0

I have some seabed bathymetry data derived from seismic reflection profiles in a 2.5D dataset (closely spaced lines).

The data was gridded by another worker in SMT Kingdom suite; I'm not familiar with the process, but the result is quite high resolution, but contains scan line errors/striping int he direction of the survey lines.

After some extensive searching via google I'm aware that this can be fixed by using fast fourier transforms (FFT) though I have no idea how to implement this!

I have ArcMap installed on my main desktop machine and would prefer a solution on this, though I do also have GRASS on another machine.

I've attached Images to try and illustrate the problem at hand.

Raw Survey data:

http://imgur.com/tWNUr5t

Inline:

https://i.stack.imgur.com/QMgG3.jpg

I can't post the crossline image, but it is in the album on imgur

PolyGeo
  • 65,136
  • 29
  • 109
  • 338
Kieran Blacker
  • 51
  • 1

1 Answers1

1

In GRASS GIS, you can use r.fillnulls:

r.fillnulls - Fills no-data areas in raster maps using spline interpolation.

or r.neighbors, if you want to apply a filter, and then r.patch between the input raster and the filtered one, in order to replace NULLs with the filtered values.

r.neighbors - Makes each cell category value a function of the category values assigned to the cells around it, and stores new cell values in an output raster map layer.

Antonio Falciano
  • 14,333
  • 2
  • 36
  • 66
  • would that still work? the errors have values that are above what they are supposed to be, ie they are about 2-5m above the actual seabed – Kieran Blacker Oct 23 '13 at 16:24