case study on an earthquake

Learning from Megadisasters: A Decade of Lessons from the Great East Japan Earthquake

March 11, 2021 Tokyo, Japan

Authors: Shoko Takemoto,  Naho Shibuya, and Keiko Sakoda

Today marks the ten-year anniversary of the Great East Japan Earthquake (GEJE), a mega-disaster that marked Japan and the world with its unprecedented scale of destruction. This feature story commemorates the disaster by reflecting on what it has taught us over the past decade in regards to infrastructure resilience, risk identification, reduction, and preparedness, and disaster risk finance.  Since GEJE, the World Bank in partnership with the Government of Japan, especially through the Japan-World Bank Program on Mainstreaming Disaster Risk Management in Developing Countries has been working with Japanese and global partners to understand impact, response, and recovery from this megadisaster to identify larger lessons for disaster risk management (DRM).

Among the numerous lessons learned over the past decade of GEJE reconstruction and analysis, we highlight three common themes that have emerged repeatedly through the examples of good practices gathered across various sectors.  First is the importance of planning. Even though disasters will always be unexpected, if not unprecedented, planning for disasters has benefits both before and after they occur. Second is that resilience is strengthened when it is shared .  After a decade since GEJE, to strengthen the resilience of infrastructure, preparedness, and finance for the next disaster, throughout Japan national and local governments, infrastructure developers and operators, businesses and industries, communities and households are building back better systems by prearranging mechanisms for risk reduction, response and continuity through collaboration and mutual support.  Third is that resilience is an iterative process .  Many adaptations were made to the policy and regulatory frameworks after the GEJE. Many past disasters show that resilience is an interactive process that needs to be adjusted and sustained over time, especially before a disaster strikes.

As the world is increasingly tested to respond and rebuild from unexpected impacts of extreme weather events and the COVID-19 pandemic, we highlight some of these efforts that may have relevance for countries around the world seeking to improve their preparedness for disaster events.

Introduction: The Triple Disaster, Response and Recovery

On March 11th, 2011 a Magnitude 9.0 earthquake struck off the northeast coast of Japan, near the Tohoku region. The force of the earthquake sent a tsunami rushing towards the Tohoku coastline, a black wall of water which wiped away entire towns and villages. Sea walls were overrun. 20,000 lives were lost. The scale of destruction to housing, infrastructure, industry and agriculture was extreme in Fukushima, Iwate, and Miyagi prefectures. In addition to the hundreds of thousands who lost their homes, the earthquake and tsunami contributed to an accident at the Fukushima Daiichi Nuclear Power Plant, requiring additional mass evacuations. The impacts not only shook Japan’s society and economy as a whole, but also had ripple effects in global supply chains. In the 21st century, a disaster of this scale is a global phenomenon.

The severity and complexity of the cascading disasters was not anticipated. The events during and following the Great East Japan Earthquake (GEJE) showed just how ruinous and complex a low-probability, high-impact disaster can be. However, although the impacts of the triple-disaster were devastating, Japan’s legacy of DRM likely reduced losses. Japan’s structural investments in warning systems and infrastructure were effective in many cases, and preparedness training helped many act and evacuate quickly. The large spatial impact of the disaster, and the region’s largely rural and elderly population, posed additional challenges for response and recovery.

Ten years after the megadisaster, the region is beginning to return to a sense of normalcy, even if many places look quite different. After years in rapidly-implemented temporary prefabricated housing, most people have moved into permanent homes, including 30,000 new units of public housing . Damaged infrastructure has been also restored or is nearing completion in the region, including rail lines, roads, and seawalls.

In 2014, three years after GEJE, The World Bank published Learning from Megadisasters: Lessons from the Great East Japan Earthquake . Edited by Federica Ranghieri and Mikio Ishiwatari , the volume brought together dozens of experts ranging from seismic engineers to urban planners, who analyzed what happened on March 11, 2011 and the following days, months, and years; compiling lessons for other countries in 36 comprehensive Knowledge Notes . This extensive research effort identified a number of key learnings in multiple sectors, and emphasized the importance of both structural and non-structural measures, as well as identifying effective strategies both pre- and post-disaster. The report highlighted four central lessons after this intensive study of the GEJE disaster, response, and initial recovery:

1) A holistic, rather than single-sector approach to DRM improves preparedness for complex disasters; 2) Investing in prevention is important, but is not a substitute for preparedness; 3) Each disaster is an opportunity to learn and adapt; 4) Effective DRM requires bringing together diverse stakeholders, including various levels of government, community and nonprofit actors, and the private sector.

Although these lessons are learned specifically from the GEJE, the report also focuses on learnings with broader applicability.

Over recent years, the Japan-World Bank Program on Mainstreaming DRM in Developing Countries has furthered the work of the Learning from Megadisasters report, continuing to gather, analyze and share the knowledge and lessons learned from GEJE, together with past disaster experiences, to enhance the resilience of next generation development investments around the world. Ten years on from the GEJE, we take a moment to revisit the lessons gathered, and reflect on how they may continue to be relevant in the next decade, in a world faced with both seismic disasters and other emergent hazards such as pandemics and climate change.

Through synthesizing a decade of research on the GEJE and accumulation of the lessons from the past disaster experience, this story highlights three key strategies which recurred across many of the cases we studied. They are:

1) the importance of planning for disasters before they strike, 2) DRM cannot be addressed by either the public or private sector alone but enabled only when it is shared among many stakeholders , 3) institutionalize the culture of continuous enhancement of the resilience .

For example, business continuity plans, or BCPs, can help both public and private organizations minimize damages and disruptions . BCPs are documents prepared in advance which provide guidance on how to respond to a disruption and resume the delivery of products and services. Additionally, the creation of pre-arranged agreements among independent public and/or private organizations can help share essential responsibilities and information both before and after a disaster . This might include agreements with private firms to repair public infrastructures, among private firms to share the costs of mitigation infrastructure, or among municipalities to share rapid response teams and other resources. These three approaches recur throughout the more specific lessons and strategies identified in the following section, which is organized along the three areas of disaster risk management: resilient infrastructure; risk identification, reduction and preparednes s ; and disaster risk finance and insurance.

Lessons from the Megadisaster

Resilient Infrastructure

The GEJE had severe impacts on critical ‘lifelines’—infrastructures and facilities that provide essential services such as transportation, communication, sanitation, education, and medical care. Impacts of megadisasters include not only damages to assets (direct impacts), but also disruptions of key services, and the resulting social and economic effects (indirect impacts). For example, the GEJE caused a water supply disruption for up to 500,000 people in Sendai city, as well as completely submerging the city’s water treatment plant. [i] Lack of access to water and sanitation had a ripple effect on public health and other emergency services, impacting response and recovery. Smart investment in infrastructure resilience can help minimize both direct and indirect impacts, reducing lifeline disruptions. The 2019 report Lifelines: The Resilient Infrastructure Opportunity found through a global study that every dollar invested in the resilience of lifelines had a $4 benefit in the long run.

In the case of water infrastructure , the World Bank report Resilient Water Supply and Sanitation Services: The Case of Japan documents how Sendai City learned from the disaster to improve the resilience of these infrastructures. [ii] Steps included retrofitting existing systems with seismic resilience upgrades, enhancing business continuity planning for sanitation systems, and creating a geographic information system (GIS)-based asset management system that allows for quick identification and repair of damaged pipes and other assets. During the GEJE, damages and disruptions to water delivery services were minimized through existing programs, including mutual aid agreements with other water supply utility operators. Through these agreements, the Sendai City Waterworks Bureau received support from more than 60 water utilities to provide emergency water supplies. Policies which promote structural resilience strategies were also essential to preserving water and sanitation services. After the 1995 Great Hanshin Awaji Earthquake (GHAE), Japanese utilities invested in earthquake resistant piping in water supply and sanitation systems. The commonly used earthquake-resistant ductile iron pipe (ERDIP) has not shown any damage from major earthquakes including the 2011 GEJE and the 2016 Kumamoto earthquake. [iii] Changes were also made to internal policies after the GEJE based on the challenges faced, such as decentralizing emergency decision-making and providing training for local communities to set up emergency water supplies without utility workers with the goal of speeding up recovery efforts. [iv]

Redundancy is another structural strategy that contributed to resilience during and after GEJE. In Sendai City, redundancy and seismic reinforcement in water supply infrastructure allowed the utility to continue to operate pipelines that were not physically damaged in the earthquake. [v] The Lifelines report describes how in the context of telecommunications infrastructure , the redundancy created through a diversity of routes in Japan’s submarine internet cable system  limited disruptions to national connectivity during the megadisaster. [vi] However, the report emphasizes that redundancy must be calibrated to the needs and resources of a particular context. For private firms, redundancy and backups for critical infrastructure can be achieved through collaboration; after the GEJE, firms are increasingly collaborating to defray the costs of these investments. [vii]

The GEJE also illustrated the importance of planning for transportation resilience . A Japan Case Study Report on Road Geohazard Risk Management shows the role that both national policy and public-private agreements can play. In response to the GEJE, Japan’s central disaster legislation, the DCBA (Disaster Countermeasures Basic Act) was amended in 2012, with particular focus on the need to reopen roads for emergency response. Quick road repairs were made possible after the GEJE in part due to the Ministry of Land, Infrastructure, Transport and Tourism (MLIT)’s emergency action plans, the swift action of the rapid response agency Technical Emergency Control Force (TEC-FORCE), and prearranged agreements with private construction companies for emergency recovery work. [viii] During the GEJE, roads were used as evacuation sites and were shown effective in controlling the spread of floods. After the disaster, public-private partnerships (PPPs) were also made to accommodate the use of expressway embankments as tsunami evacuation sites. As research on Resilient Infrastructure PPPs highlights, clear definitions of roles and responsibilities are essential to effective arrangements between the government and private companies. In Japan, lessons from the GEJE and other earthquakes have led to a refinement of disaster definitions, such as numerical standards for triggering force majeure provisions of infrastructure PPP contracts. In Sendai City, clarifying the post-disaster responsibilities of public and private actors across various sectors sped up the response process. [ix] This experience was built upon after the disaster, when Miyagi prefecture conferred operation of the Sendai International Airport   to a private consortium through a concession scheme which included refined force majeure definitions. In the context of a hazard-prone region, the agreement clearly defines disaster-related roles and responsibilities as well as relevant triggering events. [x]

Partnerships for creating backup systems that have value in non-disaster times have also proved effective in the aftermath of the GEJE. As described in Resilient Industries in Japan , Toyota’s automotive plant in Ohira village, Miyagi Prefecture lost power for two weeks following GEJE. To avoid such losses in the future, companies in the industrial park sought to secure energy during power outages and shortages by building the F-Grid, their own mini-grid system with a comprehensive energy management system. The F-Grid project is a collaboration of 10 companies and organizations in the Ohira Industrial Park. As a system used exclusively for backup energy would be costly, the system is also used to improve energy efficiency in the park during normal times. The project was supported by funding from Japan’s “Smart Communities'' program. [xi] In 2016, F-grid achieved a 24 percent increase in energy efficiency and a 31 percent reduction in carbon dioxide emissions compared to similarly sized parks. [xii]

Schools are also critical infrastructures, for their education and community roles, and also because they are commonly used as evacuation centers. Japan has updated seismic resilience standards for schools over time, integrating measures against different risks and vulnerabilities revealed after each disaster, as documented in the report Making Schools Resilient at Scale . After the 2011 GEJE, there was very little earthquake-related damage; rather, most damage was caused by the tsunami. However, in some cases damages to nonstructural elements like suspending ceilings in school gymnasiums limited the possibility of using these spaces after the disaster. After the disaster, a major update was made to the policies on the safety of nonstructural elements in schools, given the need for higher resilience standards for their function as post-disaster evacuation centers [xiii] .

Similarly, for building regulations , standards and professional training modules were updated taking the lessons learned from GEJE. The Converting Disaster Experience into a Safer Built Environment: The Case of Japan report highlights that, legal framework like, The Building Standard Law/Seismic Retrofitting Promotion Law, was amended further enhance the structural resilience of the built environment, including strengthening structural integrity, improving the efficiency of design review process, as well as mandating seismic diagnosis of large public buildings. Since the establishment of the legal and regulatory framework for building safety in early 1900, Japan continued incremental effort to create enabling environment for owners, designers, builders and building officials to make the built environment safer together.

Cultural heritage also plays an important role in creating healthy communities, and the loss or damage of these items can scar the cohesion and identity of a community. The report Resilient Cultural Heritage: Learning from the Japanese Experience shows how the GEJE highlighted the importance of investing in the resilience of cultural properties, such as through restoration budgets and response teams, which enabled the relocation of at-risk items and restoration of properties during and after the GEJE. After the megadisaster, the volunteer organization Shiryō-Net was formed to help rescue and preserve heritage properties, and this network has now spread across Japan. [xiv] Engaging both volunteer and government organizations in heritage preservation can allow for a more wide-ranging response. Cultural properties can play a role in healing communities wrought by disasters: in Ishinomaki City, the restoration of a historic storehouse served as a symbol of reconstruction [xv] , while elsewhere repair of cultural heritage sites and the celebration of cultural festivals served a stimulant for recovery. [xvi] Cultural heritage also played a preventative role during and after the disaster by embedding the experience of prior disasters in the built environment. Stone monuments which marked the extent of historic tsunamis served as guides for some residents, who fled uphill past the stones and escaped the dangerous waters. [xvii] This suggests a potential role for cultural heritage in instructing future generations about historic hazards.

These examples of lessons from the GEJE highlight how investing in resilient infrastructure is essential, but must also be done smartly, with emphasis on planning, design, and maintenance. Focusing on both minimizing disaster impacts and putting processes in place to facilitate speedy infrastructure restoration can reduce both direct and indirect impacts of megadisasters.  Over the decade since GEJE, many examples and experiences on how to better invest in resilient infrastructure, plan for service continuity and quick response, and catalyze strategic partnerships across diverse groups are emerging from Japan.

Risk Identification, Reduction, and Preparedness

Ten years after the GEJE, a number of lessons have emerged as important in identifying, reducing, and preparing for disaster risks. Given the unprecedented nature of the GEJE, it is important to be prepared for both known and uncertain risks. Information and communication technology (ICT) can play a role in improving risk identification and making evidence-based decisions for disaster risk reduction and preparedness. Communicating these risks to communities, in a way people can take appropriate mitigation action, is a key . These processes also need to be inclusive , involving diverse stakeholders--including women, elders , and the private sector--that need to be engaged and empowered to understand, reduce, and prepare for disasters. Finally, resilience is never complete . Rather, as the adaptations made by Japan after the GEJE and many past disasters show, resilience is a continuous process that needs to be adjusted and sustained over time, especially in times before a disaster strikes.

Although DRM is central in Japan, the scale of the 2011 triple disaster dramatically exceeded expectations. After the GEJE, as Chapter 32 of Learning From Megadisasters highlights, the potential of low-probability, high-impact events led Japan to focus on both structural and nonstructural disaster risk management measures. [xviii] Mitigation and preparedness strategies can be designed to be effective for both predicted and uncertain risks. Planning for a multihazard context, rather than only individual hazards, can help countries act quickly even when the unimaginable occurs. Identifying, preparing for, and reducing disaster risks all play a role in this process.

The GEJE highlighted the important role ICT can play in both understanding risk and making evidence-based decisions for risk identification, reduction, and preparedness. As documented in the World Bank report Information and Communication Technology for Disaster Risk Management in Japan , at the time of the GEJE, Japan had implemented various ICT systems for disaster response and recovery, and the disaster tested the effectiveness of these systems. During the GEJE, Japan’s “Earthquake Early Warning System” (EEWS) issued a series of warnings. Through the detection of initial seismic waves, EEWS can provide a warning of a few seconds or minutes, allowing quick action by individuals and organizations. Japan Railways’ “Urgent Earthquake Detection and Alarm System” (UrEDAS) automatically activated emergency brakes of 27 Shinkansen train lines , successfully bringing all trains to a safe stop. After the disaster, Japan expanded emergency alert delivery systems. [xix]

The World Bank’s study on Preparedness Maps shows how seismic preparedness maps are used in Japan to communicate location specific primary and secondary hazards from earthquakes, promoting preparedness at the community and household level. Preparedness maps are regularly updated after disaster events, and since 2011 Japan has promoted risk reduction activities to prepare for the projected maximum likely tsunami [xx] .

Effective engagement of various stakeholders is also important to preparedness mapping and other disaster preparedness activities. This means engaging and empowering diverse groups including women, the elderly, children, and the private sector. Elders are a particularly important demographic in the context of the GEJE, as the report Elders Leading the Way to Resilience illustrates. Tohoku is an aging region, and two-thirds of lives lost from the GEJE were over 60 years old. Research shows that building trust and social ties can reduce disaster impacts- after GEJE, a study found that communities with high social capital lost fewer residents to the tsunami. [xxi] Following the megadisaster, elders in Ofunato formed the Ibasho Cafe, a community space for strengthening social capital among older people. The World Bank has explored the potential of the Ibasho model for other contexts , highlighting how fueling social capital and engaging elders in strengthening their community can have benefits for both normal times and improve resilience when a disaster does strike.

Conducting simulation drills regularly provide another way of engaging stakeholders in preparedness. As described in Learning from Disaster Simulation Drills in Japan , [xxii] after the 1995 GHAE the first Comprehensive Disaster Management Drill Framework was developed as a guide for the execution of a comprehensive system of disaster response drills and establishing links between various disaster management agencies. The Comprehensive Disaster Management Drill Framework is updated annually by the Central Disaster Management Council. The GEJE led to new and improved drill protocols in the impacted region and in Japan as a whole. For example, the 35th Joint Disaster simulation Drill was held in the Tokyo metropolitan region in 2015 to respond to issues identified during the GEJE, such as improving mutual support systems among residents, governments, and organizations; verifying disaster management plans; and improving disaster response capabilities of government agencies. In addition to regularly scheduled disaster simulation drills, GEJE memorial events are held in Japan annually to memorialize victims and keep disaster preparedness in the public consciousness.

Business continuity planning (BCP) is another key strategy that shows how ongoing attention to resilience is also essential for both public and private sector organizations. As Resilient Industries in Japan demonstrates, after the GEJE, BCPs helped firms reduce disaster losses and recover quickly, benefiting employees, supply chains, and the economy at large. BCP is supported by many national policies in Japan, and after the GEJE, firms that had BCPs in place had reduced impacts on their financial soundness compared to firms that did not. [xxiii] The GEJE also led to the update and refinement of BCPs across Japan. Akemi industrial park in Aichi prefecture, began business continuity planning at the scale of the industrial park three years before the GEJE. After the GEJE, the park revised their plan, expanding focus on the safety of workers. National policies in Japan promote the development of BCPs, including the 2013 Basic Act for National Resilience, which was developed after the GEJE and emphasizes resilience as a shared goal across multiple sectors. [xxiv] Japan also supports BCP development for public sector organizations including subnational governments and infrastructure operators. By 2019, all of Japan’s prefectural governments, and nearly 90% of municipal governments had developed BCPs. [xxv] The role of financial institutions in incentivizing BCPs is further addressed in the following section.

The ongoing nature of these preparedness actions highlights that resilience is a continuous process. Risk management strategies must be adapted and sustained over time, especially during times without disasters. This principle is central to Japan’s disaster resilience policies. In late 2011, based on a report documenting the GEJE from the Expert Committee on Earthquake and Tsunami Disaster Management, Japan amended the DCBA (Disaster Countermeasures Basic Act) to enhance its multi-hazard countermeasures, adding a chapter on tsunami countermeasures. [xxvi]

Disaster Risk Finance and Insurance

Disasters can have a large financial impact, not only in the areas where they strike, but also at the large scale of supply chains and national economy. For example, the GEJE led to the shutdown of nuclear power plants across Japan, resulting in a 50% decrease in energy production and causing national supply disruptions. The GEJE has illustrated the importance of disaster risk finance and insurance (DRFI) such as understanding and clarifying contingent liabilities and allocating contingency budgets, putting in place financial protection measures for critical lifeline infrastructure assets and services, and developing mechanisms for vulnerable businesses and households to quickly access financial support. DRFI mechanisms can help people, firms, and critical infrastructure avoid or minimize disruptions, continue operations, and recover quickly after a disaster.

Pre-arranged agreements, including public-private partnerships, are key strategies for the financial protection of critical infrastructure. The report Financial Protection of Critical Infrastructure Services (forthcoming) [xxvii] shows how pre-arranged agreements between the public sector and private sector for post-disaster response can facilitate rapid infrastructure recovery after disasters, reducing the direct and indirect impacts of infrastructure disruptions, including economic impacts. GEJE caused devastating impacts to the transportation network across Japan. Approximately 2,300 km of expressways were closed, representing 65 percent of expressways managed by NEXCO East Japan , resulting in major supply chain disruptions [xxviii] .  However, with the activation of pre-arranged agreements between governments and local construction companies for road clearance and recovery work, allowing damaged major motorways to be repaired within one week of the earthquake. This quick response allowed critical access for other emergency services to further relief and recovery operations.

The GEJE illustrated the importance of clearly defining post-disaster financial roles and responsibilities among public and private actors in order to restore critical infrastructure rapidly . World Bank research on Catastrophe Insurance Programs for Public Assets highlights how the Japan Railway Construction, Transport and Technology Agency  (JRTT) uses insurance to reduce the contingent liabilities of critical infrastructure to ease impacts to government budgets in the event of a megadisaster. Advance agreements between the government, infrastructure owners and operators, and insurance companies clearly outline how financial responsibilities will be shared in the event of a disaster. In the event of a megadisaster like GEJE, the government pays a large share of recovery costs, which enables the Shinkansen bullet train service to be restored more rapidly. [xxix]

The Resilient Industries in Japan   report highlights how diverse and comprehensive disaster risk financing methods are also important to promoting a resilient industry sector . After the GEJE, 90% of bankruptcies linked to the disaster were due to indirect impacts such as supply chain disruptions. This means that industries located elsewhere are also vulnerable: a study found that six years after GEJE, a greater proportion of bankruptcy declarations were located in Tokyo than Tohoku. [xxx] Further, firms without disaster risk financing in place had much higher increases in debt levels than firms with preexisting risk financing mechanisms in place. [xxxi] Disaster risk financing can play a role pre-disaster, through mechanisms such as low-interest loans, guarantees, insurance, or grants which incentivize the creation of BCPs and other mitigation and preparedness measures.  When a disaster strikes, financial mechanisms that support impacted businesses, especially small or medium enterprises and women-owned businesses, can help promote equitable recovery and help businesses survive. For financial institutions, simply keeping banks open after a major disaster can support response and recovery. After the GEJE, the Bank of Japan (BoJ) and local banks leveraged pre-arranged agreements to maintain liquidity, opening the first weekend after the disaster to help minimize economic disruptions. [xxxii] These strategies highlight the important role of finance in considering economic needs before a disaster strikes, and having systems in place to act quickly to limit both economic and infrastructure service impacts of disasters.

Looking to the Future

Ten years after the GEJE, these lessons in the realms of resilient infrastructure, risk identification, reduction and preparedness, and DRFI are significant not only for parts of the world preparing for tsunamis and other seismic hazards, but also for many of the other types of hazards faced around the globe in 2021. In Japan, many of the lessons of the GEJE are being applied to the projected Nankai Trough and Tokyo Inland earthquakes, for example through modelling risks and mapping evacuation routes, implementing scenario planning exercises and evacuation drills , or even prearranging a post-disaster reconstruction vision and plans. These resilience measures are taken not only individually but also through innovative partnerships for collaboration across regions, sectors, and organizations including public-private agreements to share resources and expertise in the event of a major disaster.

The ten-year anniversary of the GEJE finds the world in the midst of the multiple emergencies of the global COVID-19 pandemic, environmental and technological hazards, and climate change. Beyond seismic hazards, the global pandemic has highlighted, for example, the risks of supply chain disruption due to biological emergencies. Climate change is also increasing hazard exposure in Japan and around the globe. Climate change is a growing concern for its potential to contribute to hydrometeorological hazards such as flooding and hurricanes, and for its potential to play a role in secondary or cascading hazards such as fire. In the era of climate change, disasters will increasingly be ‘unprecedented’, and so GEJE offers important lessons on preparing for low-probability high-impact disasters and planning under uncertain conditions in general.

Over the last decade, the World Bank has drawn upon the GEJE megadisaster experience to learn how to better prepare for and recover from low-probability high-impact disasters. While we have identified a number of diverse strategies here, ranging from technological and structural innovations to improving the engagement of diverse stakeholders, three themes recur throughout infrastructure resilience, risk preparedness, and disaster finance. First, planning in advance for how organizations will prepare for, respond to, and recover from disasters is essential, i.e. through the creation of BCPs by both public and private organizations. Second, pre-arranged agreements amongst organizations for sharing resources, knowledge, and financing in order to mitigate, prepare, respond and recover together from disasters and other unforeseen events are highly beneficial. Third, only with continuous reflection, learning and update on what worked and what didn’t work after each disasters can develop the adaptive capacities needed to manage ever increasing and unexpected risks. Preparedness is an incremental and interactive process.

These lessons from the GEJE on the importance of BCPs and pre-arranged agreements both emphasize larger principles that can be brought to bear in the context of emergent climate and public health crises. Both involve planning for the potential of disaster before it strikes. BCPs and pre-arranged agreements are both made under blue-sky conditions, which allow frameworks to be put in place for advanced mitigation and preparedness, and rapid post-disaster response and recovery. While it is impossible to know exactly what future crises a locale will face, these processes often have benefits that make places and organizations better able to act in the face of unlikely or unpredicted events. The lessons above regarding BCPs and pre-arranged agreements also highlight that neither the government nor the private sector alone have all the tools to prepare for and respond to disasters. Rather, the GEJE shows the importance of both public and private organizations adopting BCPs, and the value of creating pre-arranged agreements among and across public and private groups. By making disaster preparedness a key consideration for all organizations, and bringing diverse stakeholders together to make plans for when a crisis strikes, these strengthened networks and planning capacities have the potential to bear benefits not only in an emergency but in the everyday operations of organizations and countries.

Back to Top

Additional Resources

Program Overview

• Japan-World Bank Program on Mainstreaming Disaster Risk Management in Developing Countries

Reports and Case Studies Featuring Lessons from GEJE

• Learning from Megadisasters: Lessons from the Great East Japan Earthquake  (PDF)
• Lifelines: The Resilient Infrastructure Opportunity  (PDF)
• Resilient Water Supply and Sanitation Services: The Case of Japan  (PDF)
• Japan Case Study Report on Road Geohazard Risk Management  (PDF)
• Resilient Infrastructure PPPs  (PDF)
• Making Schools Resilient at Scale  (PDF)
• Converting Disaster Experience into a Safer Built Environment: The Case of Japan  (PDF)
• Resilient Cultural Heritage: Learning from the Japanese Experience  (PDF)
• Information and Communication Technology for Disaster Risk Management in Japan
• Resilient Industries in Japan : Lessons Learned in Japan on Enhancing Competitiveness in the Face of Disasters by Natural Hazards (PDF)
• Preparedness Maps for Community Resilience: Earthquakes. Experience from Japan  (PDF)
• Elders Leading the Way to Resilience  (PDF)
• Ibasho: Strengthening community-driven preparedness and resilience in Philippines and Nepal by leveraging Japanese expertise and experience  (PDF)
• Learning from Disaster Simulation Drills in Japan  (PDF)
• Catastrophe Insurance Programs for Public Assets  (PDF)
• PPP contract clauses unveiled: the World Bank’s 2017 Guidance on PPP Contractual Provisions
• Learning from Japan: PPPs for infrastructure resilience

Audiovisual Resources on GEJE and its Reconstruction Processes in English

• NHK documentary: 3/11-The Tsunami: The First 3 Days
• NHK: 342 Stories of Resilience and Remembrance
• Densho Road 3.11: Journey to Experience the Lessons from the Disaster - Tohoku, Japan
• Sendai City: Disaster-Resilient and Environmentally-Friendly City
• Sendai City: Eastern Coastal Area Today, 2019 Fall

[i]   Resilient Water Supply and Sanitation Services  report, p.63

[ii]   Resilient Water Supply and Sanitation Services  report, p.63

[iii]   Resilient Water Supply and Sanitation Services  report, p.8

[iv]   Resilient Water Supply and Sanitation Services  report, p.71

[v]   Resilient Water Supply and Sanitation Services  report, p.63

[vi]   Lifelines: The Resilient Infrastructure Opportunity  report, p.115

[vii] Lifelines: The Resilient Infrastructure Opportunity  report, p.133

[viii]   Japan Case Study Report on Road Geohazard Risk Management  report, p.30

[ix]   Resilient Infrastructure PPPs  report, p.8-9

[x]   Resilient Infrastructure PPPs  report, p.39-40

[xi]   Resilient Industries in Japan  report, p.153.

[xii]   Lifelines: The Resilient Infrastructure Opportunity  report, p. 132

[xiii]   Making Schools Resilient at Scale  report, p.24

[xiv]   Resilient Cultural Heritage  report, p.62

[xv]   Learning from Megadisasters  report, p.326

[xvi]   Resilient Cultural Heritage  report, p.69

[xvii]   Learning from Megadisasters  report, p.100

[xviii] Learning from Megadisasters  report, p.297.

[xix]  J-ALERT, Japan’s nationwide early warning system, had 46% implementation at GEJE, and in communities where it was implemented earthquake early warnings were successfully received. Following GEJE, GOJ invested heavily in J-ALERT adoption (JPY 14B), bearing 50% of implementation costs. In 2013 GOJ spent JPY 773M to implement J-ALERT in municipalities that could not afford the expense. In 2014 MIC heavily promoted the L-ALERT system (formerly “Public Information Commons”), achieving 100% adoption across municipalities. Since GEJE, Japan has updated the EEWS to include a hybrid method of earthquake prediction, improving the accuracy of predictions and warnings.

[xx]  Related resources: NHK, “#1 TSUNAMI BOSAI: Science that Can Save Your Life”  https://www3.nhk.or.jp/nhkworld/en/ondemand/video/3004665/  ; NHK “BOSAI: Be Prepared - Hazard Maps”  https://www3.nhk.or.jp/nhkworld/en/ondemand/video/2084002/

[xxi]  Aldrich, Daniel P., and Yasuyuki Sawada. "The physical and social determinants of mortality in the 3.11 tsunami." Social Science & Medicine 124 (2015): 66-75.

[xxii]   Learning from Disaster Simulation Drills in Japan  Report, p. 14

[xxiii]  Matsushita and Hideshima. 2014. “Influence over Financial Statement of Listed Manufacturing Companies by the GEJE, the Effect of BCP and Risk Financing.” [In Japanese.] Journal of Japan Society of Civil Engineering 70 (1): 33–43.  https://www.jstage.jst.go.jp/article/jscejsp/70/1/70_33/_pdf/-char/ja .

[xxiv]   Resilient Industries in Japan  report, p. 56

[xxv]  MIC (Ministry of Internal Affairs and Communications). 2019. “Survey Results of Business Continuity Plan Development Status in Local Governments.” [In Japanese.] Press release, MIC, Tokyo.  https://www.fdma.go.jp/pressrelease/houdou/items/011226bcphoudou.pdf .

[xxvi]   Japan Case Study Report on Road Geohazard Risk Management  report, p.17.

[xxvii]  The World Bank. 2021. “Financial Protection of Critical Infrastructure Services.” Technical Report – Contribution to 2020 APEC Finance Ministers Meeting.

[xxviii]   Resilient Industries in Japan  report, p. 119

[xxix]  Tokio Marine Holdings, Inc. 2019. “The Role of Insurance Industry to Strengthen Resilience of Infrastructure—Experience in Japan.” APEC seminar on Disaster Risk Finance.

[xxx]  TDB (Teikoku DataBank). 2018. “Trends in Bankruptcies 6 Years after the Great East Japan Earthquake.” [In Japanese.] TDB, Tokyo.  https://www.tdb.co.jp/report/watching/press/pdf/p170301.pdf .

[xxxi]  Matsushita and Hideshima. 2014. “Influence over Financial Statement of Listed Manufacturing Companies by the GEJE, the Effect of BCP and Risk Financing.” [In Japanese.] Journal of Japan Society of Civil Engineering 70 (1): 33–43.  https://www.jstage.jst.go.jp/article/jscejsp/70/1/70_33/_pdf/-char/ja .

[xxxii]   Resilient Industries in Japan  report, p. 145

Warning message

Welcome back! Log in to access your website

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• NEWS EXPLAINER
• 12 September 2023
• Clarification 13 September 2023
• Update 14 September 2023

Why was the Morocco earthquake so deadly?

• Michael Marshall 0

Michael Marshall is a science journalist in Devon, UK.

You can also search for this author in PubMed   Google Scholar

Collapsing buildings have been responsible for many of the casualties in the Morocco earthquake. Credit: Mohamed Messara/EPA-EFE/Shutterstock

Morocco is dealing with the aftermath of its most devastating earthquake for decades. The tremor, which hit on 8 September in the High Atlas mountain range, around 70 kilometres southwest of Marrakesh, has killed more than 2,800 people, with thousands more injured. The death toll seems likely to rise as rescue and recovery efforts continue.

Access options

Access Nature and 54 other Nature Portfolio journals

Get Nature+, our best-value online-access subscription

24,99 € / 30 days

cancel any time

Subscribe to this journal

Receive 51 print issues and online access

185,98 € per year

only 3,65 € per issue

Rent or buy this article

Prices vary by article type

Prices may be subject to local taxes which are calculated during checkout

doi: https://doi.org/10.1038/d41586-023-02880-3

Updates & Corrections

Clarification 13 September 2023 : This article has been updated to make clear that the 1,739 historical earthquakes did not all happen in Morocco, but their activity was felt there.

Update 14 September 2023 : This article was updated on 14 September with additional information from researchers in Morocco.

Peláez, J. A. et al. Seismol. Res. Lett. 78 , 614–621 (2007).

Article   Google Scholar  

Download references

Reprints and permissions

Related Articles

• Engineering

Corner- and edge-mode enhancement of near-field radiative heat transfer

Article 17 APR 24

NASA admits plan to bring Mars rocks to Earth won’t work — and seeks fresh ideas

News 15 APR 24

Selenium alloyed tellurium oxide for amorphous p-channel transistors

Article 10 APR 24

Seismological evidence for a multifault network at the subduction interface

Taiwan hit by biggest earthquake in 25 years: why scientists weren’t surprised

News 03 APR 24

Melting ice solves leap-second problem — for now

News & Views Forum 27 MAR 24

The ‘Anthropocene’ is here to stay — and it’s better not as a geological epoch

Correspondence 26 MAR 24

Are we in the Anthropocene yet?

Editorial 20 MAR 24

Research Associate - Brain Cancer

Houston, Texas (US)

Baylor College of Medicine (BCM)

Senior Manager, Animal Care

Research associate - genomics, postdoctoral associate- artificial intelligence, postdoctoral research fellow.

Looking for postdoctoral fellowship candidates to advance genomic research in the study of lung diseases at Brigham and Women's Hospital and HMS.

Boston, Massachusetts

Brigham and Women's Hospital (BWH)

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Earthquake case studies

Earthquake case studies Below are powerpoint presentations discussing the primary and secondary effects and immediate and long-term responses for both the Kobe, Japan and Kashmir, Pakistan earthquakes.

Effects of the Italian earthquake – http://www.bbc.co.uk/learningzone/clips/the-italian-earthquake-the-aftermath/6997.html Responses to Italian earthquake – http://www.bbc.co.uk/learningzone/clips/the-italian-earthquake-the-emergency-response/6998.html The Kobe earthquake – http://www.bbc.co.uk/learningzone/clips/the-kobe-earthquake/3070.html General effects & responses & Kobe (Rich) & Kashmir (Poor)

O Ltb Eartqaukes Cs from donotreply16 Kobe earthquake (Rich country)

Koberevision from cheergalsal Haiti 2010 – Poor country Picture Facts On 12th January, an earthquake measuring 7.0 on the Richter scale struck close to Haiti’s capital Port-au-Prince The earthquake occurred at a destructive plate margin between the Caribbean and North American Plates, along a major fault line. The earthquakes focus was 13km underground, and the epicentre was just 25km from Port-au-Prince Haiti has suffered a large number of serious aftershocks after the main earthquake

Primary effects About 220,000 people were killed and 300,000 injured The main port was badly damaged, along with many roads that were blocked by fallen buildings and smashed vehicles Eight hospitals or health centres in Port-au-Prince collapsed or were badly damaged. Many government buildings were also destroyed About 100,000 houses were destroyed and 200,000 damaged in Port-au-Prince and the surrounding area. Around 1.3 million Haitians were displaced (left homeless)

Secondary effects Over 2 million Habitats were left without food and water. Looting became a serious problem The destruction of many government buildings hindered the government’s efforts to control Haiti, and the police force collapsed The damage to the port and main roads meant that critical aid supplies for immediate help and longer-term reconstruction were prevented from arriving or being distributed effectively Displaced people moved into tents and temporary shelters, and there were concerns about outbreaks of disease. By November 2010, there were outbreaks of Cholera There were frequent power cuts The many dead bodies in the streets, and under the rubble, created a health hazard in the heat. So many had to be buried in mass graves

Short-term responses The main port and roads were badly damaged, crucial aid (such as medical supplies and food) was slow to arrive and be distributed. The airport couldn’t handle the number of planes trying to fly in and unload aid American engineers and diving teams were used to clear the worst debris and get the port working again, so that waiting ships could unload aid The USA sent ships, helicopters, 10,000 troops, search and rescue teams and $100 million in aid The UN sent troops and police and set up a Food Aid Cluster to feed 2 million people Bottled water and water purification tablets were supplied to survivors Field hospitals were set up and helicopters flew wounded people to nearby countries The Haitian government moved 235,000 people from Port-au-Prince to less damaged cities

Long-term responses Haiti is dependent on overseas aid to help it recover New homes would need to be built to a higher standard, costing billions of dollars Large-scale investment would be needed to bring Haiti’s road, electricity, water and telephone systems up to standard, and to rebuild the port Sichuan, China 2008 – Poor country case study Picture On 12th May at 14:28pm, the pressure resulting from the Indian Plate colliding with the Eurasian Plate was released along the Longmenshan fault line that runs beneath. This led to an earthquake measuring 7.9 on the Richter scale with tremors lasting 120 seconds.

Primary effects · 69,000 people were killed · 18,000 missing · 374,000 were injured · between 5 -11 million people were missing · 80% of buildings collapsed in rural areas such as Beichuan county due to poorer building standards · 5 million buildings collapsed

Secondary effects · Communication were brought to a halt – neither land nor mobile phones worked in Wenchuan · Roads were blocked and damaged and some landslides blocked rivers which led to flooding · Fires were caused as gas pipes burst · Freshwater supplies were contaminated by dead bodies

Immediate responses · 20 helicopters were assigned to rescue and relief effects immediately after the disaster · Troops parachuted in or hiked to reach survivors · Rescuing survivors trapped in collapsed buildings was a priority · Survivors needed food, water and tents to shelter people from the spring rains. 3.3 million new tents were ordered.

Long-term responses · Aid donations specifically money – over £100 million were raised by the Red Cross · One million temporary small were built to house the homeless · The Chinese government pledged a $10 million rebuilding funds and banks wrote off debts by survivors who did not have insurance

• school Campus Bookshelves
• menu_book Bookshelves
• perm_media Learning Objects
• login Login
• how_to_reg Request Instructor Account
• hub Instructor Commons
• Download Page (PDF)
• Download Full Book (PDF)
• Periodic Table
• Physics Constants
• Scientific Calculator
• Reference & Cite
• Tools expand_more
• Readability

selected template will load here

This action is not available.

3.9: Case Studies

• Last updated
• Save as PDF
• Page ID 11215

• Chris Johnson, Matthew D. Affolter, Paul Inkenbrandt, & Cam Mosher
• Salt Lake Community College via OpenGeology

Video explaining the seismic activity and hazards of the Intermountain Seismic Belt and the Wasatch Fault, a large intraplate area of seismic activity.

North American Earthquakes

Basin and Range Earthquakes —Earthquakes in the Basin and Range Province, from the Wasatch Fault (Utah) to the Sierra Nevada (California), occur primarily in normal faults created by tensional forces. The Wasatch Fault, which defines the eastern extent of the Basin and Range province, has been studied as an earthquake hazard for more than 100 years.

New Madrid Earthquakes (1811-1812) —Historical accounts of earthquakes in the New Madrid seismic zone date as far back as 1699 and earthquakes continue to be reported in modern times [ 11 ]. A sequence of large (M w >7) occurred from December 1811 to February 1812 in the New Madrid area of Missouri [ 12 ]. The earthquakes damaged houses in St. Louis, affected the stream course of the Mississippi River, and leveled the town of New Madrid. These earthquakes were the result of intraplate seismic activity [ 9 ].

Charleston (1868) —The 1868 earthquake in Charleston South Carolina was a moment magnitude 7.0, with a Mercalli intensity of X, caused significant ground motion, and killed at least 60 people. This intraplate earthquake was likely associated with ancient faults created during the breakup of Pangea. The earthquake caused significant liquefaction [ 13 ]. Scientists estimate the recurrence of destructive earthquakes in this area with an interval of approximately 1500 to 1800 years.

Great San Francisco Earthquake and Fire (1906) —On April 18, 1906, a large earthquake, with an estimated moment magnitude of 7.8 and MMI of X, occurred along the San Andreas fault near San Francisco California. There were multiple aftershocks followed by devastating fires, resulting in about 80% of the city being destroyed. Geologists G.K. Gilbert and Richard L. Humphrey, working independently, arrived the day following the earthquake and took measurements and photographs [ 14 ].

Alaska (1964) —The 1964 Alaska earthquake, moment magnitude 9.2, was one of the most powerful earthquakes ever recorded. The earthquake originated in a megathrust fault along the Aleutian subduction zone. The earthquake caused large areas of land subsidence and uplift, as well as significant mass wasting.

Video from the USGS about the 1964 Alaska earthquake.

Loma Prieta (1989) —The Loma Prieta, California, earthquake was created by movement along the San Andreas Fault. The moment magnitude 6.9 earthquake was followed by a magnitude of 5.2 aftershock. It caused 63 deaths, buckled portions of the several freeways, and collapsed part of the San Francisco-Oakland Bay Bridge.

This video shows how shaking propagated across the Bay Area during the 1989 Loma Prieta earthquake.

This video shows the destruction caused by the 1989 Loma Prieta earthquake.

Global Earthquakes

Many of history’s largest earthquakes occurred in megathrust zones, such as the Cascadia Subduction Zone (Washington and Oregon coasts) and Mt. Rainier (Washington).

Shaanxi, China (1556) —On January 23, 1556 an earthquake of an approximate moment magnitude 8 hit central China, killing approximately 830,000 people in what is considered the most deadly earthquake in history. The high death toll was attributed to the collapse of cave dwellings ( yaodong ) built in loess deposits, which are large banks of windblown, compacted sediment (see Chapter 5 ). Earthquakes in this are region are believed to have a recurrence interval of 1000 years. [ 15 ].

Lisbon, Portugal (1755) —On November 1, 1755 an earthquake with an estimated moment magnitude range of 8–9 struck Lisbon, Portugal [ 13 ], killing between 10,000 to 17,400 people [ 16 ]. The earthquake was followed by a tsunami.

Valdivia, Chile (1960) —The May 22, 1960 earthquake was the most powerful earthquake ever measured, with a moment magnitude of 9.4–9.6 and lasting an estimated 10 minutes. It triggered tsunamis that destroyed houses across the Pacific Ocean in Japan and Hawaii and caused vents to erupt on the Puyehue-Cordón Caulle (Chile).

Video describing the tsunami produced by the 1960 Chili earthquake.

Tangshan, China (1976) —Just before 4 a.m. (Beijing time) on July 28, 1976 a moment magnitude 7.8 earthquake struck Tangshan (Hebei Province), China, and killed more than 240,000 people. The high death toll is attributed to people still being asleep or at home and most buildings being made of URM.

Sumatra, Indonesia (2004) —On December 26, 2004, slippage of the Sunda megathrust fault generated a moment magnitude 9.0–9.3 earthquake off the coast of Sumatra, Indonesia [ 17 ]. This megathrust fault is created by the Australia plate subducting below the Sunda plate in the Indian Ocean [ 18 ]. The resultant tsunamis created massive waves as tall as 24 m (79 ft) when they reached the shore and killed more than an estimated 200,000 people along the Indian Ocean coastline.

Haiti (2010) —The moment magnitude 7 earthquake that occurred on January 12, 2010, was followed by many aftershocks of magnitude 4.5 or higher. More than 200,000 people are estimated to have died as a result of the earthquake. The widespread infrastructure damage and crowded conditions contributed to a cholera outbreak, which is estimated to have caused thousands more deaths.

Tōhoku, Japan (2011) —Because most Japanese buildings are designed to tolerate earthquakes, the moment magnitude 9.0 earthquake on March 11, 2011, was not as destructive as the tsunami it created. The tsunami caused more than 15,000 deaths and tens of billions of dollars in damage, including the destructive meltdown of the Fukushima nuclear power plant.

9. Hildenbrand TG, Hendricks JD (1995) Geophysical setting of the Reelfoot rift and relations between rift structures and the New Madrid seismic zone. U.S. Geological Survey, Washington; Denver, CO

11. Feldman J (2012) When the Mississippi Ran Backwards: Empire, Intrigue, Murder, and the New Madrid Earthquakes of 1811 and 1812. Free Press

12. Fuller ML (1912) The New Madrid earthquake. Central United States Earthquake Consortium, Washington, D.C.

13. Talwani P, Cox J (1985) Paleoseismic evidence for recurrence of Earthquakes near Charleston, South Carolina. Science 229:379–381

14. Gilbert GK, Holmes JA, Humphrey RL, et al (1907) The San Francisco earthquake and fire of April 18, 1906 and their effects on structures and structural materials. U.S. Geological Survey, Washington, D.C.

15. Boer JZ de, Sanders DT (2007) Earthquakes in human history: The far-reaching effects of seismic disruptions. Princeton University Press, Princeton

16. Aguirre B.E. (2012) Better disaster statistics: The Lisbon earthquake. J Interdiscip Hist 43:27–42

17. Rossetto T, Peiris N, Pomonis A, et al (2007) The Indian Ocean tsunami of December 26, 2004: observations in Sri Lanka and Thailand. Nat Hazards 42:105–124

18. Satake K, Atwater BF (2007) Long-Term Perspectives on Giant Earthquakes and Tsunamis at Subduction Zones. Annual Review of Earth and Planetary Sciences 35:349–374. https://doi.org/10.1146/annurev.earth.35.031306.140302

Case Study: Predicting the Next Big Earthquake

Recent earthquake activity.

USGS Recent Worldwide Earthquake Activity To explore individual earthquakes in more depth, click on the UTC Date-Time field. Show me how Hide Details for accessing USGS Recent Worldwide Earthquake Activity Scroll the list to look over earthquakes that have occurred in the last seven days. To explore individual earthquakes in more depth, follow the COMMENTS links. Scroll to the bottom of the list to view recent Earthquakes plotted on a world map. What is the magnitude of the most recent recorded earthquake? How many earthquakes were recorded for the last seven days? Of those earthquakes, how many were of a magnitude 7.0 or greater? IRIS Seismic Monitor Click on the map to zoom to specific regions. Click on individual earthquakes to see lists of others nearby. Show me how Hide Details for accessing the IRIS Seismic Monitor Click on the map to zoom to specific regions. Click on individual earthquakes to see lists of others nearby. Where are earthquakes concentrated?

Where does Earth Quake?

Predicting the Next Big One!

« Previous Page       Next Page »

• Investigating Earthquakes with ArcExplorer GIS
• Teaching Notes
• Step-by-Step Instructions
• Tool and Data
• Going Further

• About this Site
• Accessibility

Citing and Terms of Use

Material on this page is offered under a Creative Commons license unless otherwise noted below.

Show terms of use for text on this page »

Show terms of use for media on this page »

• None found in this page
• Last Modified: February 27, 2024
• Short URL: https://serc.carleton.edu/18034 What's This?

The Parkfield, California, Earthquake Experiment

September 28, 2004— m 6.0 earthquake captured.

The Parkfield Experiment is a comprehensive, long-term earthquake research project on the San Andreas fault. Led by the USGS and the State of California, the experiment's purpose is to better understand the physics of earthquakes - what actually happens on the fault and in the surrounding region before, during and after an earthquake. Ultimately, scientists hope to better understand the earthquake process and, if possible, to provide a scientific basis for earthquake prediction. Since its inception in 1985, the experiment has involved more than 100 researchers at the USGS and collaborating universities and government laboratories. Their coordinated efforts have led to a dense network of instruments poised to "capture" the anticipated earthquake and reveal the earthquake process in unprecedented detail.

Moderate-size earthquakes of about magnitude 6 have occurred on the Parkfield section of the San Andreas fault at fairly regular intervals - in 1857, 1881, 1901, 1922, 1934, and 1966. The first, in 1857, was a foreshock to the great Fort Tejon earthquake which ruptured the fault from Parkfield to the southeast for over 180 miles. Available data suggest that all six moderate-sized Parkfield earthquakes may have been "characteristic" in the sense that they all ruptured the same area on the fault. If such characteristic ruptures occur regularly, then the next quake would have been due before 1993.

These pages describe the scientific background for the experiment, including the tectonic setting at Parkfield, the historical earthquake activity on this section of the San Andreas fault, the monitoring and data collecting activities currently being carried out, and plans for future research. Data are available to view in real-time and download.

Scientific Advances

While the greatest scientific payoff is expected when the earthquake occurs, our understanding of the earthquake process has already been advanced through research results from Parkfield. Some of the highlights are described.

Real-time data from instrumentation networks running at Parkfield are available for viewing and downloading.

Parkfield Earthquake Shake Table Exhibit

The Art-Science of Earthquakes by D.V. Rogers November 23, 2009 ( video )

The exhibit was a geologically interactive, seismic machine earthwork temporarily installed in Parkfield in 2008. Rogers presented the history, conceptual premise, documentation of the work, and also put forward the idea of how early 21st century cultural practice could be used to encourage earthquake awareness and preparedness.

Pictures and interactive, 360-degree panorama .

Lessons From the Best-Recorded Quake in History

USGS Public Lecture by Andy Michael October 26, 2006 ( video )

New data from the 2004 Parkfield earthquake provide important lessons about earthquake processes, prediction, and the hazards assessments that underlie building codes and mitigation policies.

Research Scientist: John Langbein , Earthquake Science Center.

Bhuj Earthquake India 2001 – A Complete Study

Bhuj earthquake india.

Gujarat : Disaster on a day of celebration : 51st Republic Day on January 26, 2001

• 7.9 on the Richter scale.
• 8.46 AM January 26th 2001
• 20,800 dead

Basic Facts

• Earthquake: 8:46am on January 26, 2001
• Epicenter: Near Bhuj in Gujarat, India
• Magnitude: 7.9 on the Richter Scale

Geologic Setting

• Indian Plate Sub ducting beneath Eurasian Plate
• Continental Drift
• Convergent Boundary

Specifics of 2001 Quake

Compression Stress between region’s faults

Depth: 16km

Probable Fault: Kachchh Mainland

Fault Type: Reverse Dip-Slip (Thrust Fault)

The earthquake’s epicentre was 20km from Bhuj. A city with a population of 140,000 in 2001. The city is in the region known as the Kutch region. The effects of the earthquake were also felt on the north side of the Pakistan border, in Pakistan 18 people were killed.

Tectonic systems

The earthquake was caused at the convergent plate boundary between the Indian plate and the Eurasian plate boundary. These pushed together and caused the earthquake. However as Bhuj is in an intraplate zone, the earthquake was not expected, this is one of the reasons so many buildings were destroyed – because people did not build to earthquake resistant standards in an area earthquakes were not thought to occur. In addition the Gujarat earthquake is an excellent example of liquefaction, causing buildings to ‘sink’ into the ground which gains a consistency of a liquid due to the frequency of the earthquake.

India : Vulnerability to earthquakes

• 56% of the total area of the Indian Republic is vulnerable to seismic activity .
• 12% of the area comes under Zone V (A&N Islands, Bihar, Gujarat, Himachal Pradesh, J&K, N.E.States, Uttaranchal)
• 18% area in Zone IV (Bihar, Delhi, Gujarat, Haryana, Himachal Pradesh, J&K, Lakshadweep, Maharashtra, Punjab, Sikkim, Uttaranchal, W. Bengal)
• 26% area in Zone III (Andhra Pradesh, Bihar, Goa, Gujarat, Haryana, Kerala, Maharashtra, Orissa, Punjab, Rajasthan, Tamil Nadu, Uttaranchal, W. Bengal)
• Gujarat: an advanced state on the west coast of India.
• On 26 January 2001, an earthquake struck the Kutch district of Gujarat at 8.46 am.
• Epicentre 20 km North East of Bhuj, the headquarter of Kutch.
• The Indian Meteorological Department estimated the intensity of the earthquake at 6.9 Richter. According to the US Geological Survey, the intensity of the quake was 7.7 Richter.
• The quake was the worst in India in the last 180 years.

What earthquakes do

• Casualties: loss of life and injury.
• Loss of housing.
• Damage to infrastructure.
• Disruption of transport and communications.
• Breakdown of social order.
• Loss of industrial output.
• Loss of business.
• Disruption of marketing systems.
• The earthquake devastated Kutch. Practically all buildings and structures of Kutch were brought down.
• Ahmedabad, Rajkot, Jamnagar, Surendaranagar and Patan were heavily damaged.
• Nearly 19,000 people died. Kutch alone reported more than 17,000 deaths.
• 1.66 lakh people were injured. Most were handicapped for the rest of their lives.
• The dead included 7,065 children (0-14 years) and 9,110 women.
• There were 348 orphans and 826 widows.

Loss classification

Deaths and injuries: demographics and labour markets

Effects on assets and GDP

Effects on fiscal accounts

Financial markets

Disaster loss

• Initial estimate Rs. 200 billion.
• Came down to Rs. 144 billion.
• No inventory of buildings
• Non-engineered buildings
• Land and buildings
• Stocks and flows
• Reconstruction costs (Rs. 106 billion) and loss estimates (Rs. 99 billion) are different
• Public good considerations

Human Impact: Tertiary effects

• Affected 15.9 million people out of 37.8 in the region (in areas such as Bhuj, Bhachau, Anjar, Ganhidham, Rapar)
• High demand for food, water, and medical care for survivors
• Humanitarian intervention by groups such as Oxfam: focused on Immediate response and then rehabilitation
• Of survivors, many require persistent medical attention
• Region continues to require assistance long after quake has subsided
• International aid vital to recovery

Social Impacts

• 80% of water and food sources were destroyed.
• The obvious social impacts are that around 20,000 people were killed and near 200,000 were injured.
• However at the same time, looting and violence occurred following the quake, and this affected many people too.
• On the other hand, the earthquake resulted in millions of USD in aid, which has since allowed the Bhuj region to rebuild itself and then grow in a way it wouldn’t have done otherwise.
• The final major social effect was that around 400,000 Indian homes were destroyed resulting in around 2 million people being made homeless immediately following the quake.

Social security and insurance

• Ex gratia payment: death relief and monetary benefits to the injured
• Major and minor injuries
•  Cash doles
• Government insurance fund
• Group insurance schemes
• Claim ratio

Demographics and labour market

• Geographic pattern of ground motion, spatial array of population and properties at risk, and their risk vulnerabilities.
• Low population density was a saving grace.
• Extra fatalities among women
• Effect on dependency ratio
• Farming and textiles

Economic  Impacts

• Total damage estimated at around $7 billion. However $18 billion of aid was invested in the Bhuj area.
• Over 15km of tarmac road networks were completely destroyed.
• In the economic capital of the Gujarat region, Ahmedabad, 58 multi storey buildings were destroyed, these buildings contained many of the businesses which were generating the wealth of the region.
• Many schools were destroyed and the literacy rate of the Gujarat region is now the lowest outside southern India.

Impact on GDP

• Applying ICOR
• Rs. 99 billion – deduct a third as loss of current value added.
• Get GDP loss as Rs. 23 billion
• Adjust for heterogeneous capital, excess capacity, loss Rs. 20 billion.
• Reconstruction efforts.
• Likely to have been Rs. 15 billion.

Fiscal accounts

• Differentiate among different taxes: sales tax, stamp duties and registration fees, motor vehicle tax, electricity duty, entertainment tax, profession tax, state excise and other taxes. Shortfall of Rs. 9 billion of which about Rs. 6 billion unconnected with earthquake.
• Earthquake related other flows.
• Expenditure:Rs. 8 billion on relief. Rs. 87 billion on rehabilitation.

Impact on Revenue Continue Reading

Comments are closed.

Privacy Overview

Advertisement

Supported by

Turkey Earthquake Trial Opens Amid Anger and Tears

More than 300 people were killed when temblors toppled an upscale residential complex. Survivors hope a court will punish the men who built it.

• Share full article

By Safak Timur and Ben Hubbard

Safak Timur reported from the courtroom in Antakya, Turkey, and Ben Hubbard from Istanbul.

The families addressed the court one by one, sobbing as they spoke the names of relatives who had been killed when their upscale apartment complex in southern Turkey toppled over during a powerful earthquake last year.

One woman, whose son had died in the collapse alongside his wife and their 3-year-old son, lashed out at the defendants — the men who had built the complex and the inspectors charged with ensuring that it was safe.

“Shame on you,” said the woman, Remziye Bozdemir. “Your children are alive, mine are dead.”

The hearing on Thursday was the first aimed at seeking accountability for the collapse of Renaissance Residence, one of the most catastrophic building failures during the earthquakes of Feb. 6, 2023 , which damaged hundreds of thousands of structures and killed more than 53,000 people across southern Turkey.

More than 300 people died inside Renaissance, and many more were wounded. An investigation and forensic analysis by The New York Times found that a tragic combination of poor design and minimal oversight had left the building vulnerable, ultimately causing its 13 stories to smash into the earth.

Since the quakes, the anger of many survivors has centered on the lax construction practices that allowed so many defective buildings to rise across a region with a history of powerful temblors. When the ground shook last year, many structures became death traps, pancaking down on their residents and killing them instantly or trapping them alive inside the rubble .

In recent months, Turkish courts have begun hearing cases seeking to assign responsibility for the deadly collapses. The Renaissance trial is one such case, which illustrates what victims’ advocates say are the limits of post-quake justice.

Eight men — four from the construction company and four employees of a private building-inspection firm — stand accused of causing foreseeable death and injury through negligence for their roles in the construction of the complex. All eight have pleaded not guilty.

Missing from the case are any of the numerous public officials who allowed the complex to be built by zoning the land, approving building plans and issuing construction permits, together failing to ensure that the project had been constructed to withstand violent quakes.

This scrutiny of private builders but not of public officials has marred efforts to ensure accountability across the quake zone, said Emma Sinclair-Webb, the Turkey director for Human Rights Watch.

“Contractors can be cowboy builders, building defective buildings, but what about the enabling environment in which they operate and the public authorities that turn a blind eye and let them go ahead?” she said.

Complicating efforts to hold such officials accountable is a Turkish law that prevents prosecutors from investigating state employees without obtaining government permission.

It remains unclear whether any public officials are on trial in earthquake-related cases.

In January, Human Rights Watch and Citizens’ Assembly , a Turkish rights group, filed requests in dozens of jurisdictions seeking information about how many requests to investigate public officials had been made and how many had been granted. Their queries turned up four instances in which decisions were pending and three in which permission to investigate had been granted, although two of those had been appealed, the groups said in a report last month.

Most jurisdictions declined to respond, citing confidentiality regulations.

This diminishes the chances for true accountability, Ms. Sinclair-Webb said.

“The full facts are not really there to be looked at if the public officials are left out of the picture,” she said.

Renaissance Residence rose on a patch of converted agricultural land near the ancient city of Antakya during a construction boom that was sweeping through the area in the 2010s, fueled by the plans of Turkey’s leader, Recep Tayyip Erdogan, for development and economic growth. By the time residents arrived in 2013, the three apartment towers, superficially joined to appear as one long, thin building, loomed over the countryside.

The complex catered to the area’s rising middle class, with a pool, underground parking and a lobby designed to mimic that of a hotel. Many early occupants considered themselves lucky to live there.

But The Times’s investigation found that, despite its air of glamour, Renaissance was rife with risky design choices that were cast in concrete with minimal oversight, leaving the structure ill prepared to withstand a powerful earthquake.

The first such quake stuck last year, with a magnitude of 7.8, followed by a second powerful temblor hours later. The first quake caused the ground floor of Renaissance to fail, making the building topple on its side and destroying many of its residents’ lives.

Cemile Incili, 59, a real estate agent who attended the hearing on Thursday, said she had survived the collapse with some injuries but had been able to hear her nephew trapped in the rubble.

“Aunt, I can’t breathe,” she recalled him saying.

His body, and that of a sister of Ms. Incili, were never recovered from the wreckage. She assumes they are dead.

She hoped the trial would mean long sentences for the men who built Renaissance as well as for the officials who allowed the building to rise.

“The state did not protect our lives or our property,” she said.

Court documents say that 269 people have been identified as having been killed in the building and that 46 others are still missing and assumed to be dead.

Prosecutors have charged the eight defendants with conscious negligence that caused multiple deaths and injuries. If convicted, they could face up to 22 years in prison.

The prosecutors have accused the contractors who built Renaissance of failing to follow the building codes in place at the time, using substandard materials and neglecting to ensure that the structure was sound. They have accused the inspectors, who worked for a private company hired by the contractors, of failing to detect flaws that should have been reported to the authorities.

The contractor who was the construction company’s lead architect, Mehmet Yasar Coskun, told the court on Thursday that he rejected the allegations. He blamed the collapse on the exceptional power of the earthquake’s shaking at the site.

“As the foundation of the building was strong, the wave demolished it from the weakest point it could find, the ground floor,” he said. “It is an atypical situation.”

Other defendants, too, said they had followed all the necessary regulations and attributed the collapse to the earthquake’s strength.

Their arguments failed to convince the survivors who attended the hearing.

Hafize Acikgoz, 42, had made it out of Renaissance alive but lost her husband and three children, who were 16, 21 and 23.

“It is just me left behind,” she said, wiping away tears. “Nothing can sooth my pain and nothing can bring them back.”

Still, she hoped that the accused would receive the longest possible sentences.

“Shouldn’t those buildings have been built considering people’s lives?” she said.

Beril Eski contributed reporting from Istanbul.

Ben Hubbard is the Istanbul bureau chief, covering Turkey and the surrounding region. More about Ben Hubbard

Case Studies: The L'Aquila & Kashmir Earthquakes

Earthquakes in high income countries - l'aquila, italy.

On the 6 th of April 2009, there was an earthquake with a magnitude of 6.3 in a town called L'Aquila in the Abruzzi region in Italy.

Primary effects - deaths and damage

• 308 people died and about 1,600 people were injured.
• More than 65,000 became homeless.
• The water supply into the Paganica (a town) was damaged, cutting them off from vital water supplies.

Secondary effects - aftershocks and infrastructure

• There were aftershocks that caused further damage after the initial earthquake.
• All telecommunications (phone) and electricity infrastructure was up and running in less than 24 hours.

Immediate responses - shelter and support

• Homeless people were given shelter, food, drinks, and medical attention. They also got free mobiles to communicate with their families.
• The army, medical personnel, and the fire department all helped clear the wreckage.
• The immediate response was helped by the fact that L'Aquila is closer than 100km to Rome and Italy is a relatively rich country.

Long-term responses - rebuilding

• The city centre was rebuilt to try to rehouse the 65,000 people who had become homeless.
• The inability of modern buildings to cope with earthquakes was investigated.
• 7 people were tried for manslaughter for not giving strong enough warnings about the earthquake.

Earthquake in Low Income Country - Kashmir, Pakistan

On the 8 th of October 2005, there was an earthquake with a magnitude of 7.6 in Pakistan (low-income country).

Primary effects of the Kashmir earthquake

• 79,000 people died and lots of buildings crumbled to the ground.
• It is hard to find an exact figure, but people estimate that 4 million people became homeless.
• Infrastructure was damaged. Millions of people had no clean water and no electricity.

Secondary effects of the Kashmir earthquake

• Landslides killed people and destroyed towns.
• Sewage pipes broke. This spread contaminated water and disease.
• The winter of 2005-2006 was very cold. 4 million people became homeless and lots of the homeless froze to death during the winter.

Immediate response to the Kashmir earthquake

• Charities and foreign governments sent funds, aid workers and helicopters.
• Charities gave out warm clothes, and tents, but a lot of support took a month to arrive because of the cold weather, damaged infrastructure, and the high number of people affected.

Long-term response to the Kashmir earthquake

• Thousands of people were relocated to new settlements, but 4 million people had been made homeless.
• The Pakistan government gave people money to try to rebuild their houses and homes, but because they were starving to death, they were forced to spend money on food instead.
• Thousands of people still lived in tents in 2015, a decade later.
• The government changed building regulations to try to stop this damage happening again.

Cause of the Kashmir earthquake

• Running through the middle of Pakistan is a collision plate boundary between the Eurasian and Indian plates, which means that Pakistan is prone to seismic activity.
• These plates have folded and forced each other upwards to form the Himalayan fold mountain range.
• The strain at this boundary was suddenly released on 8th October, 2005.

1 The Challenge of Natural Hazards

1.1 Natural Hazards

1.1.1 Types of Natural Hazards

1.1.2 Hazard Risk

1.1.3 Consequences of Natural Hazards

1.1.4 End of Topic Test - Natural Hazards

1.1.5 Exam-Style Questions - Natural Hazards

1.2 Tectonic Hazards

1.2.1 Tectonic Plates

1.2.2 Tectonic Plates & Convection Currents

1.2.3 Plate Margins

1.2.4 Volcanoes

1.2.5 Effects of Volcanoes

1.2.6 Responses to Volcanic Eruptions

1.2.7 Earthquakes

1.2.8 Earthquakes 2

1.2.9 Responses to Earthquakes

1.2.10 Case Studies: The L'Aquila & Kashmir Earthquakes

1.2.11 Earthquake Case Study: Chile 2010

1.2.12 Earthquake Case Study: Nepal 2015

1.2.13 Living with Tectonic Hazards 1

1.2.14 Living with Tectonic Hazards 2

1.2.15 End of Topic Test - Tectonic Hazards

1.2.16 Exam-Style Questions - Tectonic Hazards

1.2.17 Tectonic Hazards - Statistical Skills

1.3 Weather Hazards

1.3.1 Global Atmospheric Circulation

1.3.2 Surface Winds

1.3.3 UK Weather Hazards

1.3.4 Tropical Storms

1.3.5 Features of Tropical Storms

1.3.6 Impact of Tropical Storms 1

1.3.7 Impact of Tropical Storms 2

1.3.8 Tropical Storms Case Study: Katrina

1.3.9 Tropical Storms Case Study: Haiyan

1.3.10 UK Weather Hazards Case Study: Somerset 2014

1.3.11 End of Topic Test - Weather Hazards

1.3.12 Exam-Style Questions - Weather Hazards

1.3.13 Weather Hazards - Statistical Skills

1.4 Climate Change

1.4.1 Evidence for Climate Change

1.4.2 Causes of Climate Change

1.4.3 Effects of Climate Change

1.4.4 Managing Climate Change

1.4.5 End of Topic Test - Climate Change

1.4.6 Exam-Style Questions - Climate Change

1.4.7 Climate Change - Statistical Skills

2 The Living World

2.1 Ecosystems

2.1.1 Ecosystems

2.1.2 Ecosystem Cascades & Global Ecosystems

2.1.3 Ecosystem Case Study: Freshwater Ponds

2.2 Tropical Rainforests

2.2.1 Tropical Rainforests - Intro & Interdependence

2.2.2 Adaptations

2.2.3 Biodiversity of Tropical Rainforests

2.2.4 Deforestation

2.2.5 Case Study: Deforestation in the Amazon Rainforest

2.2.6 Sustainable Management of Rainforests

2.2.7 Case Study: Malaysian Rainforest

2.2.8 End of Topic Test - Tropical Rainforests

2.2.9 Exam-Style Questions - Tropical Rainforests

2.2.10 Deforestation - Statistical Skills

2.3 Hot Deserts

2.3.1 Overview of Hot Deserts

2.3.2 Biodiversity & Adaptation to Hot Deserts

2.3.3 Case Study: Sahara Desert

2.3.4 Desertification

2.3.5 Case Study: Thar Desert

2.3.6 End of Topic Test - Hot Deserts

2.3.7 Exam-Style Questions - Hot Deserts

2.4 Tundra & Polar Environments

2.4.1 Overview of Cold Environments

2.4.2 Adaptations in Cold Environments

2.4.3 Biodiversity in Cold Environments

2.4.4 Case Study: Alaska

2.4.5 Sustainable Management

2.4.6 Case Study: Svalbard

2.4.7 End of Topic Test - Tundra & Polar Environments

2.4.8 Exam-Style Questions - Cold Environments

3 Physical Landscapes in the UK

3.1 The UK Physical Landscape

3.1.1 The UK Physical Landscape

3.2 Coastal Landscapes in the UK

3.2.1 Types of Wave

3.2.2 Weathering & Mass Movement

3.2.3 Processes of Erosion & Wave-Cut Platforms

3.2.4 Headlands, Bays, Caves, Arches & Stacks

3.2.5 Transportation

3.2.6 Deposition

3.2.7 Spits, Bars & Sand Dunes

3.2.8 Case Study: Landforms on the Dorset Coast

3.2.9 Types of Coastal Management 1

3.2.10 Types of Coastal Management 2

3.2.11 Coastal Management Case Study - Holderness

3.2.12 Coastal Management Case Study: Swanage

3.2.13 Coastal Management Case Study - Lyme Regis

3.2.14 End of Topic Test - Coastal Landscapes in the UK

3.2.15 Exam-Style Questions - Coasts

3.3 River Landscapes in the UK

3.3.1 The River Valley

3.3.2 River Valley Case Study - River Tees

3.3.3 Erosion

3.3.4 Transportation & Deposition

3.3.5 Waterfalls, Gorges & Interlocking Spurs

3.3.6 Meanders & Oxbow Lakes

3.3.7 Floodplains & Levees

3.3.8 Estuaries

3.3.9 Case Study: The River Clyde

3.3.10 River Management

3.3.11 Hard & Soft Flood Defences

3.3.12 River Management Case Study - Boscastle

3.3.13 River Management Case Study - Banbury

3.3.14 End of Topic Test - River Landscapes in the UK

3.3.15 Exam-Style Questions - Rivers

3.4 Glacial Landscapes in the UK

3.4.1 Erosion

3.4.2 Landforms Caused by Erosion

3.4.3 Landforms Caused by Transportation & Deposition

3.4.4 Snowdonia

3.4.5 Land Use in Glaciated Areas

3.4.6 Tourism in Glacial Landscapes

3.4.7 Case Study - Lake District

3.4.8 End of Topic Test - Glacial Landscapes in the UK

3.4.9 Exam-Style Questions - Glacial Landscapes

4 Urban Issues & Challenges

4.1 Urban Issues & Challenges

4.1.1 Urbanisation

4.1.2 Urbanisation Case Study: Lagos

4.1.3 Urbanisation Case Study: Rio de Janeiro

4.1.4 UK Cities

4.1.5 Case Study: Urban Regen Projects - Manchester

4.1.6 Case Study: Urban Change in Liverpool

4.1.7 Case Study: Urban Change in Bristol

4.1.8 Sustainable Urban Life

4.1.9 End of Topic Test - Urban Issues & Challenges

4.1.10 Exam-Style Questions - Urban Issues & Challenges

4.1.11 Urban Issues -Statistical Skills

5 The Changing Economic World

5.1 The Changing Economic World

5.1.1 Measuring Development

5.1.2 Classifying Countries Based on Wealth

5.1.3 The Demographic Transition Model

5.1.4 Physical & Historical Causes of Uneven Development

5.1.5 Economic Causes of Uneven Development

5.1.6 How Can We Reduce the Global Development Gap?

5.1.7 Case Study: Tourism in Kenya

5.1.8 Case Study: Tourism in Jamaica

5.1.9 Case Study: Economic Development in India

5.1.10 Case Study: Aid & Development in India

5.1.11 Case Study: Economic Development in Nigeria

5.1.12 Case Study: Aid & Development in Nigeria

5.1.13 Economic Development in the UK

5.1.14 Economic Development UK: Industry & Rural

5.1.15 Economic Development UK: Transport & North-South

5.1.16 Economic Development UK: Regional & Global

5.1.17 End of Topic Test - The Changing Economic World

5.1.18 Exam-Style Questions - The Changing Economic World

5.1.19 Changing Economic World - Statistical Skills

6 The Challenge of Resource Management

6.1 Resource Management

6.1.1 Global Distribution of Resources

6.1.2 Food in the UK

6.1.3 Water in the UK 1

6.1.4 Water in the UK 2

6.1.5 Energy in the UK

6.1.6 Resource Management - Statistical Skills

6.2.1 Areas of Food Surplus & Food Deficit

6.2.2 Food Supply & Food Insecurity

6.2.3 Increasing Food Supply

6.2.4 Case Study: Thanet Earth

6.2.5 Creating a Sustainable Food Supply

6.2.6 Case Study: Agroforestry in Mali

6.2.7 End of Topic Test - Food

6.2.8 Exam-Style Questions - Food

6.2.9 Food - Statistical Skills

6.3.1 The Global Demand for Water

6.3.2 What Affects the Availability of Water?

6.3.3 Increasing Water Supplies

6.3.4 Case Study: Water Transfer in China

6.3.5 Sustainable Water Supply

6.3.6 Case Study: Kenya's Sand Dams

6.3.7 Case Study: Lesotho Highland Water Project

6.3.8 Case Study: Wakel River Basin Project

6.3.9 Exam-Style Questions - Water

6.3.10 Water - Statistical Skills

6.4.1 Global Demand for Energy

6.4.2 Factors Affecting Energy Supply

6.4.3 Increasing Energy Supply: Renewables

6.4.4 Increasing Energy Supply: Non-Renewables

6.4.5 Carbon Footprints & Energy Conservation

6.4.6 Case Study: Rice Husks in Bihar

6.4.7 Exam-Style Questions - Energy

6.4.8 Energy - Statistical Skills

Jump to other topics

Unlock your full potential with GoStudent tutoring

Affordable 1:1 tutoring from the comfort of your home

Tutors are matched to your specific learning needs

30+ school subjects covered

Responses to Earthquakes

Earthquake Case Study: Chile 2010

Earthquake shakes U.S. East Coast

An earthquake struck the East Coast of the United States on Friday morning, according to the U.S. Geological Survey, causing buildings to shake and rattling nerves from Maryland to Maine.

The USGS measured the quake as a 4.8 temblor with its epicenter near Lebanon, New Jersey. It struck a little before 10:30 a.m. ET. An aftershock of magnitude-4.0 hit right around 6 p.m. ET.

The morning earthquake was the strongest recorded in the Northeast in more than a decade, according to USGS records .

There were no immediate reports of major destruction or any fatalities. Local and regional officials from cities in the earthquake zone said inspections had been launched to ensure that buildings, bridges and other infrastructure were not damaged.

Follow here for live updates on the earthquake.

James Pittinger, mayor of Lebanon, New Jersey, called the earthquake “the craziest thing I’ve ever experienced.”  In an interview with MSNBC , he said he had not received reports of any significant damage so far, but added that the shaking caused his dog to run for cover and objects to fall off his shelves.

While a 4.8-magnitude temblor is not considered a major earthquake, even minor shaking can cause damage on the East Coast, which does not take similar precautions as other earthquake hot spots around the world.

New York Gov. Kathy Hochul said the quake was felt across the state.

“My team is assessing impacts and any damage that may have occurred, and we will update the public throughout the day,” she wrote on X .

New York City Mayor Eric Adams said in an afternoon news briefing that no major injuries or impacts to infrastructure were reported, and that people in the city should “go about their normal day.”

Ground stops were temporarily issued at Newark Liberty International Airport in New Jersey and John F. Kennedy International Airport in New York City, according to the Federal Aviation Administration's website. Flight disruptions at the Newark airport continued into the afternoon .

The Port Authority Transit Corp., which operates a rapid transit route between Pennsylvania and New Jersey, suspended service in the aftermath of the quake.

“Crews will inspect the integrity of the line out of an abundance of caution,” PATCO said in an update on X . “Once inspection is complete, service will resume. No timeframe. Updates to follow.”

New York’s Metropolitan Transportation Authority said that there had been no impact to its service but that teams will be inspecting train lines. New Jersey Transit alerted riders of 20-minute delays due to bridge inspections following the earthquake.

While earthquakes in the northeast U.S. are rare, Buffalo, New York, was struck by a 3.8-magnitude quake in February 2023 — the strongest recorded in the area in 40 years.

A 4.1-magnitude earthquake struck the tri-state area in 2017, centered near Little Creek, Delaware,  according to the U.S. Geological Survey . And before that, a 5.8-magnitude quake  shook central Virginia in 2011,  and was felt across much of the East Coast, forcing hundreds of thousands people to evacuate buildings in New York, Washington and other cities.

New Jersey Gov. Phil Murphy said in a post on X that the state has activated its emergency operations center and asked the public not to call 911 unless they are experiencing an emergency.

Frederik J. Simons, a professor of geosciences at Princeton University, told NBC News that the earthquake occurred on a shallow fault system in New Jersey and lasted about 35 seconds.

“The shallower or the closer it is, the more we feel it as humans,” he said.

The quake originated at a depth of less than 3 miles,  according to the USGS . 

Earthquakes on the East Coast can be felt at a great distance and can cause more pronounced shaking in comparison to those on the West Coast because rocks in the region are often older, harder and more dense.  

“These are competent rocks that transmit energy well,” Simons said.

The earthquake ruptured within a fault zone known as the Ramapo system, Simons said. It’s a zone in relatively ancient rock that contains old faults and cracks from ancient tectonic processes. These old faults slowly accumulate stress and occasionally something slips, Simons said.

“There are cracks in it and now and then a little motion accumulates, the stress keeps growing, at very slow rates,” he said. “It’s like an old house creaking and groaning.”

Simons said this was one of the largest earthquakes in New Jersey in recent history. The last notable one was a magnitude-3.1 temblor in Freehold Township in September 2020. 

“I’m on campus at Princeton University for the biggest one I’ve felt in a lifetime,” he said. “This shaking was violent, strong and long.”

Some videos captured the moment of the earthquake, including one from a coffee shop in New Jersey.

The East Coast quake struck two days after a powerful 7.4-magnitude temblor shook the island of Taiwan, killing at least 12 people and injuring more than 1,000 others. The two incidents are not thought to be related, said Dara Goldberg, a USGS geophysicist.

“We’re much too far of a distance for the stress on the fault of Taiwan to affect New York,” she said.

Denise Chow is a reporter for NBC News Science focused on general science and climate change.

Evan Bush is a science reporter for NBC News. He can be reached at [email protected].

Study suggests section of San Andreas Fault may be primed for major quake; Dr. Lucy Jones disagrees

Southern Californians constantly hear about the looming possibility of the next big earthquake . Now, some scientists are warning that areas around the San Andreas Fault could possibly see a strong earthquake as soon as this year.

According to a newly published study in the journal Frontiers in Earth Science , the Parkfield section of the fault in Monterey County experiences earthquakes regularly - about every 22 years.

The last time the remote area was hit with an earthquake was 2004, leading scientists to keep an eye on the seismic activity there. After analyzing data from the weeks before the 2004 earthquake, scientists have seen some similar patterns that could indicate another one is coming. Scientists are looking for clues like strain on rocks or seismic wave patterns, saying in the study that there are hints that Parkfield is entering the final phase of its quiet, dormant period.

But Dr. Lucy Jones disagrees with the study.

"Just because something is seen before one quake doesn't make that a warning. You know what has happened within 24 hours of every big quake we've ever seen? The sun has risen," Jones said Thursday in an online post.

While the Parkfield area could experience earthquakes around magnitude 6.0, its remote location means the chances of impacting people or property is low.

Related Topics

Top stories.

LASD deputy rushed to hospital after shooting in West Covina

• 12 minutes ago

La Cañada HS, other schools locked down after threat report

• 37 minutes ago

Trump trial: Here's what we know about 12 jurors, 6 alternates

Woman dies after being stabbed in neck on Metro train in Studio City

Pizza City Fest returns to LA Live with unlimited pizza

• 6 minutes ago

Everything we know about Marvel's new 'Deadpool & Wolverine' movie

3 young boys arrested for allegedly beating man in Long Beach

ADHD patients still face drug shortages due to increased demand

• 34 minutes ago

Machine learning-based detection of TEC signatures related to earthquakes and tsunamis: the 2015 Illapel case study

• Open access
• Published: 20 April 2024
• Volume 28 , article number  106 , ( 2024 )

Cite this article

You have full access to this open access article

• Federica Fuso   ORCID: orcid.org/0009-0008-2126-4028 1 , 2 ,
• Laura Crocetti   ORCID: orcid.org/0000-0003-2538-4111 2 ,
• Michela Ravanelli   ORCID: orcid.org/0000-0001-9788-7434 3 &
• Benedikt Soja   ORCID: orcid.org/0000-0002-7010-2147 2  

65 Accesses

Explore all metrics

Earthquakes and tsunamis can trigger acoustic and gravity waves that could reach the ionosphere, generating electron density disturbances, known as traveling ionospheric disturbances. These perturbations can be investigated as variations in ionospheric total electron content (TEC) estimated through global navigation satellite systems (GNSS) receivers. The VARION (Variometric Approach for Real-Time Ionosphere Observation) algorithm is a well-known real-time tool for estimating TEC variations. In this context, the high amount of data allows the exploration of a VARION-based machine learning classification approach for TEC perturbation detection. For this purpose, we analyzed the 2015 Illapel earthquake and tsunami for its strength and high impact. We use the VARION-generated observations (i.e., dsTEC/dt) provided by 115 GNSS stations as input features for the machine learning algorithms, namely, Random Forest and XGBoost. We manually label time frames of TEC perturbations as the target variable. We consider two elevation cut-off time series, namely, 15° and 25°, to which we apply the classifier. XGBoost with a 15° elevation cut-off dsTEC/dt time series reaches the best performance, achieving an F 1 score of 0.77, recall of 0.74, and precision of 0.80 on the test data. Furthermore, XGBoost presents an average difference between the labeled and predicted middle epochs of TEC perturbation of 75 s. Finally, the model could be seamlessly integrated into a real-time early warning system, due to its low computational time. This work demonstrates high-probability TEC signature detection by machine learning for earthquakes and tsunamis, that can be used to enhance tsunami early warning systems.

Avoid common mistakes on your manuscript.

Introduction

Natural hazards such as volcanic eruptions, earthquakes, and tsunamis can perturb the ionosphere (Astafyeva 2019 ; Huang et al. 2019 ; Calais and Minster 1995 ; Peltier and Hines 1976 ; Hargreaves 1992 ; Occhipinti 2015 ; Rolland et al. 2010 ; Meng et al. 2019 ; Artru et al. 2005 ; Chou et al. 2017 ; Zettergren et al. 2017 ). In detail, these events can generate acoustic and gravity waves (AGWs), that are amplified as atmosphere density decreases and can reach the ionosphere. These waves interact with the ionospheric plasma and cause ionospheric electron density disturbances, known as traveling ionospheric disturbances (TIDs; Galvan et al. 2012 ; Astafyeva 2019 ). Here, we mention acoustic gravity waves generated near the epicenter (AGWepi) and internal gravity waves (IGWtsu; Occhipinti 2015 ). AGWepi, related to the uplift at the source, reaches the ionosphere in around 8 min, whereas IGWtsu, linked to tsunami offshore propagation, takes about 45–60 min (Lognonné et al. 2006 ; Occhipinti et al. 2011 ).

These perturbations are detected through variations in ionospheric total electron content (TEC; Coster et al. 2013 ; Hofmann-Wellenhof et al. 2008 ), retrieved by global navigation satellite system (GNSS) measurements, and measured in TEC units (1 TECU = 10 16 electrons/m 2 ). Specifically, we refer to the slant TEC (sTEC) measured by dual-frequency GNSS receivers and encountered by the GNSS signal during its path through the ionosphere from the satellite to the receiver.

The VARION algorithm (Variometric Approach for Real-Time Ionosphere Observation) is an established tool to estimate sTEC variations from GNSS observations (Savastano et al. 2017 ; Ravanelli et al. 2021 ). It is based on a single time difference of geometry-free combinations, being suitable for real-time applications (Ravanelli et al. 2020 ).

TIDs detection considering sTEC time series has been conducted using traditional techniques reliant on human expertise, including the analysis of the ionospheric power index (Manta et al. 2020 ), the wavelet analysis threshold-based (Torrence and Compo 1998 ), and the 2D principal component analysis (Lin 2022 ). Despite the effectiveness of these traditional approaches, there is increasing recognition that artificial intelligence (AI), particularly machine learning, holds potential for advancing TIDs detection. Machine learning uses data-driven algorithms for autonomous decision-making, offering computational efficiency and data handling capacity (Kuglitsch et al. 2022 ; Crocetti et al. 2021 ). During the past years, the application of machine learning algorithms has become widespread in ionospheric research for different purposes: automatic detection methods to study TID signatures (Brissaud and Astafyeva 2022 ; Constantinou et al. 2023a , b ); forecasting TEC data (Cesaroni et al. 2020 ; Huang and Yuan 2014 ; Natras et al. 2022 ; Liu et al. 2020 ); nowcasting TEC data (Zhukov et al. 2018 ; Camporeale 2019 ; Łoś et al. 2020 ); improvement of regional and global TEC models (Zhukov et al. 2020 ); ionospheric scintillation parameter predictions (Atabati et al. 2021 ; McGranaghan et al. 2018 ; Linty et al. 2019 ); and the general analysis of TEC variations induced by earthquakes and tsunamis (Zhukov et al. 2020 ).

Within the framework of automatic detection methods to study TID signatures, the previous studies by Constantinou et al. ( 2023a , b ) considered computer vision and convolutional neural networks, while Brissaud and Astafyeva ( 2022 ) applied the Random Forest classifier, similar to our study, focusing on classifying ionospheric waveforms into TIDs and noise, picking TID arrival times, and associating arrivals across a satellite network in near real-time. In this context, our study aims to use machine learning algorithms for classifying ionospheric TEC variations caused by earthquakes and tsunamis using the large amount of GNSS time-series data (provided by every satellite-station link available). It aligns with the existing body of research while contributing novel insights into the classification of TIDs in large-scale GNSS data sets. Automatic TIDs detection, as planned by the NASA-JPL GUARDIAN system (Martire et al. 2023 ), underscores the importance of our research within the scientific community.

This paper investigates the 2015 Illapel earthquake and tsunami, a valid case study with a well-documented high magnitude and tsunami signature (Reddy et al. 2017 ; Ravanelli et al. 2021 ; Shrivastava et al. 2021 ).

The main objective is to examine if and how machine learning algorithms are suitable to find TEC time-series signatures related to earthquakes and tsunamis.

To reach our aim, we consider two machine learning classifiers, namely, Random Forest and XGBoost (Breiman 2001 ; Chen and Guestrin 2016 ), and apply them to the first temporal derivative of the sTEC (dsTEC/dt), representing VARION-core output. It depicts the rate of change of sTEC with respect to time, to which an elevation cut-off of 15° and 25° is implemented. Several experiments are carried out to determine the optimally performing model, such as data pre-processing, and feature-selection. The following Section provides an overview of the analyzed event, the study region, and the data used. The Methods Section describes the methodologies: the VARION algorithm and the machine learning techniques. Moreover, it presents the set-up of the machine learning classification, the explanations of the pre-processing features, and the assessment of the model performance. In the Results Section, the outcomes are presented, analyzed, and discussed. Here, we also validate the model by applying it to time series with no seismic-induced variations in sTEC. Specifically, we consider data with alterations in dsTEC/dt values not related to the earthquake, testing the algorithm's capability to exclusively detect variations directly linked to the earthquake, aligning with our primary objective. Finally, the last Section summarizes the outcomes derived from this study, identifies the most effective configuration for an efficient model, and offers a prospective outlook on potential enhancements yet to be explored.

Study context and data set overview

On September 16, 2015, at 19:54:33 Chile Standard Time (22:54:33 UTC), a devastating earthquake with a moment magnitude of Mw 8.3 occurred 46-km offshore of the Coquimbo region of Chile. The primary seismic event lasted between 3 and 5 min, followed by multiple aftershocks (Ravanelli et al. 2021 ). Both the NOAA Pacific Tsunami Warning Centre ( https://www.tsunami.gov/ ; https://earthquake.usgs.gov/earthquakes/eventpage/us20003k7a/executive ) and the Servicio Hidrográfico y Oceanográfico de la Armada (Chile’s National Tsunami Warning System) ( https://www.armada.cl/noticias-navales/shoa-difunde-oportuna-alerta-de-tsunami ; https://www.snamchile.cl/ ) issued tsunami threat messages and alarms, respectively. Within 10 min, tsunami waves measuring 4.5 m struck the Chilean shoreline between Chañaral (~ 26°S) and Constitución (~ 35°S), causing substantial impact ( https://www.tsunami.gov/events/PAAQ/2015/09/16/nuskyv/22/WEAK51/WEAK51.txt ). Pichidangui (~ 32°S) was reached by the first wave 13.70 min after the mainshock. The coastal area between Coquimbo (~ 30°S) and Valparaíso (~ 33°S) recorded the highest wave height (over 1.5 m), leading to significant flooding in Coquimbo (Shrivastava et al. 2021 ).

Data from 115 GNSS stations mostly located in Chile but also spread all over the South American continent (Fig.  1 ) are processed with VARION (Ravanelli et al. 2021 ). The GNSS receivers collect data at 10-, 15-, and 30-s rates. Following Reddy et al. 2017 research, we consider G12, G24, and G25 satellites, resulting in a data set composed of 345 observations.

Map showing the 115 GNSS receivers (red points) and the epicenter of the 2015 Illapel earthquake (yellow star)

In detail, our data set consists of 345 VARION-obtained sTEC variations over time, i.e., dsTEC/dt [TECU/s], which represent the primary and the real-time output of VARION and do not require any additional post-processing (Ravanelli et al. 2021 ). Specifically, the dsTEC/dt time series are representative as the earthquake has high magnitude and causes evident signatures in the ionosphere. However, while our data set provides insights into TID detection, it has limitations. Indeed, seismic and non-seismic external factors such as space weather, noise, satellite angles, geomagnetic field, and observation geometry influence the induced TEC observation (Meng et al. 2022 ; Bagiya et al. 2019 ). We analyze both GPS days, i.e., DOY 259 and 260 (the earthquake day and the day after), considering the time series of the whole DOY 259 and part of the time series of DOY 260, namely, until satellites descend above the elevation cut-off (Fig.  2 ). Moreover, the data set considers only 31 selected links satellite-stations following four criteria:

Data availability on both days (DOY 259 and 260)

Considerable (that can be visually identified) variations in the dsTEC/dt time series (related to the earthquake)

Only one gap during the day (due to satellite visibility)

Data availability from the first observation of the day (which is on the first 15 s of the day), for both DOY 259 and 260

Two of the 31 dsTEC/dt time series that constitute our data set. Here, the link composed of the G12 satellite and the PAZU station is shown on the top plot (a); whereas the bottom plot (b) presents a zoomed view of the time series of the link composed of the G25 satellite and the MRCG station, specifically focusing on the part of the day when the earthquake develops. This portion, occurring after the gap due to the satellite visibility, highlights the seismic-induced variations in the data

The frequency of the observations related to the selected 31 samples is 15 s.

Methodology

The VARION algorithm is applied to the observations of GPS DOY 259 and 260. VARION estimates real-time sTEC variations relying on stand-alone GNSS receivers and standard GNSS broadcast products. It is based on single time differences of a geometry-free combination of GNSS carrier-phase measurements (Savastano et al. 2017 ; Ravanelli et al. 2020 , 2021 ). A crucial tool in this study, VARION provides the time series for both train the machine learning algorithm and validate the results.

In this analysis, we use the elevation cut-off angles of 15° and 25° to mitigate the impact of observational noise, a prevalent challenge in ionospheric studies. These cut-off angles represent a strategic filter, effectively excluding observations derived by satellites with lower elevation angles where data tend to be noisier due to increased atmospheric interference. This aligns with common practices in the literature, where such elevation cut-offs are used to enhance the signal-to-noise ratio and improve the overall quality of the data set, ensuring a more robust and reliable foundation for our analysis (Occhipinti et al. 2011 ; Astafyeva 2019 ; Ravanelli et al. 2023 ).

To achieve our final aim of detecting the TIDs caused by earthquakes and tsunamis, we formulate a binary classification problem using supervised machine learning algorithms. Data are classified into two categories: 0 (if there are no sTEC variations related to the earthquake and tsunami in a 30-min window that runs through the time series) and 1 (if there are sTEC variations). The TIDs detection is based on established criteria derived from prior studies available in the literature. These include factors such as the arrival time of the perturbation in the ionosphere, its shape, absence of geomagnetic disturbances, and frequency content, as identified and validated in the previous research (Reddy et al. 2017 ; Ravanelli et al. 2021 ; Shrivastava et al. 2021 ; Sanchez et al. 2023 ).

We consider Random Forest (RF) and XGBoost (XGB) classifiers among the several available for classification tasks (Zhang et al. 2017 ), for their well-known good performances (Brissaud and Astafyeva 2022 ). Indeed, Crocetti et al. 2021 show that RF and XGB algorithms perform better than others (i.e., Linear Support Vector Classification, Perceptron, K-Nearest Neighbor, and more). Specifically, RF excels in handling high-dimensional data and mitigating overfitting through bootstrap aggregation, while XGBoost improves weak learners iteratively, enhancing accuracy and computational efficiency. Both algorithms, as ensemble learning methods, capture intricate relationships within the data and enhance predictive accuracy. They are suitable for our purposes, especially since we do not require deep learning methods (i.e., Convolutional Neural Networks—CNNs; Albawi et al. 2017 ) given our moderate data set size and the division into short chunks of the time series.

In detail, RF consists of decision trees, nonlinear models with multiple linear boundaries. Decision tree nodes use data-related questions linked to specific feature values, recursively dividing layers into child nodes. The iterative process creates a tree with a predefined depth. The algorithm selects bootstrapped samples and a random feature subset for model evaluation. The final prediction aggregates results from all decision trees (Breiman 2001 ). In contrast, XGBoost builds trees sequentially, each minimizing the error of the previous tree. It starts with a constant, iteratively trains trees on residuals, and combines them with the previous model to reduce error. Finally, strong learner results from combining all weak learners (Wang and Liu 2020 ; Chen and Guestrin 2016 ).

The classification is conducted for the selected 31 links (described in the Data set Section), split into 80% (25 links) and 20% (6 links) for training and testing. The dsTEC/dt time series of each link are divided into individual chunks, as described in the following Section. The model is trained based on the dsTEC/dt time series of the 25 training links, while tested considering the ones of the six testing links. Finally, the validation of the model and the evaluation of its performance are analyzed by considering 18 unseen dsTEC/dt time series, also related to the Illapel event. However, these show variations in dsTEC/dt not linked to the seismic event. This additional analysis allows us to assess the model's ability to distinguish variations in dsTEC/dt specifically linked to the earthquake and tsunami from those unrelated to the seismic event.

Feature matrix

We use the dsTEC/dt time series of 31 links satellite-stations as features for the machine learning algorithms. In detail, we exclude the gap due to the satellite visibility of each time series and consider the time frames (1) from the first 15 s of DOY 259 until the start of the gap of DOY 259 and (2) from the end of the gap of DOY 259 until the start of the gap of the DOY 260 of each time series.

The time series are split into k chunks to create the feature matrices. Each chunk is m  = 30 min long and shifted by n  = 1 min from the next one: The first chunk contains dsTEC/dt values from the first 15 s (the first observation for both the days) until m, while the second one from 75 s (15 s +  n ) until 75  s  +  m , and so on. The feature matrices have the dimension [ k , 120] as 120 is the number of 15 s in m . Finally, we obtain 31 feature matrices (one per link), where we combine 25 to create the feature matrix used for training (80%), and 6 to create the one used for testing (20%). The structure of the feature matrix is shown in Fig. 3 .

Structure of the feature matrix related to the time values for DOY 259 (both for training and for testing). For every time value, the matrix is filled with the corresponding dsTEC/dt value. The red cells are an illustrative and hypothetical example to show the dsTEC/dt values perturbed from the earthquake and tsunami, corresponding to the value “1” in the target vector

Target vector

The target vector denotes whether sTEC variations due to the tsunami occurred within their corresponding chunks, classifying them as “1” if the perturbation is present or “0” if not. As previous mentioned, this attribution is based on well-defined conditions validated in the literature, which demonstrate the link between sTEC perturbations and the seismic event. To perform this binary classification, the time frame of sTEC perturbation related to the event is manually labeled from the time series, i.e., the initial and finishing times of the sTEC perturbation for every link satellite-station. Therefore, the target vector is labeled as 1 whenever the 30-min chunks contain any point in time of the manually labeled time frame of the sTEC perturbation.

In this way, we create a target vector with the dimension [ k , 1] for every link (Fig. 3 ). As for the feature matrix, we use the same 80% of the links for training and 20% for testing.

Data pre-processing

The pre-processing of the data set consists of cleaning and preparation of the data, to improve the quality of the data set and ensure better performance of the model. In this study, we consider standardization and normalization, two methods used to scale the data set. In detail, standardization transforms the data to a specific range (e.g., 0 and 1 or − 1 and + 1); while normalization changes the data so that they resemble a normal distribution (Ali and Faraj 2014; Vieira et al. 2020 ).

Finally, we also include additional features in the feature matrix, namely, the value range, defined as the maximum value minus the minimum one before the normalization, and the variance of each chunk, both individually and together.

Model evaluation

To evaluate the performance of the model, we consider the confusion matrix, the difference between the labeled and predicted middle epochs of the perturbations time frames, the receiver operating characteristic (ROC) curve and the area under the curve (ROC-AUC).

In detail, the confusion matrix shows the number of false negatives (FNs), false positives (FPs), true negatives (TNs), and true positives (TPs) generated from the machine learning classification (Crocetti et al. 2021 ). In our case, TPs and TNs indicate correctly classified chunks with (or without, for TNs) earthquake- and tsunami-induced sTEC variations. Conversely, FNs identify chunks with undetected sTEC variations, while FPs designate chunks without sTEC variations, wrongly classified as containing them. Due to our aim of not overlooking any tsunamis, FNs are considered the most crucial errors in the analysis. However, to evaluate the best model, we consider the confusion matrix in terms of the well-known performance measures of precision, recall, and F 1 score (Ting 2017 ).

Moreover, we compare the labeled perturbation time frame with the one generated by the model, represented by the time frame containing all the TPs (assumed to be at the middle epoch of the 30-min time period of the chunk). To achieve this goal, we calculate the numerical difference between the middle epoch of both time frames and assess how many 15 s of difference there are between them (according to the time-series resolution).

Results and discussion

To reach the final aim of finding signatures in TEC time series related to earthquakes and tsunamis using machine learning algorithms, we performed several experiments considering two classifiers, i.e., Random Forest and XGBoost, and two elevation cut-off angles, i.e., 15° and 25° (see the following Section). We conducted hyperparameter tuning through grid search for each of the four combinations classifier elevation cut-off. We also assessed the influence of additional features (see the Section related to the impact of added features) and pre-processing techniques (refer to the Section concerning the pre-processing effects) on the model's performance. The model with the best performance in terms of F 1 score was then selected for each combination, resulting in four models. Finally, our final best model was selected in terms of F 1 score, difference between the middle epochs of the labeled and predicted perturbation time frames, ROC curve and ROC-AUC.

The best result is presented in the first part of this section, while the following subsections show how we came to this conclusion by comparing the best model with the others.

We trained our model on a machine with a 2.20-GHz i7-Intel Core processor, 32 GB of memory, and an Intel(R) UHD Graphics 630 graphics card. So, it can even be trained on a laptop without a proper GPU.

From this study, the best model is the XGBoost classifier that uses 15° elevation cut-off dsTEC/dt time series and includes the value range of each chunk as an additional feature. A hyperparameter tuning was performed to optimize the overall model performance. In the case of XGBoost, the number of boosting rounds (or the number of trees to be built in the ensemble), the maximum depth of trees, the number of columns to be randomly sampled for each tree, and the learning rate (the step size at each iteration while the model optimizes toward its objective) were analyzed. To determine the most suitable hyperparameters, we used a grid search, which systematically searches through a grid of a manually predefined set of hyperparameter combinations. In this study, the performance of all hyperparameter value combinations is evaluated based on a threefold cross-validation. This involves partitioning the data set into three subsets, with two-thirds used for direct model training and one-third reserved for validation. This process was repeated three times, ensuring each subset served as a validation set once. The average performance across these three runs for each hyperparameter combination was then compared to identify the set of hyperparameter values that reach the best performance. Table 1 shows the tested combinations and the best and default hyperparameter values for XGB-15°, our best model.

Regarding the testing samples, the best-performing model correctly classifies 183 of 247 (74.09%) samples of sTEC variations related to the earthquake and the tsunami (TP). Furthermore, 2975 of 3021 (98.49%) of the testing samples are correctly classified, containing no earthquake-induced sTEC variations (TN). Thus, 64 of 247 samples (25.91%) are wrongly classified as containing no seismic-induced sTEC variations (FN), while 46 of 3021 (1.51%) are the number of wrongly classified chunks as containing sTEC variations related to the event (FP), as shown in Table  2 . The model achieves an F 1 score of 0.77, recall of 0.74, and precision of 0.8; while the accuracy considering the training and the testing data is 0.98 and 0.97, respectively (Table  3 ). Figure  4 depicts these results for two links and the difference between the labeled and predicted middle epochs of the perturbation time frames.

Time series of two sample links: the one composed by the G24 satellite and PAZU station, used for training ( a ); and the one formed by the G24 satellite and UDAT station, used for testing ( b ). In both, the performance of the best model is presented. The plots show the position of the FNs and TPs in the time series together with the labeled and predicted middle epochs of the perturbation time frame

The average numerical difference between the epoch of the TID is 70.8 s, considering the 25 links used to train the model, and 75 s for the 6 links used for testing (Table  4 ). As the time series are 15-s step discontinuous, achieving a 5-step average differences between the two epochs both in training and in testing is relevant. This highlights the algorithm's precise temporal detection of the AGWepi-related perturbations, underlining its critical role in timely TIDs identification for effective early warning systems. Furthermore, having similar results for training and testing data shows the model’s generalization ability, avoiding overfitting problems. In this context, the model's capability to generalize is ensured by incorporating an additional feature (i.e., value range), which prevents overfitting and captures essential information. The generalization is also achieved by the threefold cross-validation. Moreover, the sensitivity to noisy observations (i.e., selection of elevation cut-off angles) improves the model's focus on relevant information.

Figure  5 illustrates the ROC curves and AUC values for both algorithms and cut-off angles. RF reaches better ROC curves and higher AUC values; however, we opted for XGB-15° considering its greater performance in the F 1 score and in the difference between the labeled and predicted middle epochs. In particular, a good performance in the last metric is essential for good results in real-time applications, aligning with our aim scope.

ROC curves depicting the performance of the different machine learning algorithms. The blue dashed line represents the performance of a random classifier. The legend includes the corresponding AUC values

Overall, the algorithm demonstrates good computational efficiency, detecting in about 2–3 min, with hyperparameters tuning as the most critical step. An effective computational performance is crucial for real-time operational requirements, essential for effective early warning systems. However, our analysis uses data sampled at larger intervals than the 1-s real-time data rate, which can increase the computational time to 30–40 min.

Finally, the algorithm holds potentials as a highly viable tool, considering its ability to operate in real-time using only sTEC time series. However, several crucial aspects must be addressed for practical implications. In fact, successful real-world implementation requires access to real-time data from networks in high-risk areas prone to tsunamis. Furthermore, incorporating buffer time into the analysis process allows for a thorough examination of the data and its quality in real-time (i.e., data integrity, and bias and outlier identification). Establishing the necessary infrastructure is also crucial. This includes dedicated servers that can perform complex computations, efficiently run AI algorithms, and transfer data to cloud platforms. Finally, the use of servers and cloud storage is essential for storing data collected in real-time, which can be used for training new models and conducting retrospective analyses.

Comparison of different elevation cut-off angles

We tested two different elevation cut-off angles, 15° and 25°, as input features for our model. The results for the two classifiers, and the elevation cut-off angles, are shown in Table  3 , where the first column shows the results of our best model.

For both classifiers, the performance related to the different cut-off angles is similar in terms of accuracy and precision. However, the recall is a bit higher when using an elevation cut-off of 25° (Table  3 ), whereas the F 1 score is higher for an elevation cut-off of 25° for RF and similar for XGBoost. Conversely, when comparing the results of the two classifiers, we note that while the precision is higher when using RF, the F 1 score and recall are significantly higher for XGBoost (Table  3 ).

Furthermore, Table  4 shows that for both classifiers, the differences between the labeled and predicted middle epochs for the links used for training and testing are smaller when using an elevation cut-off of 15° instead of 25°. Even though the average difference between the middle epochs for the training links is higher for XGBoost than for RF, the one for the testing links is smaller (Table  4 ). Moreover, RF presents a high dissimilarity between the average difference of the middle epochs considering the training and testing links, so the model proves not to be suitable to generalize well over unseen data sets.

Thus, we conclude that the XGBoost algorithm using dsTEC/dt time series with an elevation cut-off of 15° is the best model since it has high accuracy, precision, recall, and F 1 score and performs best when investigating the average differences between labeled and predicted middle epochs.

Impact of additional features

We evaluate the impact of adding additional features, namely, the value range and variance of the chunk, to the feature matrix used in the machine learning algorithm. The choice of incorporating those features in our model is motivated by their capacity to capture different aspects of the dsTEC/dt time series. The value range represents the amplitude of variations within each chunk, providing insights into the overall perturbation magnitude. Meanwhile, variance quantifies the internal dynamics and temporal variability, describing the temporal characteristics of the ionospheric perturbations. For all combinations (XGB-15°, XGB-25°, RF-15°, and RF-25°), this addition improves the results in terms of precision, recall, and F 1 score, as shown in Table  5 . Indeed, they enable the model to better distinguish seismic-induced variations in dsTEC/dt time series, thus improving predictive accuracy through a deeper understanding of ionospheric perturbations. In particular, in XGB-15°, RF-25°, and XGB-25°, including value range or variance leads to similar results, having thus a similar impact on the model. On the other hand, in RF-15°, including the value range has a greater impact than including variance as it doubles the values of the metrics obtained with the raw data. This sensitivity of the algorithm to the range values aligns with the nature of Random Forest, which benefits from capturing a broad spectrum of information for robust decision-making, and the value range seems to capture the data variability more effectively than the variance. Finally, the study of the feature importance reveals that, for our best model (XGB-15°), the value range has the greatest impact. Thus, this shows that the value range significantly contributes to reduce errors in the tree ensemble.

Including both value range and variance in the feature matrix improves the raw case results. However, in XGB-15° and XGB-25°, the outcomes are worse than the ones related to the addition of value range or variance individually. Only in the case of RF-25°, adding both value range and variance presents better results than including them individually. This is likely because we do not have that much training data, and having more data can help the model better learn the patterns and relationship within the features. In this case, the joint inclusion of the features might introduce redundancies or interactions that the model struggles to effectively leverage.

Impact of the pre-processing

In this section, we evaluate the impact of standardization and normalization of the feature matrix chunks, done separately for each chunk, on the performances of the different models (Table  6 ), which commonly should be small on tree-based algorithms (García et al. 2015 ; Dougherty 2013 ). We highlight that only standardization applied to XGB-15° outperforms the raw case. This improvement can be attributed to the sensitivity of the XGBoost algorithm to the scale of input features. Standardization helps align the features to a common scale, facilitating more effective convergence during the boosting process. Furthermore, with 15° elevation cut-off, we observe more noisy data compared to the ones at 25°. This leads to more pronounced fluctuations and a high amount of data, underscoring the greater impact of standardization.

In contrast, in RF-15°, RF-25°, and XGB-25°, both standardization and normalization have a negative impact on the model, worsening the raw case results in terms of F 1 score and recall. This outcome shows that RF and XGB are in this case not highly sensitive to the scale of features.

Validation of the model

To validate our results, we apply the best model to unseen dsTEC/dt time series related to the same event but generated from different satellite-station links. Specifically, we select the time series with variations in dsTEC/dt values occurring before the seismic event, ensuring that these perturbations are not caused by the earthquake. To select them, we apply some variations to the first two selection criteria used before:

Data availability only on DOY 259, excluding DOY 260 (as those time series are not registering at the end of DOY 259, when the earthquake occurred, so there would be a gap between DOY 259 and 260)

Considerable (that can be visually identified) variations in the dsTEC/dt time series, but not related to the earthquake (caused by noise or ionospheric background)

In this way, 18 dsTEC/dt time series are selected, enabling us to evaluate the performance of the model. Indeed, the expectations for this application are to predominantly capture all or most TNs, and minimize the occurrence of FPs. This is particularly crucial, as we are specifically considering time-series data with variations in sTEC that is not induced by the earthquake, ensuring a focus on our primary objective. The confusion matrix shows good results: 96.89% TNs and 3.11% FPs (in numbers: 5119 of TNs and 164 of FPs). These successful results are also confirmed by an accuracy of 0.97 (Fig.  6 ).

Time series of G14 satellite and LSCH station showing the results obtained applying the model to one of the unseen dsTEC/dt time series, where the earthquake does not occur. The plot shows the FPs and TNs in the time series ( a ), with a zoom on the time frame where the FPs are detected ( b )

Conclusions

This study successfully used two machine learning algorithms, Random Forest and XGBoost, to detect TEC variations induced by the significant 2015 Illapel earthquake and tsunami, known for its substantial ionospheric TEC signatures. We approached the problem as a supervised binary classification task and used the VARION-generated dsTEC/dt time series as the input for our model. We followed specific criteria to select 31 links satellite-station. Then, we split the corresponding dsTEC/dt time series into individual 30-min chunks to create the feature matrix used in the machine learning algorithms.

We considered different classifiers, elevation cut-off angles, additional features, and pre-processing techniques. The best result, based on F 1 score, average difference between labeled and predicted middle epochs of the perturbation time frames, and ROC curves, was with XGBoost classifier applied to 15° elevation cut-off time dsTEC/dt time series. The best performance was achieved by including the value range of the chunks as an additional feature and by tuning the hyperparameter using grid search.

Applying our final model to unseen test data, we obtained an overall performance of an F 1 score of 0.77, a recall of 0.74, and a precision of 0.80. Indeed, focusing on the testing samples, 183 of 247 (74.09%) were correctly classified to contain sTEC variations related to the earthquake and the tsunami (TP). Furthermore, 2975 of 3021 (98.49%) of the testing samples were correctly classified, containing no sTEC variations caused by an earthquake (TN). Thus, 64 samples (25.91%) were wrongly classified as containing no sTEC variations caused by the event (FN), and 46 (1.51%) were the number of wrongly classified chunks as containing sTEC variations related to the earthquake and tsunami (FP).

Our model, with a smaller data set and single event, achieves competitive TP and TN rates when compared to the study by Brissaud and Astafyeva (Brissaud) with a substantial data set of 12 earthquake events, underscoring its potential utility in operational early warning systems.

This model demonstrated a 75-s average difference in predicting perturbation time frames for testing links, equivalent to an average difference of five steps considering the 15-s steps time series. This highlights the algorithm's potential for early detection of ionospheric perturbations caused by earthquakes and tsunamis, aiding in early warnings purposes.

The model’s versatility allows application in an operational real-time setting using real-time GNSS data, as it only needs the VARION-generated real-time sTEC time series (dsTEC/dt). In that case, buffer time conversion of the sTEC time series, tools for store the processed data, powerful servers, and adequate computational resources need to be considered.

The model takes a few minutes to detect TIDs, presenting a very low percentage of FPs (1.51%) and showing a high computational efficiency, crucial for effective early warning systems. However, real-time situations may increase the computational time due to higher data frequency (1 s) and bigger data set size, emphasizing the need for continuous improvement.

Finally, we validated our results applying the best model to dsTEC/dt time series related to the same event but generated from different satellite-station links. In detail, those time series had variations in dsTEC/dt values occurring before the seismic event, thus not caused by the earthquake. We obtained promising results: 96.89% TNs, 3.11% FPs, and 97% accuracy. However, we know that validation process also involves applying the model to similar events. We acknowledge the importance of external validation on entirely different data sets to assess the model's performance in different scenarios, which will be discussed in the future works.

In conclusion, we have demonstrated a powerful tool for timely and accurate identification of ionospheric perturbations linked to seismic events. Our study not only provides a valuable contribution to the field of ionospheric research but also sets the stage for the integration of advanced machine learning techniques into operational early warning systems, improving our ability to respond proactively to seismic events and associated hazards.

However, recognizing the current model's limitations, particularly referring to the low frequency and low copiousness of the data, future studies should fine-tune the algorithm for real-time, high-resolution data, widening the analysis to different kinds of TIDs and data sets. Furthermore, the model should be improved by incorporating additional features and optimizing the computational efficiency considering parallel processing, together with the analysis of its generalization capabilities. This will enhance the model's robustness and effectiveness in real-time application. Moreover, a database that stores the events (such as earthquakes, tsunamis, and volcanic eruptions) with their characteristic features (time frame, waveform, frequency content, and period) should be established. In this way, it will be possible to collect several events that can be adopted within different algorithms, reaching a way of continuous learning. Those outcomes could then be used as tools that enable early warning systems to combine data derived from the ionosphere with other information to achieve integrated systems that work synergically, enhancing the overall effectiveness of disaster prediction and mitigation strategies.

Data availability

The GNSS data set used in this study is kindly provided by the Centro Sismológico Nacional, Universidad de Chile (CSN), the Low-Latitude Ionospheric Sensor Network (LISN), the Red Argentina de Monitoreo Satelital Continuo (RAMSAC) of Instituto Geográfico Nacional de Argentina (IGN), the Rede Brasileira de Monitoramento Contínuo dos sistemas GNSS (RBMC), the NASA Global GNSS Network (GGN), the Deutsches Geodätisches Forschungsinstitut (DGFI-TUM), the Geodesy & Geodynamics Group of The Ohio State University (G2/OSU), the Universidad de Concepción, the Jicamarca Radio Observatory, the Institut national de l’information géographique et forestiére (IGN-France), the Centre National d’Études Spatiales (CNES), the European Space Operations Centre (ESA/ESOC), the University of Miami, the Instituto Geofísico de Ecuador (EPN), the Servicio Geográfico Militar (IGM), and the Universidad Nacional del Sur (UNS).

Albawi S, Mohammed TA, Al-Zawi S (2017) Understanding of a convolutional neural network. In: International conference on engineering and technology (ICET), Antalya, Turkey, pp 1–6. https://doi.org/10.1109/ICEngTechnol.2017.8308186

Ali PJM, Faraj RH (2014) Data normalization and standardization: a technical report. Mach Learn Tech Rep 1:1–6. https://doi.org/10.13140/RG.2.2.28948.04489

Article   Google Scholar  

Artru J, Ducic V, Kanamori H, Lognonné P, Murakami M (2005) Ionospheric detection of gravity waves induced by tsunamis. Geophys J Int 160(3):840–848. https://doi.org/10.1111/j.1365-246X.2005.02552.x

Astafyeva E (2019) Ionospheric detection of natural hazards. Rev Geophys 57(4):1265–1288. https://doi.org/10.1029/2019RG000668

Atabati A, Alizadeh MM, Schuh H, Tsai L (2021) Ionospheric scintillation prediction on S4 and ROTI parameters using artificial neural network and genetic algorithm. Remote Sens 13:2092. https://doi.org/10.3390/rs13112092

Bagiya MS, Sunil AS, Rolland L, Nayak S, Ponraj M, Thomas D, Ramesh DS (2019) Mapping the impact of non-tectonic forcing mechanisms on GNSS measured coseismic ionospheric perturbations. Sci Rep 9:18640. https://doi.org/10.1038/s41598-019-54354-0

Article   CAS   Google Scholar  

Breiman L (2001) Random forests. Mach Learn 45:5–32. https://doi.org/10.1023/A:1010933404324

Brissaud Q, Astafyeva E (2022) Near-real-time detection of co-seismic ionospheric disturbances using machine learning. Geophys J Int 230(3):2117–2130. https://doi.org/10.1093/gji/ggac167

Calais E, Minster JB (1995) GPS detection of ionospheric perturbations following the January 17, 1994, Northridge Earthquake. Geophys Res Lett 22(9):1045–1048. https://doi.org/10.1029/95GL00168

Camporeale E (2019) The challenge of machine learning in space weather: nowcasting and forecasting. Space Weather 17:1166–1207. https://doi.org/10.1029/2018SW002061

Cesaroni C, Spogli L, Aragon-Angel A, Fiocca M, Dear V, De Franceschi G, Romano V (2020) Neural network based model for global total electron content forecasting. J Space Weather Space Clim 10:11. https://doi.org/10.1051/swsc/2020013

Chen T, Guestrin C (2016) XGBoost: a scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining (KDD '16). Association for Computing Machinery, pp 785–794. https://doi.org/10.1145/2939672.2939785

Chou MY, Lin CCH, Yue J, Tsai HF, Sun YY, Liu JY, Chen CH (2017) Concentric traveling ionosphere disturbances triggered by Super Typhoon Meranti. Geophys Res Lett 44:1219–1226. https://doi.org/10.1002/2016GL072205

Constantinou V, Ravanelli M, Liu H, Bortnik J (2023a) A Deep Learning Approach for Detection of Internal Gravity Waves in Earth’s Ionosphere, IGARSS 2023 - 2023 IEEE International Geoscience and Remote Sensing Symposium, Pasadena, CA, USA, pp. 1178–1181. https://doi.org/10.1109/IGARSS52108.2023.10282501

Constantinou V, Ravanelli M, Liu H, Bortnik J (2023b) Deep learning driven detection of tsunami related internal GravityWaves: a path towards open-ocean natural hazards detection. https://doi.org/10.48550/arXiv.2308.04611

Coster A, Williams J, Weatherwax A, Rideout W, Herne D (2013) Accuracy of GPS total electron content: GPS receiver bias temperature dependence. Radio Sci 48:190–196. https://doi.org/10.1002/rds.20011

Crocetti L, Schartner M, Soja B (2021) Discontinuity detection in GNSS station coordinate time series using machine learning. Remote Sens 13(19):3906. https://doi.org/10.3390/rs13193906

Dougherty G (2013) Feature extraction and selection. In: Pattern recognition and classification: an introduction. Springer, New York, pp 123–141. https://doi.org/10.1007/978-1-4614-5323-9_7

Galvan DA, Komjathy A, Hickey MP, Stephens P, Snively J, Tony Song Y, Butala MD, Mannucci AJ (2012) Ionospheric signatures of Tohoku-Oki tsunami of march 11, 2011: Model comparisons near the epicenter. Radio Sci. https://doi.org/10.1029/2012RS005023

García S, Luengo J, Herrera F (2015) Introduction. In: Data preprocessing in data mining. Intelligent system reference library, vol 72. Springer, Cham, pp 1–17. https://doi.org/10.1007/978-3-319-10247-4_1

Hargreaves JK (1992) The solar-terrestrial environment: an introduction to geospace—the science of the terrestrial upper atmosphere, ionosphere, and magnetosphere. Cambridge University Press, Cambridge. https://doi.org/10.1017/CBO9780511628924

Book   Google Scholar  

Hofmann-Wellenhof B, Lichtenegger HI, Wasle E (2008) GNSS—global navigation satellite systems: GPS, GLONASS, Galileo, and more. Springer, Vienna. https://doi.org/10.1007/978-3-211-73017-1

Huang Z, Yuan H (2014) Ionospheric single-station TEC short term forecast using RBF neural network. Radio Sci 49(4):283–292. https://doi.org/10.1002/2013RS005247

Huang CY, Helmboldt JF, Park J, Pedersen TR, Willemann R (2019) Ionospheric detection of explosive events. Rev Geophys 57:78–105. https://doi.org/10.1029/2017RG000594

Jin S, Occhipinti G, Jin R (2015) GNSS ionospheric seismology: recent observation evidences and characteristics. Earth Sci Rev 147:54–64. https://doi.org/10.1016/j.earscirev.2015.05.003

Kuglitsch M, Pelivan I, Ceola S, Menon M, Xoplaki E (2022) Facilitating adoption of AI in natural disaster management through collaboration. Nat Commun 13:1579. https://doi.org/10.1038/s41467-022-29285-6

Lin JW (2022) Generalized two-dimensional principal component analysis and two artificial neural network models to detect traveling ionospheric disturbances. Nat Hazards 111:1245–1270. https://doi.org/10.1007/s11069-021-05093-x

Linty N, Farasin A, Favenza A, Dovis F (2019) Detection of GNSS ionospheric scintillations based on machine learning decision tree. IEEE Trans Aerosp Electron Syst 55(1):303–317. https://doi.org/10.1109/TAES.2018.2850385

Liu L, Zou S, Yao Y, Wang Z (2020) Forecasting global ionospheric TEC using deep learning approach. Space Weather 18:e2020SW002501. https://doi.org/10.1029/2020SW002501

Lognonné P, Artru J, Garcia R, Crespon F, Ducic V, Jeansou E, Occhipinti G, Helbert J, Moreaux G, Godet P-E (2006) Ground-based GPS imaging of ionospheric post-seismic signal. Planet Space Sci 54(5):528–540. https://doi.org/10.1016/j.pss.2005.10.021

Łoś M, Smolak K, Guerova G, Rohm W (2020) GNSS-based machine learning storm nowcasting. Remote Sens 12(16):2536. https://doi.org/10.3390/rs12162536

Manta F, Occhipinti G, Feng L, Hill EM (2020) Rapid identification of tsunamigenic earthquakes using GNSS ionospheric sounding. Sci Rep 10:11054. https://doi.org/10.1038/s41598-020-68097-w

Martire L, Krishnamoorthy S, Vergados P, Romans LJ, Szilágyi B, Meng X, Anderson JL, Komjáthy A, Bar-Sever YE (2023) The GUARDIAN system-a GNSS upper atmospheric real-time disaster information and alert network. GPS Solut 27(1):32. https://doi.org/10.1007/s10291-022-01365-6

McGranaghan RM, Mannucci AJ, Wilson BD, Mattmann CA, Chadwick R (2018) New capabilities for prediction of high-latitude ionospheric scintillation: a novel approach with machine learning. Space Weather 16:1817–1846. https://doi.org/10.1029/2018SW002018

Meng X, Vergados P, Komjathy A, Verkhoglyadova O (2019) Upper atmospheric responses to surface disturbances: an observational perspective. Radio Sci 54:1076–1098. https://doi.org/10.1029/2019RS006858

Meng X, Ravanelli M, Komjathy A, Verkhoglyadova OP (2022) On the north-south asymmetry of co-seismic ionospheric disturbances. Geophys Res Lett 49:e2022GL098090. https://doi.org/10.1029/2022GL098090

Natras R, Soja B, Schmidt M (2022) Ensemble machine learning of random forest, AdaBoost and XGBoost for vertical total electron content forecasting. Remote Sens 14:3547. https://doi.org/10.3390/rs14153547

Occhipinti G (2015) The seismology of the planet Mongo. In: Morra G, Yuen DA, King SD, Lee SM, Stein S (eds) Subduction dynamics. American Geophysical Union (AGU), Washington, pp 169–182. https://doi.org/10.1002/9781118888865.ch9

Chapter   Google Scholar  

Occhipinti G, Coïsson P, Makela JJ, Allgeyer S, Kherani A, Hebert H, Logonné P (2011) Three-dimensional numerical modeling of tsunami-related internal gravity waves in the Hawaiian atmosphere. Earth Planet Space 63:847–851. https://doi.org/10.5047/eps.2011.06.051

Occhipinti G, Rolland L, Lognonne P, Watada S (2013) From sumatra 2004 to Tohoku-Oki 2011: the systematic GPS detection of the ionospheric signature induced by tsunamigenic earthquakes. J Geophys Res Space Physics 118(6):3626–3636. https://doi.org/10.1002/jgra.50322

Peltier WR, Hines CO (1976) On the possible detection of tsunamis by a monitoring of the ionosphere. J Geophys Res 81(12):1995–2000. https://doi.org/10.1029/JC081i012p01995

Ravanelli M, Occhipinti G, Savastano G, Komjathy A, Shume EB, Crespi M (2021) GNSS total variometric approach: first demonstration of a tool for real-time tsunami genesis estimation. Sci Rep 11(1):1–12. https://doi.org/10.1038/s41598-021-82532-6

Ravanelli M, Astafyeva E, Munaibari E, Rolland L, Mikesell TD (2023) Ocean-ionosphere disturbances due to the 15 January 2022 Hunga-Tonga Hunga-Ha’apai eruption. Geophys Res Lett. https://doi.org/10.1029/2022GL101465

Ravanelli M, Crespi M, Foster J (2020) Tids detection from ship-based GNSS receiver: first test on 2010 Maule tsunami. In: IGARSS 2020—2020 IEEE international geoscience and remote sensing symposium. Waikoloa, HI, USA, pp 6846–6849. https://doi.org/10.1109/IGARSS39084.2020.9324549

Ravanelli M (2021) An innovative approach for real-time GNSS ionosphere seismology: assessment, potentialities, applications and issues. Ph.D. thesis, Sapienza University

Reddy CD, Shrivastava MN, Seemala GK, González G, Baez JC (2017) Ionospheric plasma response to Mw 8.3 Chile Illapel Earthquake on September 16, 2015. In: Braitenberg C, Rabinovich A (eds) The Chile-2015 (Illapel) Earthquake and Tsunami. Pageoph Topical Volumes. Birkhäuser, Cham, pp 145–155. https://doi.org/10.1007/978-3-319-57822-4_12

Rolland L, Occhipinti G, Lognonné P, Loevenbruck A (2010) Ionospheric gravity waves detected offshore Hawaii after tsunamis. Geophys Res Lett. https://doi.org/10.1029/2010GL044479

Sanchez SA, Kherani EA, Astafyeva E, de Paula ER (2023) Rapid detection of co-seismic ionospheric disturbances associated with the 2015 Illapel, the 2014 Iquique and the 2011 Sanriku-Oki Earthquakes. J Geophys Res (space Phys). https://doi.org/10.1029/2022JA031231

Savastano G, Komjathy A, Verkhoglyadova O, Mazzoni A, Crespi M, Wei Y, Mannucci AJ (2017) Real-time detection of tsunami ionospheric disturbances with a stand-alone GNSS receiver: a preliminary feasibility demonstration. Sci Rep 7:46607. https://doi.org/10.1038/srep46607

Shrivastava MN, Maurya AK, Gonzalez G, Sunil PS, Gonzalez J, Salazar P, Aranguiz R (2021) Tsunami detection by GPS-derived ionospheric total electron content. Sci Rep 11:12978. https://doi.org/10.1038/s41598-021-92479-3

Ting KM (2017) Confusion matrix. In: Sammut C, Webb GI (eds) Encyclopedia of machine learning and data mining. Springer, Boston. https://doi.org/10.1007/978-1-4899-7687-1_50

Torrence C, Compo GP (1998) A practical guide to wavelet analysis. Bull Am Meteorol Soc 79(1):61–78. https://doi.org/10.1175/1520-0477(1998)079%3c0061:APGTWA%3e2.0.CO;2

Vieira S, Lopez Pinaya WH, Mechelli A (2020) Chapter 2—main concepts in machine learning. In: Mechelli A, Vieira S (eds) machine learning. Academic Press/Elsevier, London, pp 21–44. https://doi.org/10.1016/B978-0-12-815739-8.00002-X

Wang XW, Liu YY (2020) Comparative study of classifiers for human microbiome data. Med Microecol 4:100013. https://doi.org/10.1016/j.medmic.2020.100013

Zettergren MD, Snively JB, Komjathy A, Verkhoglyadova OP (2017) Nonlinear ionospheric responses to large-amplitude infrasonic-acoustic waves generated by undersea earthquakes. J Geophys Res Space Physics 122:2272–2291. https://doi.org/10.1002/2016JA023159

Zhang C, Liu C, Zhang X, Almpanidis G (2017) An up-to-date comparison of state-of-the-art classification algorithms. Expert Syst Appl 82:128–150. https://doi.org/10.1016/j.eswa.2017.04.003

Zhukov AV, Sidorov DN, Mylnikova A, Yasyukevich YV (2018) Machine learning methodology for ionosphere total electron content nowcasting. Int J Artif Intell 16:144–157. https://doi.org/10.13140/RG.2.2.19349.83685

Zhukov AV, Yasyukevich YV, Bykov AE (2020) GIMLI: Global ionospheric total electron content model based on machine learning. GPS Solut 25(1):19. https://doi.org/10.1007/s10291-020-01055-1

Download references

Acknowledgements

F. F. was supported by the Zegna Foundation Grant and by Doctoral Program fellowship within the Data Science Course of Sapienza University of Rome founded by INGV. M.R. was supported by the AXA Research Fund and the Intergovernmental Oceanographic Commission of UNESCO (IOC-UNESCO) on Coastal Livelihood grant (B83C23002490007). L.C. and B.S. were supported by the ETH Zurich.

Open access funding provided by Università degli Studi di Roma La Sapienza within the CRUI-CARE Agreement.

Author information

Authors and affiliations.

Department of Computer, Control and Management Engineering (DIAG), Sapienza University of Rome, Via Ariosto, 25, 00185, Rome, Italy

Federica Fuso

Institute of Geodesy and Photogrammetry, ETH Zurich, Robert-Gnehm-Weg 15, 8093, Zurich, Switzerland

Federica Fuso, Laura Crocetti & Benedikt Soja

Department of Civil, Constructional and Environmental Engineering (DICEA), Sapienza University of Rome, 00184, Rome, Italy

Michela Ravanelli

You can also search for this author in PubMed   Google Scholar

Contributions

F.F. wrote the main manuscript, F.F. and L.C. carried out the experiments, and M.R. and B.S. conceived the experiments. All authors reviewed the manuscript.

Corresponding author

Correspondence to Federica Fuso .

Ethics declarations

Conflict of interest.

The authors declare no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Fuso, F., Crocetti, L., Ravanelli, M. et al. Machine learning-based detection of TEC signatures related to earthquakes and tsunamis: the 2015 Illapel case study. GPS Solut 28 , 106 (2024). https://doi.org/10.1007/s10291-024-01649-z

Download citation

Received : 26 October 2023

Accepted : 26 March 2024

Published : 20 April 2024

DOI : https://doi.org/10.1007/s10291-024-01649-z

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• GNSS Ionospheric Seismology
• VARION algorithm
• Machine learning
• Tsunami early warning systems

Advertisement

• Find a journal
• Publish with us
• Track your research

• 0 Shopping Cart

Nepal Earthquake 2015

A case study of an earthquake in a low income country (LIC).

Nepal, one of the poorest countries in the world, is a low-income country. Nepal is located between China and India in Asia along the Himalayan Mountains.

A map to show the location of Nepal in Asia

What caused the Nepal Earthquake?

The earthquake occurred on a  collision plate boundary between the Indian and Eurasian plates.

What were the impacts of the Nepal earthquake?

Infrastructure.

• Centuries-old buildings were destroyed at UNESCO World Heritage Sites in the Kathmandu Valley, including some at the Changu Narayan Temple and the Dharahara Tower.
• Thousands of houses were destroyed across many districts of the country.

Social and economic

• Eight thousand six hundred thirty-two dead and 19,009 injured.
• It was the worst earthquake in Nepal in more than 80 years.
• People chose to sleep outside in cold temperatures due to the risk of aftershocks causing damaged buildings to collapse.
• Hundreds of thousands of people were made homeless, with entire villages flattened.
• Harvests were reduced or lost that season.
• Economic losses were estimated to be between nine per cent to 50 per cent of GDP by The United States Geological Survey (USGS).
• Tourism is a significant source of revenue in Nepal, and the earthquake led to a sharp drop in the number of visitors.
• An avalanche killed at least 17 people at the Mount Everest Base Camp.
• Many landslides occurred along steep valleys. For example, 250 people were killed when the village of Ghodatabela was covered in material.

What were the primary effects of the 2015 earthquake in Nepal?

The primary effects of the 2015 earthquake in Nepal include:

• Nine thousand people died, and 19,000 people were injured – over 8 million people were affected.
• Three million people were made homeless.
• Electricity and water supplies, along with communications, were affected.
• 1.4 million people needed support with access to water, food and shelter in the days and weeks after the earthquake
• Seven thousand schools were destroyed.
• Hospitals were overwhelmed.
• As aid arrived, the international airport became congested.
• 50% of shops were destroyed, affecting supplies of food and people’s livelihoods.
• The cost of the earthquake was estimated to be US$5 billion.

What were the secondary effects of the 2015 earthquake in Nepal?

The secondary effects of the 2015 earthquake in Nepal include:

• Avalanches and landslides were triggered by the quake, blocking rocks and hampering the relief effort.
• At least nineteen people lost their lives on Mount Everest due to avalanches.
• Two hundred fifty people were missing in the Langtang region due to an avalanche.
• The Kali Gandaki River was blocked by a landslide leading many people to be evacuated due to the increased risk of flooding.
• Tourism employment and income declined.
• Rice seed ruined, causing food shortage and income loss.

What were the immediate responses to the Nepal earthquake?

• India and China provided over $1 billion of international aid .
• Over 100 search and rescue responders, medics and disaster and rescue experts were provided by The UK, along with three Chinook helicopters for use by the Nepali government.
• The GIS tool “Crisis mapping” was used to coordinate the response.
• Aid workers from charities such as the Red Cross came to help.
• Temporary housing was provided, including a ‘Tent city’ in Kathmandu.
• Search and rescue teams, and water and medical support arrived quickly from China, the UK and India.
• Half a million tents were provided to shelter the homeless.
• Helicopters rescued people caught in avalanches on Mount Everest and delivered aid to villages cut off by landslides.
• Field hospitals were set up to take pressure off hospitals.
• Three hundred thousand people migrated from Kathmandu to seek shelter and support from friends and family.
• Facebook launched a safety feature for users to indicate they were safe.

What were the long-term responses to the Nepal earthquake?

• A $3 million grant was provided by The Asian Development Bank (ADB) for immediate relief efforts and up to $200 million for the first phase of rehabilitation.
• Many countries donated aid. £73 million was donated by the UK (£23 million by the government and £50 million by the public). In addition to this, the UK provided 30 tonnes of humanitarian aid and eight tonnes of equipment.
• Landslides were cleared, and roads were repaired.
• Lakes that formed behind rivers damned by landslides were drained to avoid flooding.
• Stricter building codes were introduced.
• Thousands of homeless people were rehoused, and damaged homes were repaired.
• Over 7000 schools were rebuilt.
• Repairs were made to Everest base camp and trekking routes – by August 2015, new routes were established, and the government reopened the mountain to tourists.
• A blockade at the Indian border was cleared in late 2015, allowing better movement of fuels, medicines and construction materials.

Premium Resources

Please support internet geography.

If you've found the resources on this page useful please consider making a secure donation via PayPal to support the development of the site. The site is self-funded and your support is really appreciated.

Related Topics

Use the images below to explore related GeoTopics.

Previous Topic Page

Topic home, amatrice earthquake case study, share this:.

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)
• Click to share on Pinterest (Opens in new window)
• Click to email a link to a friend (Opens in new window)
• Click to share on WhatsApp (Opens in new window)
• Click to print (Opens in new window)

If you've found the resources on this site useful please consider making a secure donation via PayPal to support the development of the site. The site is self-funded and your support is really appreciated.

Search Internet Geography

Top posts and pages.

Latest Blog Entries

Pin It on Pinterest

• Click to share
• Print Friendly

What caused Dubai floods? Experts cite climate change, not cloud seeding

• Medium Text

DID CLOUD SEEDING CAUSE THE STORM?

CAN'T CREATE CLOUDS FROM NOTHING

Coming soon: Get the latest news and expert analysis about the state of the global economy with Reuters Econ World. Sign up here.

Reporting by Alexander Cornwell; editing by Maha El Dahan and Alexandra Hudson

Our Standards: The Thomson Reuters Trust Principles. New Tab , opens new tab

World Chevron

US rights report on India cites abuses in Manipur, harassment of media and minorities

The U.S. State Department's annual human rights assessment found "significant" abuses in India's northeastern Manipur state last year and attacks on minorities, journalists and dissenting voices in the rest of the country.