Shapefile, KML, GeoJSON, and CSV: Which File Types Matter for Exploration Mapping?
Shapefile and GeoJSON carry boundaries (claims, targets), KML comes from Google Earth and some government portals, and CSV carries point data like drill collars. Exploration Maps imports all four — when you get a choice, ask for GeoJSON.
Quick reference
| Format | Extensions | Best for | Watch out for |
|---|---|---|---|
| Shapefile | .shp + .dbf + .shx + .prj | Claims/target boundaries from GIS and registries | It's 4+ files — a missing .prj loses the projection |
| GeoJSON | .geojson / .json | Everything — the cleanest interchange format | Large files can be slow; simplify heavy geometry |
| KML/KMZ | .kml / .kmz | Google Earth sketches, some registry exports | Styling is discarded on import; geometry survives |
| CSV | .csv | Drill collars, samples — any point list | Coordinate column naming and UTM vs lat/long |
Shapefile — the GIS incumbent
Decades old and still what most registries and consultants hand over. A 'shapefile' is really a bundle: .shp (geometry), .dbf (attributes), .shx (index), .prj (projection). Keep the set together — especially the .prj, without which no tool can know where your data belongs. Convert to GeoJSON at mapshaper.org (drag the whole set or a .zip in, export GeoJSON) for the smoothest import.
GeoJSON — ask for this one
One text file, human-readable, no companion files to lose, imports directly. When a consultant or data room asks what format you want, say GeoJSON in WGS84. Its only weakness is bulk — survey-grade boundaries carry more vertices than a presentation map needs, and mapshaper's Simplify fixes that in seconds.
KML — the Google Earth format
KML (and its zipped sibling KMZ) is how boundaries get sketched and shared by people who don't run GIS. Geometry imports fine; Google Earth's styling (icons, colours) doesn't — you'll restyle with layer roles anyway, which is faster than fixing inherited styles.
CSV — points with coordinates
A spreadsheet with latitude/longitude columns is the natural format for drill collars, soil samples, and showings. On import you map which columns are coordinates and identifiers — hole ID and coordinates are enough; extra columns (depth, azimuth, assay) survive for labels and callouts.
For the collar-specific workflow — column mapping, callouts, badge labels — see How to Import CSV Data into a Mining Map.
The universal fix
-proj wgs84), simplifies heavy geometry, and exports clean GeoJSON. It untangles 90% of exploration data problems.