Data Interpolation on a spherical or geographic coordinate system

I am just starting a new project where I want to use the Agents.jl package to create an agent based model. The nice thing about Agents.jl is that it supports geographic meshes or grids, like the WGS84 projection, which is the standard latitude and longitude coordinates. So I can create a geographic basemap using OSM for any location in the world.

Next, I have some data that is geographically referenced with latitude and longitude coordinates. Each point represents an incident, and I want to approximate the risk surface between points. So I take the discrete incident data and create a smooth interpolation between the points that represents the risk.

I needed to understand if there is a library to do this in Julia? I am also wondering if this is considered “kriging,” if I remember my geostatistics well enough? Of course, this would initially be deterministic interpolation, so no randomness. Could I use the Kriging tools in Julia GeoStats to do something like this. Haha, is it as simple as just matching vocabulary where interpolation on a Cartesian grid is equivalent to kriging on a spherical grid?

You may be looking for GMT’s sphinterpolate and/or sphtriangulate that you can use from GMT.jl. I’m liking to the GMT man pages and not the GMT.jl ones … because the later don’t yet exist.

And, btw, WGS84 is an ellipsoid, not a projection system.