I am using ArcMap 10.5.1
I am trying to update a global bathymetry grid (at 1 min resolution) with local sounding data (resampled to 1 min resolution). The sounding data is sparse and irregular in coverage (criss-crossing lines across the area). I would like to correct the global grid with the ship sounding data. Rather than just replace the global values by the local ones, I would like to correct the global grid with the local soundings interpolated over a larger scale. I have tried spline in tension interpolation over the entire study area but the output was patchy.
I have also tried to follow the methodology from Jeffrey Evans in this post: Interpolation of multibeam bathymetry but I was unable to figure out how to filter the data using a Gaussian weighted filter.
My question is - How can I smooth a patchy spline interpolated map with a Gaussian filter (sigma = 2) so that the interpolated region is significant within 2 standard deviations of the original sampling point?