>Hello,
>I have 5 years of binary data from MM5. The data has lat2d and lon2d
>dimensions.
>I wish to plot the annual cycle of rainfall (xy plot) for a given
>location (say 40N, 65W). In a normal netcdf file I could have used
>coordinate subscripting for this.
>I need some ideas on how I can extract the location of interest. I
>converted the binary data (Lambert Conformal projection) into
>netcdf but the dimensions of my rainfall variable is as follows;
>
>(0) ==========> printVarInfo: PREC_E
>
>
>Variable: x (parameter)
>Type: float
>Total Size: 706560 bytes
> 176640 values
> Number of Dimensions: 3
> Dimensions and sizes: [time | 60] x [MLON | 64] x [NLAT | 46]
> Coordinates:
> time: [204101..204512]
> Number Of Attributes: 2
> units : mm/day
> long_name : total precip (perturbed vegetation)
> (0) Minimum: 0 Maximum: 75.4019
>
>This is still not helpful. Maybe, there is a better way to handle this.
>I should be grateful, if someone can suggest a way to do this.
The following is not tested. It (hopefully) finds the
subscript indices of the grid point nearest LAT, LON.
Given the high resolution of the MM5 grid, it is hardly worth
interpolating to a specific point.
function indMinDist(lat2d[*][*]:numeric, lon2d[*][*]:numeric \
,LAT[1]:numeric, LON[1]:numeric)
; determine subscript(s) of the nearest grid point to (LAT,LON)
local lat1d, lon1d, dist, dmin, ijmin, ij2d
begin
lat1d = ndtooned(lat2d)
lon1d = ndtooned(lon2d) ; -180 to 180
; great circle distance is best BUT
; this is close enough in this case
dist = sqrt( (lat1d-LAT)^2 + (lon1d-LON)^2 )
dmin = min(dist) ; min distance
ijmin = ind(dist.eq.dmin) ; index of desire pt in 1d
ij2d = ind_resolve( ijmin, dimsizes(lat2d) )
return (ij2d)
end
prc = .... ; (time,lat2d,lon2d)
lat2d =
lon2d =
LAT = 40.
LON = -65.
ij = indMinDist(lat2d, lon2d, LAT, LON)
ii = ij(0,0) ; lon2d
jj = ij(0,1) ; lat2d
prc_timeSeries = prc(:,jj,ii) ; time series of pts at
; grid pt nearest LAT,LON
printVarSummary( prc_timeSeries )
================
You can remove the annual cycle from the above time series.
good luck
_______________________________________________
ncl-talk mailing list
ncl-talk@ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
This archive was generated by hypermail 2b29 : Mon Feb 07 2005 - 07:50:42 MST