[Thread Prev][Thread Next][Index]

Re: [ferret_users]intelligent interpolation method

Hi Peter,
In the Ferret documentation for SCAT2GRIDLAPLACE_XY and for SCAT2GRIDGAUSS_XY are descriptions of the algorithms that these functions use.  Scat2gridLaplace uses a combination of Laplacian and Spline interpolation, controlled by the parameter called CAY which is sent in to the function.  Scat2GridGauss uses a Gaussian interpolation.  More details are in the documentation.  An easy way to find the right discussions in the documentation in the Ferret Users guide is from the Index at http://ferret.pmel.noaa.gov/Ferret/documentation/users-guide/index-1/index/


Peter Szabo wrote:

Suppose we have additional data (they say we have it in here) for the interpolation, how can i do that in a proper way? What kind of interpolation is e.g. SCAT2GRID using? We need to know whether it is taking into account the roughness, surface relief, statistical linking between surface and variable, etc., dont we?
Someone suggested me to have a look at the method that Parry & Carter, 1998 says (Parry, M. and T. Carter. 1998. Climate Impact And Adaptation Assessment. London: Earthscan Publications Ltd.). Have you heard of this?

Regards, Peter

On Thu, Jun 4, 2009 at 9:52 PM, Ryo Furue <furue@xxxxxxxxxx> wrote:
Hi Peter,

| it is not putting more information into the grid, but interpolates
| to a smoother one.
| I need an interpolation method which uses a surface and interpolates
| the data onto it: suppose i have a data set which highest surface
| point for Hungary is around 400 metres. On a smoother resolution of
| a model the highest surface point is at 800 metres. And during the
| process an intelligent interpolation should take into account that
| the surface is not that same as it was in the coarse resolution
| data.

>From your description, I can tell what result you want.
But, I can't tell what additional data you have.

Suppose we have a gridded topography map at a resolution
of 0.1 x 0.1.  In it, the highest point of Hungary is 400m,
say.  Can we convert it to a 0.01 x 0.01 map in which the
maximum altitude of Hungary is 800m?  How do we know "800m"
is the desirable value?  How do we know where that altitude
(800m) is located?   We need an additional set of data
to modify the 0.1 x 0.1 topography.

To proceed, let me assume that you have some additional data

  lon    lat    altitude [m]
  19.00  47.53  800
  87.63  29.69  8848
  37.36  -3.07  5891.8
  . . . . . . . . . . .

You could convert your original 0.1 x 0.1 gridded data
into "station" data (scattered data), mix it with the
additional datapoints above, and regrid it, using
one of the SCAT2GRID functions.

This method may be highly inefficient.  I hope there are
better ones.


[Thread Prev][Thread Next][Index]

Contact Us
Dept of Commerce / NOAA / OAR / PMEL / TMAP

Privacy Policy | Disclaimer | Accessibility Statement