Geolocating a photo for verification: landmarks, shadows, and maps

Purpose and boundaries

Geolocation is a core verification skill for journalists and fact-checkers checking whether footage really shows the claimed event. Apply it to published newsworthy material, not to locating private individuals from their personal photos, and avoid publishing details that would endanger people in the image.

Clues to extract

List everything distinctive before opening a map: shop names and phone number formats, language and script on signs, road markings and which side traffic drives on, architecture, vegetation, mountains on the skyline, and the direction and length of shadows. Each narrows the search area.

• Text: signs, plates, phone formats
• Infrastructure: road markings, poles, bollards
• Skyline: hills, towers, bridges
• Nature: vegetation and terrain
• Shadows: direction and length for time and orientation

Lab: verify a documented image

Use a public geolocation exercise or a news photo whose location was confirmed in a published fact-check. Extract at least five clues, narrow to a city using text clues, then match the skyline and street layout in satellite and street-level imagery. Record the coordinates you reach and compare them with the published answer. Use a sun position tool such as SunCalc to check that shadow direction fits the claimed time.

clue log
1 sign text in Portuguese; phone prefix suggests region
2 tram tracks with overhead wires
3 yellow-and-white tiled facade
4 hill with antenna tower to the north-east
5 shadows point north-west: late morning
result: candidate street; matched facade and tower alignment in street-level imagery
confidence: high (3 independent matches)

Reporting confidence

Report how many independent features match and which ones do not. One matching sign is weak; a skyline, street layout, and facade lining up together is strong. Archive the imagery you compared against, since street-level and satellite images are updated over time.