Estimate photo location from sun position and shadows

The sun's apparent position in the sky is a function of latitude, date, and local time. At any moment on Earth, solar azimuth (compass direction from observer to sun) and solar elevation (angle above horizon) follow predictable astronomy. A photo showing a vertical object's shadow therefore encodes partial information about where and when the image was captured—if you can measure shadow length and direction reliably.

Shadow analysis rarely yields a unique coordinate. It excels at elimination: ruling out wrong hemispheres, impossible time-of-day claims, or seasonal mislabels in viral posts. A long shadow at noon suggests high latitude in winter or low sun angle near sunrise—not a tropical beach at solar noon.

This method is educational and investigative, not forensic-grade without calibrated metadata. Phone clocks drift, social timestamps reflect upload time not capture time, and daylight saving transitions confuse naive calculators. Treat sun/shadow output as a hypothesis filter you corroborate with maps, EXIF when available, and independent visual clues.

Solar azimuth is measured clockwise from north (0°) through east (90°), south (180°), and west (270°). In the northern hemisphere, shadows at local solar noon point roughly north (sun south of zenith). In the southern hemisphere, noon shadows point roughly south. Morning shadows point west of the object; afternoon shadows point east—reversed in intuition many beginners get wrong.

Shadow length relative to object height reveals sun elevation. A one-meter pole casting a one-meter shadow implies roughly 45° elevation. Very short shadows mean high sun (tropics at noon, or summer mid-latitudes). Very long shadows mean low sun (winter, dawn, dusk, or high latitudes).

The shadow tip points away from the sun. Draw a line from object base through shadow tip extended outward—that line runs toward the sun's azimuth from the camera's ground plane perspective, modulo slope and vertical object tilt. Leaning poles, curved tree trunks, and wide tree canopies distort simple stick-and-shadow math.

Step one: identify a near-vertical reference—person, lamppost, signpost, building corner. Step two: determine whether the photo was taken near local solar noon or morning/evening. Social posts sometimes include 'good morning' text; EXIF DateTimeOriginal helps; otherwise assume unknown and widen error bands.

Near solar noon in the northern hemisphere, shadows fall northward (sun toward south). Near solar noon in the southern hemisphere, shadows fall southward. Equatorial zones at equinox can produce ambiguous short shadows—combine with vegetation and climate cues.

If someone claims a photo was taken at 3 p.m. local in July but shadows suggest sun almost setting behind the camera while facades face east, either the clock is wrong or the caption location is wrong. Viral misinformation often fails this consistency check before you identify a single landmark.

Indoor photos with window light can mimic outdoor sun direction—distinguish direct sun outdoors from diffused skylight through glass. Window mullion shadows indoors follow exterior sun but compress contrast; do not apply outdoor pole-shadow formulas without confirming outdoor capture.

Solar elevation at noon on a given date varies strongly with latitude. On June 21, solar noon elevation in London (~51°N) is roughly 62°; in Singapore (~1°N) it is about 67° with less seasonal swing. On December 21, London noon elevation drops near 15°—long shadows even at midday.

If you can estimate sun elevation from shadow length and you know the calendar date (from EXIF, foliage, or event context), you can bracket latitude using sun calculators or the interactive estimator on this page. Two latitude solutions can exist for some elevation-date pairs—always cross-check hemisphere from shadow direction.

Seasonal vegetation validates date assumptions: bare deciduous trees suggest winter in temperate zones; snow on ground confirms cold season but not hemisphere alone (mountains in both Chile and Japan carry snow). Combine sun math with biome signals before publishing latitude conclusions.

Polar regions during midnight sun or polar night break naive noon-shadow rules—if shadows are short at clock midnight, you are in Arctic summer; educational sun tools may need explicit date ranges for extreme latitudes above 66°.

Given a fixed location and date, shadow azimuth rotates through the day as the sun crosses the sky. Morning: shadows point west of vertical objects (sun in east). Afternoon: shadows point east (sun in west). The rate of azimuth change is fastest near sunrise and sunset, slowest near noon.

When EXIF GPS exists but timestamp is missing or disputed, shadow azimuth at known coordinates can estimate local capture time within tens of minutes if object verticality is good—useful for sequencing multi-photo timelines in investigations.

Wide-angle lenses distort shadow angles near frame edges—measure shadows near image center or correct for perspective. Drone oblique shots require projecting shadows onto ground plane models; simple phone-photo stick math fails on steep aerial views.

Cloud cover diffuses shadows and softens edges—low-confidence flag. Wait for hard-shadow moments in video keyframes if analyzing stills extracted from footage; pick the clearest frame for measurement.

Night sky geolocation uses star field rotation and constellations visible at latitude—Polaris elevation approximates northern latitude; Southern Cross region implies south of equator. Moon phase and position add date constraints when combined with ephemeris tools.

Consumer photos rarely capture enough stars for plate-solving without long exposure; moonless rural night shots with EXIF 30-second exposure are exceptions. OSINT on typical JPEGs focuses on sun/shadow, not astrometry.

Moon shadow direction follows the same azimuth logic as sun for outdoor night scenes with moonlight—rare in urban light pollution. Do not confuse artificial floodlight shadows (multiple directions) with single-source moon or sun.

For breaking news, treat viral 'blood moon over city X' claims skeptically—composite art layers moon at impossible scale. Ephemeris check: was the moon even above horizon at claimed time in claimed city?

Unknown or wrong timestamp is the dominant failure mode. Upload time on Instagram differs from capture time by hours or days. Always document whether you used EXIF DateTimeOriginal, platform post time, or witness statement—and widen uncertainty accordingly.

Sloped ground tilts shadow vectors. A shadow on a 10° uphill slope can shift apparent azimuth several degrees—significant when distinguishing morning from afternoon in narrow windows. Look for curb lines and building verticals to assess ground plane.

Reflective surfaces and fill flash erase shadow evidence. Event photography with on-camera flash at night produces weak environmental shadow cues. HDR processing can compress shadow direction cues by blending exposures.

Composite and AI-generated images may invent consistent shadows—or fail consistency across objects. Always check multiple shadow casters in one frame; inconsistency is stronger evidence than a single perfect shadow is proof of authenticity.

Daylight saving and timezone errors: a photo EXIF stamped UTC without offset conversion produces wrong local sun calculations. Convert to true local solar time or standard time with correct offset before running calculators.

Sun/shadow analysis sits mid-pipeline: after EXIF GPS parse (if present) and before or parallel with AI region ranking. Use it to validate or falsify caption claims, not as sole geolocation output.

Pair hemisphere inference with language on signs—Spanish text plus southern-hemisphere noon shadow consistent with Buenos Aires strengthens a hypothesis; Spanish plus impossible sun angle for claimed Madrid winter morning weakens it.

Combine with satellite geometry: if shadow suggests afternoon sun from the west and facades show lit western faces, camera likely faced east toward buildings— narrows which street segment in a known town when map orientation aligns.

whereisthis.place processes EXIF before AI vision; manual sun analysis remains an analyst skill for stripped files. Use the sun-angle estimator to teach trainees before they touch paid API credits on obvious timezone fraud cases.

In journalism verification memos, phrase sun conclusions carefully: 'Shadow geometry consistent with northern hemisphere, temperate latitude, late afternoon in October' rather than 'definitely Ohio.' Graded certainty preserves credibility when later evidence upgrades or downgrades the hypothesis.

Free sun calculators accept date, time, latitude, longitude and return azimuth and elevation—use them in reverse: fix date/time, sweep latitude until modeled shadow matches observation. Interactive estimators on this site simplify elevation-from-shadow for teaching.

Google Earth Pro sunlight slider animates sun position for a map pin and date—excellent for verifying whether a known candidate location reproduces shadow direction in a comparable Street View or user photo after you have a rough fix.

Suncalc.org and similar tools overlay sun path on map when coordinates known; for unknown coordinates, iterate candidate cities from AI rank and test each for shadow plausibility—a manual ensemble filter.

Photoshop shadow measurement plugins exist for forensic labs; OSINT analysts often use protractor overlays on screenshot duplicates—low tech, auditable in published debunk threads with annotated diagrams audiences understand.

An analyst receives a photo labeled 'evening demonstration, Stockholm, March 15' showing protesters on a plaza. No EXIF—screenshot from a social post uploaded at 22:04 UTC. Visible lamppost shadows fall roughly northwest; shadow length about 1.2× pole height.

If capture were true local evening in Stockholm (~59°N) on March 15, sun would be low in the west—shadows should point east, not northwest. Northwest shadows imply sun in southeast sky: morning, not evening, for a north-hemisphere plaza with standard orientation.

Running sun calculator: March 15 Stockholm, 09:30 local solar time yields southeast azimuth and elevation matching measured shadow within reasonable error. Claimed 'evening' contradicts shadow geometry unless plaza map orientation is highly unusual—satellite check shows main square aligned north-south; northwest shadows confirm morning.

Independent visual clue: bakery awning OCR 'öppet' (Swedish) supports Sweden but not time claim. Reverse image search finds earlier post same day at 08:12 local with same crowd configuration—original morning rally reframed as evening narrative.

Outcome: sun analysis did not geolocate Stockholm alone—language and search did—but falsified the misinformation hook ('evening' timing). Memo grade: location plausible Sweden; time claim debunked; always request original file for EXIF in future similar cases.

Teach sun/shadow methods in journalism schools and OSINT workshops as consistency checks, not as standalone legal evidence. Courts may accept expert astronomy testimony with chain of custody on originals—your Twitter thread protractor overlay is persuasion, not subpoena-grade.

Avoid overclaiming precision: 'within 5° latitude' from a phone photo shadow is often optimistic. Report uncertainty bands and alternative explanations (wrong time, sloped ground, composite).

When coaching sources, explain that timestamp and location metadata in camera apps affect later verification—separate from sun analysis but part of holistic literacy.

Combine with whereisthis.place EXIF-first workflow: free client-side GPS read when originals exist saves hours of sun math when coordinates were embedded all along.

At 40°N (New York, Madrid, Beijing), summer noon shadows are short—roughly 30–40% of object height in June. Winter noon shadows stretch to 150% or more of object height in December. A viral photo claiming 'midsummer heatwave' with winter-length noon shadows in a temperate city should trigger timestamp or location skepticism before you invest in landmark matching.

At 0–10° latitude (Singapore, Nairobi, Quito), seasonal shadow swing is smaller year-round. Midday shadows stay relatively short; time-of-day discrimination from shadow length alone is harder—lean on vegetation seasonality and cultural event calendars instead.

At 55°N (Edinburgh, Copenhagen), summer golden hours extend late into evening with long but still northerly-pointing shadows before true polar-day extremes. Mislabeling 'midnight sun party' photos from Edinburgh as Arctic Norway fails sun elevation caps—June sun still sets briefly at 55°N.

Southern hemisphere seasons invert: December brings short shadows to Sydney and Cape Town; June lengthens them. Analysts in northern countries routinely forget to flip season assumptions when debunking global south content—document hemisphere before interpreting foliage plus shadow together.

Import photo into any editor allowing straight-line angle measurement (Photoshop ruler tool, GIMP measure, Krita). Draw line from object base along shadow centerline toward tip—read angle relative to image bottom edge, then adjust for camera roll if horizon visible.

If horizon level, camera roll correction is simple rotation. If horizon tilted due to uphill shot, estimate ground plane from multiple vertical edges (door frames, window mullions) and average their deviation from true vertical.

For human subjects, shadow length from feet to head shadow tip works when person stands on flat pavement. Crouching, jumping, or leaning invalidates height baseline—pick lampposts or bollards instead.

Video frames: scrub to moment when sun not intermittently blocked by clouds; export highest-resolution still. Phone video 1080p still beats screenshot of player—compression artifacts blur shadow edges.

Publish debunk graphics with annotated lines audiences can reproduce—transparency beats black-box 'expert says wrong sun' assertions in community trust metrics.

Forward mode: given latitude, longitude, date, local time → azimuth and elevation. Reverse mode for OSINT: given date, time, observed azimuth/elevation → latitude band solutions. Longitude rarely comes from sun alone unless time zone offset known from other clues.

Equation-of-time and atmospheric refraction corrections shift apparent solar position by minutes near horizon—negligible for rough hemisphere checks, relevant when claiming ±5-minute capture time from shadow at sunset.

Standard time vs solar time: civil clocks use time zones up to ~30° wide; solar noon differs from 12:00 clock by longitude offset within zone plus equation of time. Analysts testing caption 'exactly noon' should allow ±30 minutes in wide zones like US Mountain Time.

Mobile apps (Sun Surveyor, PhotoPills) overlay sun path on live camera—field verification tool when visiting candidate sites, not substitute for remote analysis of unknown uploads.