Geolocate a photograph
Place an image on the map using nothing but what is visible in the frame — signs, skyline, terrain, and shadows.
- 01
Do not trust the caption
Assume the stated location is wrong until proven otherwise. Your job is to derive the location from the pixels.
Output: A clean slate.
- 02
Inventory every clue
List all readable text, language, licence plates, phone prefixes, brand names, architecture, road markings, vegetation, and terrain. Zoom to the maximum.
Output: A clue sheet.
- 03
Reverse-image search
Run the image (and tight crops of distinctive regions) through reverse-search engines to find earlier or higher-resolution copies and any prior captions.
Output: Earlier copies and candidate locations.
- 04
Read the light
Use shadows and a sun-position tool together with the photo’s claimed time to estimate the compass heading and latitude band.
Output: A time/heading constraint.
- 05
Match on satellite and street
Search satellite imagery for the terrain and road layout, then confirm on street-level imagery. Match until at least three independent features line up.
Output: Coordinates on the map.
- 06
Pin it and date it
Drop the location into the Mapping workspace, attach the source and your confidence, and note the estimated capture date.
Output: A documented, mapped finding.
Field notes
- Three matching features beats one convincing one — force yourself to find corroboration.
- Power lines, pylons, and bus-stop signage are quieter, more reliable anchors than big landmarks.
- If it will not resolve, write down exactly what blocked you; that is a result too.