INTERPLATLON Interpolation for working with latitude and longitude.
INTERPLATLON performs 2D or 3D interpolation appropriate for latitude
and longitude. It deals with issues arising from the periodicity of
longitude in addition to performing several other convenient functions.
__________________________________________________________________
2D Interpolation
Z=INTERPLATLON(LATO,LONO,[],ZO,LAT,LON) performs a 2D interpolation
from the array ZO, which has latitude varying in LENGTH(LATO) rows and
longitude varying in LENGTH(LONO) columns, onto locations (LAT,LON).
Here LAT and LON are 2D arrays of the same size. LATO and LONO should
1D arrays such that ZO is plaid; otherwise see SPHEREINTERP.
A difficulty in interpolating latitude and longitude is that longitude
is periodic, for example, if LONO=0.5:1:359.5. Then the interpolation
can fail if 0INTERPLATLON checks for LONO
being periodic, and if so it handles the interpolation appropriately by
appending the last column of ZO to its front and the last to is back.
ZO may have multiple elements along its 3rd, 4th, or 5th dimension. Z
will then have its first two dimensions matching LAT and LON, and
higher dimensions matching ZO.
Z=INTERPLATLON(LATO,LONO,[],ZO,LAT,LON,[],BOOL), where BOOL is a
boolean array with size SIZE(LAT), sets the corresponding entries of
Z to NaNs. BOOL could for example be true over land. If Z has more
than 2 dimensions, this is done for each SIZE(LAT) element of Z.
__________________________________________________________________
Interpolation algorithm
INTERPLATLON uses Matlab's INTERP2 with linear interpolation. Data
points for which linear interpolation does not return a valid value
are filled with results from INTERP2 using the nearest-neighbor method.
INTERPLATON(...,'fill'), if LAT and LON are matrices of the from output
by MESHGRID, will instead fill missing data points from the linear
interpolation with the results of a bin-averaging.
This typically affects only a small number of points, but is useful in
controlling exactly what constitutes missing data.
For example, when interpolating an ocean field from a finer grid to a
coaser grid, linear interpolation will result in somewhat degraded
continental boundaries. The 'fill' algorithm puts back in a few data
points that would otherwise be missed by linear interpolation.
__________________________________________________________________
3D Interpolation
Z=INTERPLATLON(LATO,LONO,TO,ZO,LAT,LON,T) similarly performs a 3D
interpolation, where TO is now time, with SIZE(ZO,3) being the same as
LENGTH(TO). LAT, LON, and T are all the same size.
Note that the 'fill' option does not work with 3D interpolation.
__________________________________________________________________
Usage: z=interplatlon(lato,lono,[],zo,lat,lon);
z=interplatlon(lato,lono,[],zo,lat,lon,[],bool);
z=interplatlon(lato,lono,to,zo,lat,lon,t);
z=interplatlon(lato,lono,to,zo,lat,lon,t,bool);
__________________________________________________________________
This is part of JLAB --- type 'help jlab' for more information
(C) 2018 J.M. Lilly --- type 'help jlab_license' for details