Mastering video verification and chronolocation for investigative journalism
How to anchor video evidence in irrefutable time and space using OSINT techniques
In the age of deepfakes and mass disinformation, video evidence is a crucial, yet highly vulnerable, component of any major investigation. A journalist cannot responsibly publish a story based on an unverified clip. Chronolocation—the process of determining precisely when a video was recorded—is the defensive tool against misinformation, allowing you to anchor a breaking news event or an exposé in irrefutable time and space.
If you learn how to do this, shaky film that was once unverified content becomes a solid journalistic fact, which is necessary for putting together a correct chain of events.
👁️ Goals for learning
Accurately geolocate a video by identifying key landmarks and unique features.
Master the use of Google Earth Pro and Google Maps Timeline to establish a historical reference point for a scene.
Corroborate visual evidence by cross-referencing weather, solar, or seasonal data.
Implement a robust chain of custody protocol for video evidence and associated chronolocation findings.
💡 Basic ideas
Key terminology
Chronolocation:
The process of determining the specific date and time an image or video was recorded based on contextual clues (weather, shadows, construction, vegetation).Geolocation:
Determining the precise geographic location (coordinates) where a photo or video was taken. The prerequisite for chronolocation.Pillars of Authentication:
The three key elements for verifying a piece of media are source, date, and location.Shadow Analysis:
An advanced chronolocation technique using the length and angle of shadows to calculate the sun’s position and, thus, the exact time and date a scene was captured.
⚠️ Important for Journalists: Ethical and legal boundaries
Consent & privacy: The “publicness” Rule
Your analysis only looks at media that the individual or group who posted it has made public. Do not try to bypass paywalls, hack into private accounts, or use tools to access APIs or data streams that are not public. The moment your research requires a login or password that is not your own, STOP.
Knowing when and where a video was made can help you find its creator. Your newsroom’s privacy policy requires you to be as ethical as possible when dealing with all such information.
Legal Considerations: Verification vs. trespass
Chronolocation uses publicly available tools like Google Earth Pro’s historical imagery, which is legal. The moment you use your findings to physically trespass or encourage others to do so, or use non-public databases/tools without explicit authorisation, you risk legal action.
Disclaimer: Before publishing any findings that include location or time data, especially if they go against official stories or involve sensitive figures, always talk to your organisation’s legal department.
🛠️ The method: Putting it into action step by step
Tools and setup needed
👷♀️ Practical execution (The “How”)
The core of chronolocation relies on identifying change over time at a fixed location.
Scenario 1: Geolocating the scene
Scenario 2: Establishing Chronology with Historical Imagery
The goal is to match a unique, time-sensitive feature in the video (e.g., construction or vegetation) with a snapshot from historical satellite imagery.
Scenario 3: Advanced Chronolocation with Shadow Analysis
Shadow analysis provides a very specific time of day and maybe even the exact date.
Identify a Shadow-Casting Object: Find a tall, clear object in the video (e.g., a traffic sign pole, lamppost, or vertical edge of a building) and its corresponding shadow.
Measure and Orient: In Google Earth Pro, measure the object's height and the length of its shadow. Note the shadow’s direction (bearing) relative to North.
Input Data to Solar Calculator: Input the geolocation (Lat/Long) and a plausible range of dates into an online shadow calculator (e.g., SunCalc).
Match: The calculator will tell you what time of day the sun would have to be at that angle to make a shadow of that length and direction on the date you chose. You have successfully changed the video’s claimed recording date if the timestamp on the footage differs between dates.
💾 The audit trail: Keeping data safe
Journalistic integrity requires an absolute chain of custody for all evidence.
Immediate Archiving: Use Archive.today to create a static, time-stamped, and verifiable record of the original posting/URL.
Local Backup: Download the video file and all extracted keyframes.
Metadata Capture: Log the following in a dedicated text file for the case:
Original URL and Archive.today link.
Include the geolocation coordinates (latitude and longitude) along with a screenshot of the annotated map.
Chronolocation range (e.g., “Must be between July 12, 2024, and August 1, 2024, based on construction.”)
Hash Generation: Generate a SHA-256 hash for the original downloaded video file. This hash is a unique digital fingerprint. If the file is ever altered, the hash will change, proving tampering. This is non-negotiable for court-ready evidence.
# Example command in a Linux environment: sha256sum videofile_original.mp4 >> audit_trail.txt🧠 Verification and analysis for reporting
Corroboration Strategy
Never publish based on a single piece of evidence. Your chronolocation findings must be cross-referenced with at least two independent sources or methods:
Technical Cross-Check: Corroborate the Google Earth Pro findings (construction status) with third-party satellite providers (e.g., Sentinel Hub, Maxar public releases) if available.
Contextual Cross-Check: If the video is geolocated, look through local news archives or social media posts from that area around the time you think it happened to find events, weather conditions, or announcements that back up what you found.
Solar Cross-Check: If your historical imagery method gives you a date range (e.g., late summer), use a solar calculator to verify the shadow analysis for that period, providing a technical loop.
Linking Data to Narrative
Translating technical findings into accessible journalistic facts is crucial.
🤖 AI Assistance in Analysis
AI can significantly accelerate the processing of large amounts of related data, but must not handle sensitive, non-public source information.
Summarising large documents or log files: If your chronolocation leads you to a large, public city council PDF about local construction projects, an LLM can quickly summarise the key sections, mentioning dates and locations.
Identifying key entities, dates, and relationships (clustering data): Feed an LLM a batch of public social media posts about the area and time frame you’ve established. Ask it to extract all mentions of “construction”, “weather event”, or “unique landmark” to quickly cluster corroborating evidence.
Translation of foreign language material: Use services like Google Translate or a non-sensitive LLM to translate foreign language signage, posters, or local news articles found during the geolocation phase.
⚠️ IMPORTANT WARNING: Don’t give public AI/LLM models source-provided or very sensitive data, like unredacted source interview transcripts or private internal files, because of the risks to data privacy and the source. Because of the high risk of hallucination, you should ALWAYS have a person check all the dates, times, and entities that AI finds. The AI gives leads and summaries, but it doesn’t give facts that can be checked.
🚀 Next steps and practice
Exercise to Practise
The “Construction Site Challenge”:
Find a recent, publicly posted video (e.g., on a local subreddit or Twitter) of a new building under construction in a major city.
Geolocate: Pinpoint the exact latitude and longitude.
Chronolocate: Use the Historical Imagery Slider in Google Earth Pro to determine the exact date range (month/year) when that construction status (e.g., structural steel complete, but no cladding) was achieved.
Corroborate: Search local public records or news archives for the groundbreaking/completion dates to confirm your window.
📖 More advanced resources
Google Earth Pro Documentation: In-depth instructions on how to use the measurement, 3D viewing, and historical imagery features.
SunEarthTools Solar Data: A powerful calculator for analysing shadows that gives you exact solar tracking.
Sentinel Hub EO Browser: You can look at recent, high-resolution public satellite images from the European Space Agency (ESA) and compare them with other images.
Heliacal Shadow Calculator: A specialised tool that focuses on checking the accuracy of shadow casting geometry.
GeoGuessr (Pro Version/Challenge): Test your geolocation skills in a high-pressure setting that is like the real thing.
✅ Important things to remember and investigative principles
Geolocation is the foundation: You cannot chronolocate until you have definitively geolocated the scene. The success of the entire operation rests on identifying the exact Lat/Long.
Look for change: Chronolocation is the science of identifying unique, time-sensitive changes (construction, seasonal vegetation, shadow patterns) and matching them to historical records.
The “stop at the login” rule: Never engage in unauthorised access. All investigative work must be strictly limited to publicly available data.
Mandatory chain of custody: The SHA-256 hash is a requirement for all evidential media. It preserves the integrity of your findings for your editor and, if necessary, a court.
AI is an assistant, not a fact-checker: Use LLMs only to process non-sensitive, publicly available bulk data (summaries, translations). A human journalist must fact-check all output to prevent the spread of AI-driven hallucinations.
🎓 Tutorials in the Deep Web, geolocation, and metadata mastery series
👁️ Coming next week…
👣 Tracking digital footprints: username & alias correlation
Uncover hidden connections! This tutorial explores identity stitching and resolution techniques. Learn to cross-reference usernames and aliases across platforms.