... interpolate to a new projection?

using SpeciesDistributionToolkit
using CairoMakie

The interpolate method can be used to project data into another coordinate system. For example, we can get bird richness data for metropolitan France and Corsica, in the Eckert IV projection:

spatial_extent = (; left = -4.87, right = 9.63, bottom = 41.31, top = 51.14)
dataprovider = RasterData(BiodiversityMapping, BirdRichness)
layer = SDMLayer(dataprovider; layer = "Birds", spatial_extent...)

🗺️  A 104 × 123 layer with 9579 UInt16 cells
   Projection: +proj=eck4 +lon_0=0 +x_0=0 +y_0=0 +datum=WGS84 +units=m +no_defs

We can check out the original data:

Code for the figure

fig, ax, hm = heatmap(
    layer;
    colormap = :navia,
    figure = (; size = (800, 400)),
    axis = (; aspect = DataAspect()),
)
Colorbar(fig[:, end + 1], hm)

And project them to the more locally appropriate EPSG:27574:

proj_string = "+proj=lcc +lat_1=42.165 +lat_0=42.165 +lon_0=0 +k_0=0.99994471 +x_0=234.358 +y_0=4185861.369 +ellps=clrk80ign +pm=paris +towgs84=-168,-60,320,0,0,0,0 +units=m +no_defs"
ws = interpolate(layer; dest = proj_string)

🗺️  A 104 × 123 layer with 8678 Float32 cells
   Projection: +proj=lcc +lat_1=42.165 +lat_0=42.165 +lon_0=0 +k_0=0.99994471 +x_0=234.358 +y_0=4185861.369 +ellps=clrk80ign +pm=paris +towgs84=-168,-60,320,0,0,0,0 +units=m +no_defs

By default, this produces a layer with the same dimension as the input, and uses bilinear interpolation:

Code for the figure

fig, ax, hm = heatmap(
    ws;
    colormap = :navia,
    figure = (; size = (800, 400)),
    axis = (; aspect = DataAspect()),
)
Colorbar(fig[:, end + 1], hm)

Overwrite a layer

There is an interpolate! method that will perform interpolation on an already existing layer. This is useful if you want to rapidly bring data from a layer to something compatible with another layer.