Blend two rasters together so they transition smoothly into each other using inverse distance weighting interpolation (IDW).
A buffer region of blend_width grid cells around an inner nest is created. In this buffer region, IDW interpolation is used to smoothly blend from the nest to the outer grid.
Key parameters involved in the blend are nnear the number of nearest neighbors to do the nearest neighbor query for IDW, the blend_width or number of grid cells to buffer the inner nest, and p which controls 1 / distance**p in the IDW calculation.
from blend import Grid
grid_coarse = Grid((0, 1, 0, 1), 1 / 100, values=1)
grid_fine = Grid((0.40, 0.70, 0.40, 0.70), 1/1000, values=5)
grid_coarse_2 = grid_fine.blend_into(grid_coarse, blend_width=50, p=1, nnear=28)
grid_coarse_2.plot(vmin=1, vmax=5.0)using some NetCDF DEMs with real geophysical data. Note the DEM class which is a child class of Grid.
from blend import DEM
dn1 = "PostSandyNCEI.nc"
dn2 = "SRTM15+.nc"
bbox1 = (-74.24, -73.75, 40.5, 41)
bbox2 = (-76.01, -70.0, 38, 42)
dem1 = DEM(dn1, bbox1)
dem2 = DEM(dn2, bbox2)
dem3 = dem1.blend_into(dem2, blend_width=500, p=2, nnear=28)
dem2.plot(vmin=-100, vmax=500, show=False, filename="unmerged.png", dpi=1000)
dem3.plot(vmin=-100, vmax=500, show=False, filename="merged.png", dpi=1000)