Geo-guide: The 101 to verify video footage in Myanmar

A simple #OSINT case study on how to geo-locate and chrono-locate video material in an Asian country like Myanmar 🇲🇲

The situation in Myanmar is deteriorating. Mobile phone video footage can provide journalists with important open-source evidence of what happens on the ground. But in order to trust it, we have to verify it first. Shocking videos surfaced in February after the coup when police decided to use live ammunition on unarmed protesters (Source: an analysis of videos by Amnesty International’s Crisis Evidence Lab).

Last week, I was asked to work on a training case study. The starting point was a video without a description. For the purpose of knowledge sharing, I’d like to provide my insight. It’s fairly basic but holds important lessons. In the end, there is a debate on what other subsequent investigations and reporting project can do.

Sidenote: As a journalist/analyst reporting on Asia, it’s good timing to improve one’s foreign geo-location skills. There are constantly new videos surfacing that show clashes between protesters and the armed police. #OSINT expert Nathan Ruther at the Australian Strategic Policy Institute started geolocating footage and observation of clashes across Myanmar for the month of April. More can be done. With some skills, more can follow suit.

Case-study: The video footage

The video was shot from a window facing first south-west and then north-west at the location of the coordinates 14.071372065554256, 98.18962045273268 in the city of Dawei, Myanmar, presumably in the morning of February 28 in 2021, between 8:00 and 9:30 am local time.

How to arrive at this conclusion? Assuming you have no further background knowledge of the video (starting from zero, so to speak), how would you approach it?

Step 1: Narrow down country/region

It looks like Myanmar, ok. But, what features confirm it? There is the unique grey-ish uniform of police forces shooting at individuals (here), distinct for this part of the world (the look of the police van further provides a basis to review your assumption).

There is the distinct Burmese round script that can be read on local shop signs seen in the video (opposite house). There is the KBZ bank branch (towards the left of the person who filmed it), a Burmese bank, seen in the video to further narrow down the location (as well as the car, opposite side of the street, with MIFIDA printed on it, a Burmese company).

From the density of buildings in the shots, we can make an educated guess that it’s probably filmed in a population-dense environment, like a city.

Before we endeavour to look for regional-specific features to nail down the city (options listed here), let’s crawl the web to make it easier on us.

Step 2: Event analysis reverse images search

We assume that we only have the video material to go on. One option is to take a screenshot and reverse-image search on the web. I screenshotted the little police van with its blue lights on, in the right bottom corner of the video.

By using the Russian image search platform Yandex, we are led straight to a Twitter post that features the a similar video, presumably from the same source/filming camera. The handle Yandex link us to isn't the original source, it merely quotes a video from another handle. From there on we can check the ‘original’ video post (one way is via checking which video post has the earliest upload date/time and by checking other social media platforms via the hashtag #CoupFeb28, which seem to specifically refer to events around the 28th of February). The commentary suggests it’s police using arms with life rounds, shot at protesters. It’s posted by the handle called @soezeya.

We are told it’s a protest that took place in Dawei, a city in south-eastern Myanmar and the capital of the Tanintharyi Region.

With its 150,000 inhabitants, the city isn't tiny but not huge, either. There are a few major streets we can check out.

Let’s look for details again. The video material suggests this street does not have any road markings (at least not that we can spot them). The other indicators are the shadows. They set at an approx. 45-90 degree angle to the street. The first assumption is the street might run from East to West. The two big obvious candidates are Arzarni Rd and Niban Rd running from West to East, come to mind. After examining them closely, we find no similar location of building features. There is something wrong. Let’s pivot.

Step 3: Making adjustments to assumptions

The streets we tested don't match. What to do? The shadows fall from South to North in February (when the twitter handle posted the video) around midday/early afternoon, according to suncalc.org. But they don't during early mornings. In the morning, the shadows tend to set more from South-East to North-West. That means roads running from South to North are possibilities, too.

There are U Kyaw Yin Rd, Yay Rd, and Anauk Rd and a few other large roads heading South to North. After checking those, the only possible one is the north-south Arzarni Rd. Note the uniquely shaped and coloured roofs and the little green decorated gate. That’s our little street.

Another way forward is by relying on features. Aforementioned KMZ Bank lists branches on its website. Two are listed in Dawei.

Despite not showing the right bank location in the addresses listed here, it draws our attention to the fact that banks are listed in Dawei and that Arzarni Rd is involved (which are listed twice, south to north and East to West, the latter of interest to us).

Step 4: Time verification

We mentioned that going by the shades caused by the sun, it’s seems its filmed in the morning (the social media comments also match this assumption). Using Suncal.org a second time, we run an analysis of the shadows thrown by a policeman in the video (see below). For the day of February 28, the scenario suggests a time between 8:00 and 9:30 am.

Step 5: Further investigations

Now that we have the time and location, we can continue digging. There are other videos that emerged for that day on social media. One (below) showing a street corner and scenes of a female body dragged across the streets in Dawei.

Then there is the question of live-ammo. The use of such is highly controversial and was condemned internationally. It left many dead. The government response killed at least 557 and injured another 2,750 over the recent months, reports in early April claimed. New investigations may centre around similar video footage where the military shoots at protesters.

From the video, at hand, we can see a rifle loaded and shot. The standing police/military men have a short debate before the first round (in the video) is shot. From the shots here, it’s hard to identify the type. It could be a MA-1 Assault rifle, a Myanma-made clone of an Israeli Galil assault rifle (by 2009, it was reported that the Tatmadaw, the military, has replaced most of their BA63 rifles with the MA-1s).

Essentially, we have to question what made the police fire in this case. What is out of fear? Two groups of protesters, it looks like, approached from both sides. Did they coldhearted following orders and shot. Was this an ad-hock call? Can we further verify the weapons used in this case? Were protesters provided with weapons out of car boots to give the military a narrative to resort to live-rounds (as suggested by some sources)?

Did the shooting military/police officers single out specific protesters they shot at (in the video, it looks like the shooter put somebody/something in the crosshairs). These are questions that could be further investigated and verified with additional sources and footage.

Investigative journalist with a technical edge, interested in open source investigations, satellite imgs, R, python, AI, data journalism and injustice