[gradsusr] set shp -poly

[Jeff Lake]

This is exactly the problem ...
the dbf is also in a format I nor GDAL has ever seen before,
could be this temp, but yet it could be this???

this is the GRIB file I'm using
http://weather.noaa.gov/pub/SL.us008001/ST.opnl/DF.gr2/DC.ndgd/GT.rtma/AR.conus/RT.01/ds.temp.bin

the RT.01 determines the cycle hour

GrADS version 2.0.1,
CentOS 5.7 64bit


and the script snippet

'clear shp'
'set lat 20 55'
'set lon -150 -45'
'set gxout shp'
'set mproj latlon'
'set ccols 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22'
'set clevs -40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 45 50 
55 60 65 70 75 80 85 90 95 100 105 110 115 120 125 130 135'
'set shp -poly rtma_temp'
'd (TMP2m-273.16)*9/5+32'
'quit'


I would let GrADS do my mapping, but I'm some what of a stickler about 
my maps looking the same
I cannot for the life of me get GrADS to match my projection
http://contours.michiganwxsystem.com/forecasts/gfs/gfsmos_d1_maxt.png
I got GEMPAK to match (and they say that is harder), but not GrADS :(

-Jeff Lake
MichiganWxSystem.com
WeatherMichigan.net
TheWeatherCenter.net
GRLevelXStuff.com


On 1/3/2012 7:43 PM, Yaqiang Wang wrote:
> It seems GrADS does not form real 'polygon' from contour lines by
> topology analysis. This is the source of the problem.
>
> On Wed, Jan 4, 2012 at 7:17 AM, Jennifer Adams<jma at cola.iges.org>  wrote:
>> Jeff,
>> The shapfile interface in GrADS creates a shapefile that contains polygons
>> that fill out the areas outlined by the shaded contours. Because of the
>> rules in the shapefile spec regarding polygon geometry, the area inside a
>> shaded contour is split up into many pieces so that individual polygons do
>> not overlap and do not contain holes (no donuts). That is why you are seeing
>> all the vertical lines. If you color in the polygons and use the same color
>> to draw the outline, you won't see the vertical lines and the plot will look
>> more normal (like the shaded contour plot in GrADS).
>>
>> Also, before you create the shapefile, you must be in the latlon projection
>> (set mproj latlon). No other projection will work. It should be possible to
>> use your GIS tool to re-project the data in the shapefile using the metadata
>> in the file provided by GrADS. From the attachement you sent, it's difficult
>> to tell what the problem is -- better to draw a plot of a variable such a
>> land mask that would clearly show whether the boundaries in the data are
>> lining up with the political boundaries (and color them in and use matching
>> outline color). If there is an issue in the way Qgis is re-projecting the
>> shapefile, then that may or may not be a GrADS problem.
>>
>> The polygons 'connect to the outside' because they are filling in the
>> lat/lon block defined by your dimension environment. You can't mask your
>> shapefile according to political boundaries, but you can choose to not draw
>> a certain contour level by using -1 instead of a color number in the 'set
>> ccols' command.
>>
>> I can't help you any further without more specific information about what
>> version of GrADS you're using, what the descriptor file looks like for your
>> data, the source of your data file (so I can grab a sample), what your
>> commands were to generate the shapefile, and what you did inside Qgis to
>> create the display in your attachment.
>> --Jennifer
>>
>>
>>
>>
>> On Jan 3, 2012, at 4:22 PM, Jeff Lake wrote:
>>
>> I believe there is a minor, okay major, flaw in the conversion to shape file
>> here is the shape file I get when converting the RTMA GRIB over (see
>> attachment)
>> the resulting polygons all 'connect' to the outside
>> this is a cint set from -30 to 140 by 5 degree's
>> this is a better result then another piece of software, but still not
>> correct..
>>
>> anybody got suggestions ??
>>
>> --
>> -Jeff Lake
>> MichiganWxSystem.com
>> WeatherMichigan.net
>> TheWeatherCenter.net
>> GRLevelXStuff.com
>>
>> <rtma.jpg>_______________________________________________
>> gradsusr mailing list
>> gradsusr at gradsusr.org
>> http://gradsusr.org/mailman/listinfo/gradsusr
>>
>>
>> --
>> Jennifer M. Adams
>> IGES/COLA
>> 4041 Powder Mill Road, Suite 302
>> Calverton, MD 20705
>> jma at cola.iges.org
>>
>>
>>
>>
>> _______________________________________________
>> gradsusr mailing list
>> gradsusr at gradsusr.org
>> http://gradsusr.org/mailman/listinfo/gradsusr
>>
>
>