netCDF for those of us
doing modeling and obs analysis.  We have some issues that Unidata
may be able to help with.  For example,
-we use the C libraries that always seem to be behind the java
libraries in development
-are their standards now for trajectory, cruise profiles, glider
data, etc?  We will have a large number of observational platforms,
and want to be able to leverage all the data in modeling and
forecasting.
Visualization - IDV.  I'm an IDV novice, but a big fan.  Working on
the HabitatSpace project with Tiffany, we ran into some little bugs
that caused headaches.  In looking at how to do 4D visualization for
obs and models, I'd really like to use IDV.  Is there help available
to make sure that it works for what we need?

If you need more background on our needs, or what we do, let me know.

Best regards,
CJ