Current methods used to forecast aftershocks — secondary quakes that can prove more deadly than initial earthquakes — can take several hours or days. New machine learning models have now been developed that can forecast where and how many aftershocks will take place following an earthquake in close to real-time.

Researchers from BGS, the University of Edinburgh and the University of Padua created the artificial intelligence (AI)-driven forecasting tools. They were developed by training machine learning models on earthquake data from California, New Zealand, Italy, Japan and Greece, all parts of the world that regularly experience earthquakes.

The rapid forecasts produced by AI-powered tools could help authorities with decision making about public safety measures and resource allocation in disaster-hit areas. The team analysed the AI models’ ability to produce forecasts of how many aftershocks will take place within the 24 hours following earthquakes of magnitude 4 or higher. They compared the performance of their models with the most widely used forecasting system, known as the epidemic-type aftershock sequence (ETAS) model, which is used operationally in Italy, New Zealand and the USA.

While both model types show similar performance at forecasting aftershock risk, the ETAS model took much longer to produce results. As it involves running a large number of simulations, the ETAS model can take up to several hours or days on a single mid-range computer.

By training the AI tools on records of past earthquakes from regions with different tectonic landscapes, researchers say their models could be used to forecast aftershock risk in most parts of the world that experience earthquakes.

The research, published in Earth, Planets and Space, was supported by the European Union Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie SPIN Innovative Training Network.

This study shows that machine learning models can produce aftershock forecasts within seconds, showing comparable quality to that of ETAS forecasts. Their speed and low computational cost offer major benefits for operational use: coupled with the near real-time development of machine learning-based, high-resolution earthquake catalogues, these models will enhance our ability to monitor and understand seismic crises as they evolve.

Foteini Dervisi, study leader, PhD student at BGS and the University of Edinburgh School of GeoSciences.

Landslides cause significant disruption to the road and rail network across Great Britain and can lead to fatalities. Identifying active slope failure is a difficult task, as monitoring is costly and time consuming, especially at a national scale.

In collaboration with the University of Florence in Italy, BGS has used a new, semi-automated method that uses artificial intelligence (AI) to identify the slopes that are actively moving, highlighting areas potentially at risk.

Previously, BGS has used interferometric synthetic aperture radar, or InSAR, for monitoring landslides. One of the benefits of InSAR is the large amount of information available, especially at a national scale; but analysing all these data present a challenge for scientists. To help tackle this problem, we have developed a semi-automated method that combines a type of AI called machine learning with clustering tools. The benefit of this approach is that we can analyse data for the whole of Great Britain, which wouldn’t have been possible before.

Results from this recent analysis highlighted around 3000 slopes that showed consistent movement of over 2.5mm per year between 2018 and 2022. These actively moving slopes affect approximately 14000km of road and 360km of railway — 2.4per cent and 1per cent of the entire national network, respectively.

The slopes deemed unstable are not all linked to landslides. Rather, they show the areas that should be focused on not only for future landslide research and mapping but also for the effect on local infrastructure, such as buildings and roads.

Our new, semi-automated approach supports the work of landslide specialists and provides a practical solution for large-scale geohazard management. The tool has helped to classify more than 300000 slopes around the UK and has highlighted 3000 slopes that have moved in a four-year period.

Satellite InSAR data has enormous potential for understanding ground deformation, but its complexity and the volume of data require advanced automated tools to extract meaningful information. Our semi-automated method helps bridge this gap by identifying the most critical areas to focus on, enabling efficient monitoring and helping to prevent serious damage.

Dr Alessandro Novellino, BGS remote sensing geologist

This approach already provides a powerful disaster-management tool, allowing decision makers to quickly identify areas that are currently at risk from ground motion. By highlighting these vulnerable areas, it supports smarter prioritisation of detailed field surveys, maintenance, and mitigation strategies, reducing costs and improving safety.

Next steps will focus on refining this national-scale analysis by integrating more detailed topographical data, to move from identifying unstable slopes to automatically mapping individual landslides within those slopes. This will enable more precise classification of landslide types and extents and the likely triggering mechanisms. The results will be shared with key stakeholders, including local authorities, infrastructure owners and the Natural Hazards Partnership.

Camilla Medici, postdoctoral researcher at the University of Florence

The research paper, , is now available to read.

The European Space Agency (ESA) has many offices around Europe but, as an Earth observation scientist myself, the Earth observation headquarters at ESA (ESRIN) office in Frascati, just outside Rome, is the pinnacle!

ESRIN coordinates and manages the ground-based activities of ESA’s Earth observation missions: data acquisition and processing, and satellitecommunication. It is the home of innovation and management of software used across the agency, and houses ESA records of legacy projects, with missions dating back to the 1970s. It also holds the largest archive of environmental data in Europe, coordinating over 20 ground stations and ground segment facilities across Europe.

Earth observation at ESRIN and BGS

ESA Earth-observing activities include satellite missions that monitor many of our planet natural processes, such as snow and ice cap accumulation and melt, wildfires, landslides, earthquakes and tectonic movements. It also tracks human-induced changes like city growth, deforestation and groundwater abstraction. Many of these processes and changes are also researched at BGS, using the data from these satellites alongside our expertise in geohazards and geological processes.

The typical challenge we face nowadays as Earth observation scientists is the sheer volume of data available to analyse — we have too much data to sift through manually. One avenue for allowing timely analyses of these large datasets is to use specific artificial intelligence (AI) models called foundation models.

What are foundation models?

Foundation models are designed to take in millions of pieces of data and find relationships between different datasets that we don’t have the time to do manually. Additionally, if the model is trained on several images of the Earth through time, it can make predictions about how our planet might change in the future. For BGS research, this could be used to help provide advice on a multitude of crucial future geological hazards faced by countries around the world; for example, how our coasts may change with sea-level rise.

Last month, I had the pleasure of attending ESA joint workshop with NASA on ‘Foundation models for Earth observation at ESA ESRIN’. We stayed near the Colli Albani volcanic complex, which has formed some of the beautiful hills and volcanic lakes surrounding this area, just south-east of Rome.

At the workshop

My job at BGS is to use both classical and newly devised methods to analyse satellite data and find patterns between the behaviour of the ground beneath us and our other geospatial datasets, and what this means for the people and surface infrastructure. This workshop was ideal for my role. I attended the sessions that focused on applications of AI models to real scenarios; on day one, sessions included using foundation models for various applications in earth sciences, weather prediction and climate science.

On day two, I attended a morning session on how scientists are adapting foundation models for geospatial and Earth observation tasks, which is exactly what I’m aiming to do! In the late morning and afternoon, I had my poster presentation slot, where I showed how my team at BGS envisions using AI alongside Earth observation and BGS data. This includes our bedrock and superficial deposit maps created by our survey geologists, hazard susceptibility maps from our hazards specialists, and more. I also presented some of the machine learning (ML) tools BGS has developed so far to help with this task. I got talking to some really engaging researchers, learning a lot from the people I spoke to about what data works well in these models (and what doesn’t!) and the field of AI in earth sciences as a whole. This was the most beneficial day to me as an early career scientist; talking to so many people from different organisations with different areas of expertise has been invaluable in my development as a scientist.

When in Rome…

A trip to Lazio wouldn’t be complete without a little sightseeing, so on my penultimate night in Rome I managed to squeeze in some touristy activities. A great thing about working for BGS is being able to experience different cultures and their food — and to have your Italian speaking skills completely humbled by the locals…!

The final day consisted of hands-on training workshops in three of the foundation models that ESA and NASA have developed over the years, which is what I was most looking forward to. The training was delivered by the scientists at NASA Impact and IBM, who helped write the models, who were all fantastically knowledgeable.

Putting my knowledge to work

Now, a few weeks after coming back and with my newfound knowledge from world-leading experts in artificial intelligence, I’ve started to piece together more about how BGS could incorporate our data into such powerful models and I’m excited to practise my new skills. Unfortunately though, I couldn’t bring back buckets of Roman carbonara… so I’ll just have to get back to Rome as soon as I can!

About the author

To date, the real-time impact data that is needed to effectively forecast and monitor geological hazard events has been unavailable or incomplete. The FloodTags platform aims to fill this gap by using large language models (LLMs) to extract real-time and historic information from social media platforms (X; YouTube; Bluesky; Facebook; Instagram) and more than 150 000 online news sources. This collaboration is a step towards providing timely, ground-level insight into geological hazards around the world.

I am aware that many organisations around the world, including BGS, rely on the manual gathering of data from social media and the news during disaster events, and to update regional and national hazard inventories. This can add a significant time lag to relevant information being interpreted, particularly during natural disasters, which means any actions taken are also delayed. We have been working with FloodTags for some time now and are delighted to formalise our collaboration in this highly valuable area of research.

Catherine Pennington, BGS Engineering Geologist, landslides.

This collaboration marks a major step forward for FloodTags. Partnering with BGS brings us the scientific expertise and data to expand into landslides and other geological hazards. Their deep knowledge of earth science opens the door to new applications for our real-time media monitoring tools. Combined with the power of large language models, this collaboration allows us to jointly deliver fast and relevant disaster insights for both hydrological and geological hazards. This helps governments and emergency services in making more informed, evidence-based decisions.

Jurjen Wagemaker, founder of FloodTags.

As a first activity under the new Memorandum of Understanding, BGS and FloodTags are in Indonesia this week topresent the first version of HazTags, an LLM-powered platform for monitoring floodsand landslides using social and news media data. They will discuss long-term collaboration in Indonesia with:

• theIndonesian national research agency, BRIN
• Centre for Volcanology and Geological Hazard Mitigation (PVMBG)
• Indonesian Red Cross (PMI)
• Meteorology, Climatology and GeophysicsAgency (BMKG)
• Ministry for Public Works (PU)
• National Agency for Disaster Management (BPBD)
• Research Centre for Disaster Mitigation (ITB)

For more information, please contact BGS press (bgspress@bgs.ac.uk) or call 07790 607 010.

Twenty years ago, BGS took a bold step into the world of virtual reality (VR), pioneering 3D visualisation in geological surveying. From its first immersive 3D room in 2005 to its global influence today, the journey of VR at BGS has been one of constant innovation, exploration and impact.

The early days: a new perspective on geology

51ÁÔÆæ has been working to improve understanding of the potential for CO₂ storage as part of its International Geoscience Research and Development programme. During early 2005, Virtalis Ltd. installed BGS first 3D visualisation facility at our headquarters in Keyworth, Nottinghamshire, featuring a state-of-the-art Christie S-4K projector and nVidia Quadro graphics. A cutting-edge Intersense motion tracking system brought geological models to life, offering an immersive way to explore complex 3D data.

Initially, these rooms served as a platform to showcase geological model outputs from the BGS  GeoScience Spatial Model (DGSM) programme. However, their potential quickly expanded beyond presentations, opening the door to practical applications in 3D modelling and landscape visualisation. BGS commissioned Virtalis to develop immersive VR experiences for and digital terrain models, overlaying geological maps, aerial photographs and satellite images to enrich understanding of geological formations.

A showcase for science

The technology quickly became a key feature of BGS, attracting a diverse range of visitors. These included school groups, MPs, top government scientific advisors, VIPs and even royalty — including the Princess Royal — who were all invited to witness BGS’s cutting-edge capabilities firsthand. These demonstrations proved to be a powerful tool for communicating the importance of geological research.

Expanding the horizon: virtual field reconnaissance

By 2006, the vision for BGS VR had grown. The next challenge was making VR an interactive, integral part of geological surveying. A cross-disciplinary project was launched, bringing together the land survey, remote sensing and data and digital systems teams to develop virtual field reconnaissance (VFR). The aim was ambitious: integrate field data collection with VR landscapes, enabling geologists to conduct initial assessments remotely before heading into the field. This initiative aligned perfectly with BGS acquisition of high-resolution aerial orthophotos and a 5m digital terrain model (DTM) and digital surface model (DSM) from Intermap Technologies.

To handle the vast datasets seamlessly, Virtalis was commissioned once again, this time to build a prototype virtual landscape visualisation system. The result was a game-changing enhancement to geological fieldwork, increasing efficiency and accuracy and giving us the ability to plan research in a way never before possible.

Evolution and global expansion

In 2010, the original immersive 3D visualisation facility (i3DVF) needed a new home due to building works. BGS created a cutting-edge VR hub, complete with an upgraded projector, screen and computing power. Six years later, in 2016, the facility saw another major upgrade: the world first 4K projection system in a geological survey organisation. The enhanced resolution, combined with new, highly detailed 2 m DTM and DSM data from the Pan-Governmental Data Agreement, took geological visualisation to an unprecedented level.

What began as a pioneering project within BGS has since spread across the world. The integration of GeoVisionary and i3DVF technology has inspired geological surveys, mining companies, universities, and environmental organisations globally. From Alaska to South Africa, Malaysia to Brazil, BGS VR expertise continues to revolutionise how geologists explore and understand our planet.

Looking ahead

As we celebrate 20 years of VR at BGS, we also look to the future. With advancements in artificial intelligence, real-time data processing and even more immersive visualisation technologies, the possibilities for geological VR are boundless. One thing is certain: BGS will remain at the forefront, pushing the boundaries of innovation and transforming how we see the Earth.

Here to the next 20 years of discovery!

Further reading

• 51ÁÔÆæ 3D Visualisation Systems

As a state of emergency is declared on the Greek island of Santorini, seismologists are increasingly turning to artificial intelligence technology to provide high-resolution images of the ongoing seismic activity, in a bid to enhance short-term forecasting accuracy.

Since the start of the crisis, a team from BGS comprising Margarita Segou, Brian Baptie, Rajat Choudhary, Wayne Shelley and Foteini Dervisi, has been employing machine learning algorithms to detect ten times as many earthquakes as standard techniques, with over 20000 tremors accurately predicted in the Santorini area alone since 1 December 2024. This approach is allowing geologists to identify for the first time small magnitude earthquakes that were previously undetected using standard approaches.

51ÁÔÆæ Seismologist Margarita Segou, who is leading the development of the groundbreaking research, says it has revolutionised the way scientists can learn from seismic activity and predict patterns.

This machine learning technique results in far richer data feeding into short-term forecasts, which can allow experts to track the evolution of events and better advise emergency services and at-risk communities.

Dr Margarita Segou, BGS Seismologist.

These algorithms allowed researchers to first note increased seismic activity across the Santorini region on 26 January 2025. In comparison, standard detection schemes did not register the same increase until 31 January and only picked up around 2000 seismic events in the Santorini area; ten times less than the new approach has detected.

Dr Segou says it is the ability to combine different sources of information more quickly that is at the heart of the advancement.

Through strong international partnerships, we can reprocess past and present data through machine learning and gain a new and priceless insight into the seismic activity in Santorini in previous phases of unrest and its links to the volcanic system.

Dr Margarita Segou.

Santorini is located on the Hellenic volcanic arc at the convergence of the African plate and the Eurasian plate, at a complex tectonic boundary. Currently, seismic events around the island show that seismicity bursts occur almost twice a day, with the tremors lasting for one to two hours.

Dr Segou adds the data is revealing some unique features.

We have evidence that this is fluid-driven, swarm-type seismicity that comes in pulses. This is not unheard of in other volcanic regions; however, this time it is evolving on top of active faults that complicate the expression of seismicity.

It is easy to get a disconnected story when we just look at moderate magnitude seismic events. It is only when we investigate the smaller magnitude events that occur between that we learn of the hidden mechanisms that take place between the large earthquakes.

It is critical that we track whether those pulses become more frequent and how they migrate in space and depth. So far, the largest quake in this swarm has been a 5.2 magnitude.

Dr Margarita Segou.

Contact

For more information, please contact 51ÁÔÆæ press (bgspress@bgs.ac.uk) or call 07790 607 010.

Extreme weather events are projected to become more common in the UK, costing £750 million per year (Bates et al., 2023). A new, £38 million infrastructure project will enhance the UK resilience to floods and droughts and will include open-air laboratories across the UK and a large-scale, live environmental data bank.

The project, titled ‘’ (FDRI), will provide infrastructure to allow aspects of the hydrological cycle in specific locations in England, Scotland and Wales to be tracked. The data produced can be used alongside artificial intelligence (AI) and machine-learning technology to model present conditions and forecast the impact of extremes.

Improving our ability to analyse UK environmental data with models and AI will:

• improve the prediction of flood and drought risk
• enable the creation of better, more cost-effective infrastructure
• allow more accurate response to water supply demands

Monitoring activities will be coordinated and innovation better directed through the network that the FDRI project will create. It will also create a near real-time data bank with outdoor laboratories in three catchments: the Severn, the Chess (Thames) and the Tweed. This will be achieved by deploying instruments for observing subtle changes in the water environment, such as:

• evaporation
• soil moisture
• weather
• groundwater
• river flow

It will also provide new digital solutions to support data and help build capacity in the hydrological community through training and skills sharing.

We are delighted to be part of this landmark project, which will provide the UK with revolutionary solutions to reduce the impact of floods and droughts.

Each year, dealing with the impact of flooding and droughts costs the UK around £750 million. It is through increased resilience and advanced prediction capabilities that the nation can reduce this cost and better protect at-risk communities.

Alan MacDonald, head of BGS Groundwater.

Funding

The £38 million project has been awarded funding by the 51ÁÔÆæ/Natural Environment Research Council (NERC). NERC and the UK Centre for Hydrology & Ecology will lead the project, with contributions from BGS, Imperial College London and the University of Bristol.

Earth changing climate is increasing the number of extreme floods and droughts, causing environmental, societal and economic damage. This investment will transform the way we can forecast these events by building data and monitoring capability.

NERC is helping to respond to climate challenges with research and innovation investments that will accelerate the green economy and deliver solutions to national priorities.

Prof Louise Heathwaite, executive chair of 51ÁÔÆæ/NERC.

The project will work closely with organisations in the environmental and government sectors, including the Environment Agency, to build modelling and help prepare for severe weather.

Reference

Bates, P D, Savage, J, Wing, O, Quinn, N, Sampson, C, Neal, J, and Smith, A. 2023. . Natural Hazards and Earth System Sciences, Vol. 23, 891–908. DOI: https://doi.org/10.5194/nhess-23-891-2023

Notes for editors

Landslides and floods triggered by earthquakes pose a great threat to human life and infrastructure. Currently, research into mitigation of these natural hazards has focused on events triggered during or shortly after earthquakes; for example, the failure of a slope shortly after a seismic event.

However, the long-term seismic effects that cause unstable landslides to accelerate without immediate failure are largely neglected. A new project, ‘Advancing knowledge of multi-hazards processes and their impact’ (AMHEI), aims to fill this research gap by looking at slope dynamics following a major earthquake and how these processes can also affect flood hazards.

There is currently little research into earthquakes affecting landslides and flooding in the longer term. We are excited to be able to utilise latest satellite technologies to better understand the relationships between these hazards.

Alessandro Novellino, BGS Remote Sensing Geoscientist.

AMHEI will use the latest satellite technologies, includingInterferometric Synthetic Aperture Radar(InSAR) combined with artificial intelligence techniques, to map and identify relationships between natural hazards in Turkey, using the February 2023 earthquake as a case study.

Funded by the European Space Agency, the project will be led by Alessandro Novellino, a remote sensing geoscientist at BGS, in collaboration with colleagues at the Faculty of Geo-Information Science and Earth Observation at the University of Twente (Netherlands) and the University of Bergen (Norway).