[Thread Prev][Thread Next][Index]

Re: curvilinear datasets



Hi Julien,

Adding a little more to Ansley's suggestions:
The command
    PLOT/VS longicrs, latitcrs
can also be very handy to help visualize the curvilinear coordinate system.
Sometimes you'll find "discontinuities" (crossing of the branch point)
embedded inside of the curvilinear grid coordinates.  In other words, as you
proceed along I holding J constant you may find that your X values look like
..., 358, 359, 360, 1, 2, ....  The 3-argument SHADE command will not handle
such coordinates as you want in these cases.  Check to see if you can "mend"
these coordinates using "IF" expressions:
    LET mended_longitude = IF j GT 50 THEN ....

    - steve

==========================

Ansley Manke wrote:

> Hi Julien,
> The appearance of data in curvilinear data sets can be od, depending
> on the grid. I haven't seen what you describe, but curvilinear data is
> fairly
> new to me and I haven't seen a wide range of examples.
>
> The regridding functions would let you put your data on a rectangular
> grid and then you can do Ferret operations such as filling or smoothing.
>
> One thing you can do to explore the data and the grid is to do plots
> of the data and coordinates separately.
>
>  yes? SHADE rain[L=1]
>  yes? SHADE longicrs
>  yes? SHADE latitcrs
>
> Sometimes visualizing the coordinates will help to understand
> more about the grid.
>
> Ansley
>
> Julien Demaria wrote:
>
> > Hi,
> >
> > I use this curvilinear netCDF dataset :
> >
> > netcdf out {
> > dimensions:
> >         time = 97 ;
> >         i_cross = 39 ;
> >         j_cross = 30 ;
> >         pressure = 22 ;
> > variables:
> >         float time(time) ;
> >                 time:units = "minutes since 2005-01-11 00:00:00" ;
> >                 time:long_name = "time" ;
> >         float latitcrs(i_cross, j_cross) ;
> >                 latitcrs:units = "degrees" ;
> >                 latitcrs:long_name = "LATITUDE (SOUTH NEGATIVE)" ;
> >         float longicrs(i_cross, j_cross) ;
> >                 longicrs:units = "degrees" ;
> >                 longicrs:long_name = "LONGITUDE (WEST NEGATIVE)" ;
> >         float rain(time, i_cross, j_cross) ;
> >                 rain:units = "cm" ;
> >                 rain:long_name = "ACCUMULATED NONCONVECTIVE
> > PRECIPITATION" ;
> >         float vnorm(time, i_cross, j_cross) ;
> >                 vnorm:units = "m s{-1}" ;
> >                 vnorm:long_name = "10-meter Speed Norm" ;
> >         float tau(time, i_cross, j_cross) ;
> >                 tau:units = "-" ;
> >                 tau:long_name = "Optical Thickness" ;
> > }
> >
> > and when I shade or fill it (fill/L=2 vnorm, longicrs, latitcrs), I
> > have tholes on borders.
> >
> > With my little knowledges about Ferret, I understand that I have to
> > regrid my dataset on a regular grid to solve this problem ?
> > For do this I need the new features "CURV_TO_RECT_MAP" and other
> > functions of Ferret 5.8, isn't it ?
> >
> > Are there some other solutions to my problem ?
> >
> > Other little problem : I try to download the new 5.8 release, but the
> > FTP always says "Control connection closed" ; are there some tricks
> > about FTP access (I believe I've successfuly downloaded the 5.7
> > release on this FTP in 2004 year, maybe you have made some recent
> > changes on your FTP...) ? Can I use a mirror ?
> >
> >
> > Many Thanks in advance,
> >
> > Julien
> >

--
º¤ø,¸¸,ø¤º°`°º¤ø,¸¸,ø¤º°`°º¤ø,¸¸,ø¤º°`°º¤º¤ø,¸¸,ø¤º

Steve Hankin, NOAA/PMEL -- Steven.C.Hankin@noaa.gov
7600 Sand Point Way NE, Seattle, WA 98115-6349
ph. (206) 526-6080, FAX (206) 526-6744



[Thread Prev][Thread Next][Index]

Dept of Commerce / NOAA / OAR / PMEL / TMAP

Contact Us | Privacy Policy | Disclaimer | Accessibility Statement