Re: Map display problem

Thanks Adam. I checked and the latitudes are oriented from south to north
(-59.5 to 89.5, no data over Antarctica). Not sure what is wrong here,
but remarkably after conversion the plot seems to match up with the map
underneath better, not worse. Plot(0) is shifted longitudinally--I think
the prime meridian (0east) and international date line (180east) are
interchanged?--as well as stretched, but plot(1) is just stretched. I
tried gsnAddCyclic=False, but that only creates a small blank space down
the prime meridian in plot(1). I suspect that the issue is the lack of
data for Antarctica; tried setting res@mpMinLatF=-59 but that only made
things worse. Attaching a screenshot (.png) of the plot I get out.
Thanks again,
Lizz

> Hi Lizz,
> I don't see anything wrong with your script visually. The calculation of
> the trend looks fine to me. In the resulting panel plot, does plot(0)
> look right but plot(1) look off? If so, something is going on with the
> f2gsh_Wrap conversion. Make sure that your rc array has latitudes
> oriented from south to north (f2gsh requirement). Use
> printVarSummary(rc) before the conversion to see if that is the case.
>
> If after checking the latitudes you still are having problems with the
> conversion please write the group back, and include a .gif,.jpg, or .png
> image of your panel plot so we can see what is going on.
> Good luck,
> Adam
>
> On 06/16/2011 10:36 AM, Lizz Ultee wrote:
>> Posting script, as requested. This also returns complaints about the
>> units in the file being wrong, which I think has something to do with
>> the
>> misalignment on the map.
>>
>> ;*************************************************
>> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
>> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl"
>> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl"
>> begin
>> ;************************************************
>> ; read in data
>> ;************************************************
>> in = addfile("./Data/tws_noah10_1m_7901_1010.nc","r")
>> water = in->water
>> lat=in->latitude
>> lat@units="degrees_north"
>> lon=in->longitude
>> lon@units="degrees_east"
>> time=in->time
>> wt = water(latitude|:,longitude|:,time|:)
>> ;************************************************
>> ; create x and calculate trend
>> ; regCoef instead of regline
>> ;************************************************
>> x = ispan(0,dimsizes(wt&time)-1,1)*1.
>> rc = regCoef(x,wt)
>>
>> rc@long_name = "regression coefficient" ;modelled on NCL example
>> script here
>> rc@units = "unitless"
>> copyatt(rc,wt)
>> ;**********************************************
>> ; interpolate to new grid
>> ;**********************************************
>> newgrid = f2gsh_Wrap(rc,(/64,128/),0)
>>
>> ;=========================================;
>> ; create plots of original and regridded data
>> ;=========================================;
>> wks = gsn_open_wks("pdf" ,"regrid1")
>> ;gsn_define_colormap(wks,"gui_default") ; choose colormap
>> plot = new(2, graphic)
>>
>> res = True
>> res@cnFillOn = True
>> res@gsnSpreadColors = True
>> res@cnLinesOn = False
>> res@cnLineLabelsOn = False
>> res@lbLabelAutoStride = True
>>
>> res@gsnDraw = False
>> res@gsnFrame = False
>>
>> plot(0) = gsn_csm_contour_map_ce(wks,rc,res)
>>
>> plot(1) = gsn_csm_contour_map_ce(wks,newgrid,res)
>> ;=========================================;
>> ; create panel plot
>> ;=========================================;
>> gsn_panel(wks, plot,(/2,1/),False)
>> end
>>
>>
>>
>>
>>> offline
>>>
>>> [1] regCoef is what you want
>>>
>>> [2] You should post a *clean* version of your script or, at least,
>>> the relevant portions of the script to ncl-talk.
>>>
>>> Please do not reply directly to me.
>>>
>>> Good luck
>>>
>>> On 6/16/11 8:23 AM, Lizz Ultee wrote:
>>>> Hi NCL Talk,
>>>> I'm a new user of NCL and do not have much programming experience
>>>> besides,
>>>> so please excuse any confusion.
>>>> I'm attempting to calculate a trend for each point of a dataset and
>>>> then
>>>> interpolate this info to a Gaussian grid. There are a few issues when
>>>> I
>>>> try to do this:
>>>> 1) Error messages display when I try to get the trend using regline.
>>>> I
>>>> can use regCoef but I'm not sure this is what I need. Previously, I
>>>> calculated the trend at specific points using regline and had no
>>>> problems,
>>>> but now (calculating for each point of the dataset) error messages
>>>> about
>>>> the dimensionality come up. The range of values returned by regline
>>>> is
>>>> not the same as those returned by regCoef, and I don't know why this
>>>> should be.
>>>> 2) I have regridded what I got out of regCoef and plotted the results
>>>> on
>>>> a
>>>> contour map, original vs. Gaussian grid. It looks fine except that
>>>> the
>>>> plot is not lined up with the underlying map; the data very clearly
>>>> shows
>>>> outlines of the continents but these are in the wrong place. Does
>>>> this
>>>> matter in the grand scheme of things? I just need values on the
>>>> Gaussian
>>>> grid (in a NetCDF for example) to proceed with analysis--the plot will
>>>> not
>>>> play a part in my final product.
>>>>
>>>> Thanks in advance for any suggestions anyone has!
>>>> Best,
>>>> Lizz Ultee
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> ncl-talk mailing list
>>>> List instructions, subscriber options, unsubscribe:
>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>
>>>
>>
>>
>> _______________________________________________
>> ncl-talk mailing list
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
> --
> ______________________________________________________________
> Adam Phillips asphilli@ucar.edu
> NCAR/Climate and Global Dynamics Division (303) 497-1726
> P.O. Box 3000
> Boulder, CO 80307-3000 http://www.cgd.ucar.edu/cas/asphilli
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
>

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk