Support draping images over parts of a grid with grdview

[KristofKoch]

Description of the problem

When draping a georeferenced image over some relief and they have different resolutions the image is stretched to fill the whole plot area instead of a correctly scaled placement.

The docs for grdview -G allow for different scales:

The drapegrid may be of a different resolution than the reliefgrid.

Full script that generated the error

#!/usr/bin/env bash

gmt begin draped_img
  gmt grdview -JL-24.9/16.55/16.3/16.7/20c -R-25.14/16.75/-24.8/16.95r CV.grd -Ba10mg10m -I+ -Gcaboverde_mod.tif -Qi150
  gmt plot land_polygons_osm_caboverde.bf2 -bi2f -W0.5p,red
gmt end show

The files CV.grd, caboverde_mod.tif and land_polygons_osm_caboverde.bf2 are provided in
data.zip (467 KB)

CV.grd is a subset of the GMT provided 3 arc second global relief, caboverde_mod.tif is a simplified georeferenced tiff image, and land_polygons_osm_caboverde.bf2 is the coast outline from OpenStreetMap to have a precise coast line known to fit with the SRTM data.

Actual outcome

Note the mismatch between the black coast line with blue water and the shaded terrain with the red coast line. The image (blue water and black coast line) appears to be stretched to fill the whole plot area instead of being correctly sized.

Expected outcome

I expected the image to be scaled (shrunk in this case) to the correct size when draped over the relief grid and not stretched to fill the whole plot area.

System information

• Operating system: macOS 10.15.4
• GMT version (gmt --version): tested in
  • 6.0.0
  • 6.1.0_9c88924_2020.05.06

[KristofKoch]

I missed to provide the context – it came from the forum.

[joa-quim]

Well the docs only promise that different resolutions are allowed but in this case the image covers only a sub-region of the grid.

Off course this case should be allowed too but grdview is awfully complicated and it processing the images as if they were grids, which adds a lot of memory overhead (each byte is represented in a float).

Using 3 grids to represent an image is something we should really drop off. It comes form last century when we couldn't deal with images.

[stale]

This issue has been automatically marked as stale because it has not had activity in the last 90 days. It will be closed if no further activity occurs within 7 days. Thank you for your contributions.

[stale]

This issue has been automatically marked as stale because it has not had activity in the last 90 days. It will be closed if no further activity occurs within 7 days. Thank you for your contributions.

[stale]

This issue has been automatically marked as stale because it has not had activity in the last 90 days. It will be closed if no further activity occurs within 7 days. Thank you for your contributions.

[KristofKoch]

Is there still interest in placing images on parts of a grid? Like a georeferenced satellite images over a larger grid to provide context to the image? I could still make use of this …

[maxrjones]

Is there still interest in placing images on parts of a grid? Like a georeferenced satellite images over a larger grid to provide context to the image? I could still make use of this …

I agree that this would be useful.

[KristofKoch]

Great. @meghanrjones, would you recommend to keep this issue open or write a new feature request?

[maxrjones]

Great. @meghanrjones, would you recommend to keep this issue open or write a new feature request?

I would keep this open, but from Joaquim's comments there may also be prerequisite refactoring of the handling of images. I am not familiar with what would be involved.