Kriging is a geostatistical interpolation method for sparse data, that also provides a measure of confidence over the interpolated area (i.e. a spatial distribution of confidence in the interpolation). It’s used for interpolated geologic surfaces (e.g. the elevation of the top of a particular geologic unit), hydraulic/geologic property distributions, and contaminant plume distributions. A decent intro to it can be found here: Kriging Interpolation - The Prediction Is Strong in this One - GIS Geography
As a person with an engineering degree coming into the field of hydrogeology with actually little geologic background, I had a hard time understanding how this interpolation routine actually works, and just what the heck I was doing when I changed input parameters like “range”, “sill” and “nugget”, or what effect the addition of new data points might have, etc. etc. Enter “Bokrige”, a tiny little kriging sandbox I built for me to play in. It allows for me to import/modify a small dataset of XYZ data, adjust kriging input params, and see the corresponding effect on the interpolated surface.
This leans fairly heavily on https://oeo4b.github.io/ , so big shoutout to Omar Olmedo for writing out the hard JS stuff for me, and of course thanks to bokeh devs that provide the tools to bring it to life 