InSAR and AI Safeguard Dams from Collapse

Wednesday, 22 January 2025
Reading Time : 4 minutes

Dam collapses can be catastrophic events for people, infrastructure, buildings and natural ecosystems. Now, better access to remote sensing data from satellites is helping to keep an electronic eye on dams all over the world, providing early warning of potential problems. These techniques protect against collapses of dams of many kinds, including those at mine tailings storage facilities, hydro-electric installations, and reservoirs for human and agricultural use.

Big improvements in surveying technology

While collapse of a dam may happen within seconds, the dam may have been putting out warning signs for years – in the form of small-scale displacements of the dam that show that failure is imminent.

Traditionally, the dam’s operators will search for these warning signs by means of ground surveys – an approach that uses periodic measurements, perhaps annually, to monitor for deformations in the dam that hint at potential problems.

But there’s growing interest in supplementing this method from agents located hundreds or thousands of kilometres above the earth. Satellite imagery is not a new technology, but it is constantly improving.

Satellite-based solutions can monitor dams in any part of the world, as well as other ground features such as the slopes of open-pit mines and quarries, slopes adjacent to roads and rail lines, buildings, bridges and other infrastructure.

InSAR technology provides insights on the ground

Interferometric Synthetic Aperture Radar (InSAR), uses orbiting satellites to collect radar images of the earth’s surface. Unlike visible or infrared light, radar waves penetrate most clouds and are equally effective in darkness. Radar images from the same area are collected on two or more different passes of the satellite over that area, and the results compared to see if there are any changes. This movement can be portrayed as a “picture” – not of the surface itself, but of how much the surface has moved or deformed during the time between images.

In many cases, InSAR data is used as an early-warning tool – pointing to areas of potential problems that can then be checked by teams on the ground.

A surprising amount of InSAR satellite data is made freely available through agencies such as NASA andEuropean Space Agency (ESA). One problem with the free data is that it tends to be lower in resolution than that which is available on a commercial basis.

Looking to the past to see the future

One of the benefits of InSAR and other satellite technologies is that there are available records for previous years. This means that if a structure such as a dam has started to show signs of deformation, it’s possible to check back in time (e.g., over the past two decades) to see what the situation was then. If there has been significant change, the contrast between then and now will give clues to the trends, and a glimpse of the future.

Artificial Intelligence gives new insights

With the vast volumes of data available from satellite imagery, it can be hard to sort through it to isolate information that really matters. As with many business processes these days, Artificial Intelligence (AI) technologies plus insights from experienced human analysts, can isolate actionable information from the flood of data.

How to get good results from InSAR analysis

For anyone with a property such as a dam, or a portfolio of properties, it may be best to start with a feasibility study. This means an investigation to see if InSAR analysis will be likely to provide the information needed (more on this below), or if other techniques may be better.

If the answer is positive, it’s good to start with a survey that uses public inSAR data, to see if this delivers what’s required. In the case the data does not meet the requirements, the analysts will recommend paying for commercial InSAR data.

Then, set up a regular monitoring program – once a month, perhaps, to survey all the properties in the portfolio to see if any have shown signs of deformation. At that point, other analysis such as ground-based surveying can provide the information needed to take action that protects the asset.

Know the limitations of InSAR

As with any technology, InSAR comes with limitations that may require mitigation or the exploration of alternative methods. For instance, if a structure such as a dam is heavily covered with vegetation, the satellite imagery may not produce a reliable result. Fortunately, there is a growing number of satellites that can “see” through vegetation to the ground beneath. Smooth surfaces may also cause problems for InSAR imagery – but by installing reflectors on the structure being monitored, it’s possible to check for deformation over time.

Experience has found that many owners of assets such as dams and mines are not aware of the benefits that InSAR can offer.

About the author:

Meisam Amani, Remote Sensing team Lead

Dr. Meisam Amani is a Remote Sensing Team Lead at WSP in Canada and the Key Specialty Leader of Remote Sensing and Geospatial Technologies at global WSP. His work focuses on remote sensing, machine learning, digital twin, and data analytics. He has over 15 years of experience in processing, analysing, and understanding earth observation datasets (e.g., satellite and drone imagery) for different applications.