Study Images In Cartopy

This repository contains a collection of scripts to study rendering images in Cartopy.

With the new functionality in Cartopy, picking the best parameters to render an image in Cartopy is a seven-dimensional problem. To help collapse the parameter-space then I will make the following assumptions:

• dpi = 300;
• resample = False; and
• if rendering tiles then the tile provider is Thunderforest and scale = 2.

Bearing that in mind, then the problem becomes only four-dimensional. The following choices have been made (in order):

1. regrid_shape = ×2.0 (sometimes ×4.0 is better but the RAM requirements are prohibitive for universal usage);
2. resolution = "large4096px" ("large8192px" has too much detail which is not captured at regrid_shape = ×2.0 and so it actually looks worse);
3. interpolation = "gaussian" (see MatPlotLib documentation of interpolations for imshow()); and
4. when calculating a tile zoom level based off a map's resolution res = ×2.0 (compare the contour lines between "bicubic" and "gaussian" for the Norway tile example).

Dependencies

This collection requires the following Python modules to be installed and available in your PYTHONPATH.

• cartopy
• geojson
• matplotlib
• numpy
• PIL
• pyguymer3
• shapely