View from above: what role do satellite images play in the Russian-Ukrainian war

When talking about the use of satellites during the current stage of the war (since February 24), one remembers the ICEYE satellite acquired by the Prytula team or the use of Starlink for communication at the front. However, there is a less visible, but nevertheless important, work of satellites, the results of which during the war are available to the general public. This is taking pictures of the destruction in cities, the consequences of hostilities for nature, and even exposing the lies of the Russian authorities regarding important geopolitical issues.

Such work is not operational – its results are not required here and now and do not become obsolete in a few days. At the same time, the “view from above”, which is not limited by borders, is constantly applied where information is methodically processed.

Let’s look at several areas where satellite images are used in wartime, and also find out how they are used in DeepState to keep the front line on the map up-to-date.

NPP occupation

In violation of all international agreements, the occupiers use nuclear power plants for military purposes. At the beginning of the full-scale invasion, it became known that the Russians were digging trenches in the “red forest” near the Chornobyl power plant. However, Zaporizhzhia NPP, which is still under the control of the occupiers, is currently in a worse situation.

Satellite images allow us to see, how the Russians make fortifications on reactor units. Despite the fact that the IAEA warns against military activity around the ZNPP, the occupiers “involve the reactors in tactical defense planning”, writes the British Ministry of Defense, which released the pictures.

Destruction tracking

Satellite images also help to reach places that independent journalists cannot access. In particular, occupied cities or cities where active hostilities are taking place. From above, you can assess the number of destroyed houses, and the general state of the cities and compare it with the pre-war situation.

Using Planet Labs data, journalists were able to show how Severodonetsk, Lysychansk, Izyum, Lyman, Avdiivka, and surrounding villages have changed. Analysis of data from the satellite made it possible to recognize the destruction and create an interactive map where it is displayed.

Environmental consequences

As well as the destruction of cities, the damage to the environment caused by the Russian occupiers is clearly visible from satellites. This is almost the only way to assess how the war affected the ecological condition of the temporarily occupied territories.

Skhemy program journalists using satellite images noticed the consequences of the fires, which arose due to shelling, in the forest near Izyum of the Kharkiv region, on the Kinburn Spit (Mykolaiv Oblast) and near the Zaporizhzhia NPP. Numerous fires also damaged fields in the Kherson region.

Probably the most famous result of tracking the result of the shelling were the pictures of the fields of Donetsk region, covered with funnels from the bombings. These photos not only clearly demonstrate how massive the shelling was, but also give ecologists an opportunity to assess the contamination of the earth by the remains of shells.

Exposure of lies

Satellite images help to investigate cases where there are no or insufficient witnesses. For example, in the case of blowing up of “Nord Stream” and “Nord Stream-2” in the Baltic Sea. The Russian side calls it sabotage, however, witnesses and satellite images prove that it could be a Russian ship with a mini-submarine on board.

Probably the most extensive project to use satellite imagery during wartime is the DeepState map. It is updated daily by analyzing a huge array of data from satellites that are obtained from open sources. We spoke to one of DeepState‘s analysts to find out exactly how they do it.

The DeepState case: how satellite images help update the virtual frontline

How are frontline changes tracked to map them?

Every day, both sides publish a lot of videos from the front. First, we try to confirm the relevance – that a certain video was actually taken today or yesterday, and not a month or a year ago. After that, we geolocate it.

On the basis of geolocations, you can roughly build a map of combat operations. If, for example, in the video, you can see how Ukrainian tanks are already driving behind a forest strip, and this video is from today – and yesterday this forest strip was marked by the enemy on the map, then you can advance the front. In this way, each geolocation found is a probable change on the front, which can be recorded and entered on the map.

What is the role of satellites in video geolocation?

Geolocation works like this. We take the video from the front and take some other material as a basis. It can be a satellite image, a picture in Google Street View, a photo from the Internet, or a video on YouTube, which also shows the territory that is in the video from the front.

Then we find the correlation between the two materials. If the correlation is confirmed, then the location on the video is indeed equal to the location on Google Maps or the YouTube video. After that, we draw these correlations on the map and thus confirm the reliability of the location.

The problem is that most videos from the front are not shot in cities, and simply opening Google maps, throwing the yellow man somewhere on the street, and seeing what is around will not work. It is mostly impossible to geolocate a village using photos from the Internet or Google Street View, because there are either not enough photos of this village on the Internet or none at all.

When we geolocate forest belts, fields, or villages, the only thing we rely on are satellite images. Without them, it would be impossible to geolocate 90% of videos from the front.

It is quite simple. Since most of the videos are shot from drones, we can already see what a certain location looks like from a height. Using the analysis method, you can narrow the search circle to a certain region, and then use satellite maps to find a place that will coincide with the place on the video.

Sometimes we receive materials where the shooting is not from a drone but from the ground. But in this case, you can also draw what the territory will look like from above. Power lines, electric poles, lights, roofs of houses, forest strips, bushes, and roads are taken into account. Even such a detail as a pillar is clearly visible on satellite images. In sunny weather, it casts a shadow – and if the pillar cannot be seen from space, then you can see a shadow that will indicate its presence.

I will tell about one case of narrowing the search circle. Once (in the video) there was a location with an open field. There is nothing, but a forest that can be seen from afar. One guy from the team immediately said that the forest in such a field was most likely artificial, and was planted there to avoid flooding. So, there is a huge body of water nearby. Thus, we found that this location was indeed in the Zaporizhzhia region near the Dnipro.

From what sources do you get satellite data?

They are obtained from sources that are freely available and can be opened by anyone on the Internet. There are many different services. One of them is Google Earth. Regardless of the resource, they all show more or less up-to-date satellite images that can be used to search for a location.

What interesting things have you seen from satellites?

I will say this – you can see a lot of interesting things from satellite images. Undoubtedly, this information is seen every day and is used by our General Staff in order to harm the enemy. Most likely, they use satellite images that were taken as recently as possible and were provided through some communication channels. Here we are not talking about open-source intelligence, but much more serious intelligence.

We, in turn, use only open resources. If we find a concentration of troops, tanks, or air bases, then, most likely, they have long been either eliminated or moved to another place. We work with maps that were updated neither today nor yesterday, but, for example, three months ago. It is obvious that this information is no longer relevant.

Більше коментарів!