Amid restrictions on optical satellite imagery, researchers have developed radar technology to assess building damage in Gaza.
Damage Analysis in Gaza Using Radar
One analyzed project reviewed in over 120 articles indicated that Israel may have damaged more than half of the buildings in northern Gaza by mid-November. Instead of using optical images, the assessment relies on publicly available satellite radar data and a specific algorithm developed to infer building damage. This method is considered an innovative scientific solution that helps journalists and the public navigate the fog of conflict.
How Radar is Used to Assess Building Damage
Radar illuminates the area with ultrasonic waves. Satellites from the Sentinel-1 group collect data from about 700 kilometers (435 miles) in space. The waves emitted from each radar satellite behave similarly to the flash of a camera. The radiation then bounces back from the area and returns to the sensor. This echo is sensitive to the structure and arrangement of objects on the Earth’s surface.
The first step is to obtain a database. We analyze dense sets of satellite radar data collected over long periods to classify parts of the imagery as stable. The second step is to monitor those areas we previously classified as stable for signals indicating instability. We then developed a statistical model to classify potential damage.
The Difference Between Your Images and Google Earth Images
Satellite images on Google are high-resolution optical maps, which is different from satellite radar. Rather than looking straight down at an area, radar illuminates the area from the side. Through radar scattering, we can detect everything from tree cover to urban layouts. We do not know anything about color or the spectrum of visible light, but we have a good idea of how structures are rearranged. With branches and leaves moving with every gust of wind, the scattering pattern will change.
What we send is a network of zeros and ones – 0 means no damage; 1 means damage is present. Each file has a different time span covered. Essentially, we are sharing an abstract overlay. Journalists can compare that with other datasets and add context with baseline maps.
Challenges and Limitations
If there is a house surrounded by vegetation, we may be less sensitive to that because the plants move, and that area of the image is not stable enough for our model.
Does this mean your method is slightly better in urban areas than in rural ones? Yes. The methods work better in densely populated urban areas that do not change much, and when there is no snow accumulation or heavy rainfall or changes in soil moisture. In an area like Gaza, which has extremely high urban density and minimal rainfall, monitoring potential damage is easier than in geographies like Ukraine.
Application in Ukraine and Sudan
We have been working in Ukraine for several years. The algorithms there differ slightly from what we do in Gaza due to climatic differences and also due to the vast size of Ukraine and the much longer duration of the conflict. We have also recently applied similar methods to monitor the effects of conflict in Sudan.
Satellite Images and Reality
There is often a belief that satellite images are objective and depict the entire truth, but they do not. Any image you see is the result of personal decisions made by a specific individual to analyze data for a certain type of application.
The image is limited. What happened five minutes ago? What happened the day before? You can take a picture of your apartment, but is that the complete truth about your apartment? You may have just cleaned it, or it might be completely wrecked after a Thanksgiving dinner. The world changes. And in the context of conflict, it changes very quickly and in extremely drastic ways.
Was
We can use one image at the beginning of this conflict and one image at the end to compare before and after. Certainly, this is useful, but we will lose everything in between. We won’t have any idea about the process. There won’t be a sense of the rhythm of the damages.
Even with the data we have, we know nothing about what happens on the human level. We don’t know what happens inside or underground. We don’t know what happens to the atmosphere with the enormous amount of pollutants being produced.
We’ve used this analogy before: satellite images are a mirror that enhances and distorts. There’s no way you can get everything in one image or through one method. In understanding landscape changes and displacement or potential economic loss, a satellite image is just the simple tip of the iceberg.
Failure to Provide Objective Images
Although we make decisions in our analysis, we strive for neutrality and transparency.
One thing that’s been reassuring over the past weeks is that people from all sides of the conflict are looking at the same maps. The fact that everyone is pointing to something in a shared space is crucial for any kind of peacebuilding or any ability to move forward and agree on the facts. Our images were shown on Israeli Channel 12 and Al Jazeera. They were all talking about it. This is powerful. It’s a new kind of – I won’t say truth – but it’s a rough agreement on reality.
Source: https://www.scientificamerican.com/article/inside-the-satellite-tech-revealing-gazas-destruction/
Leave a Reply