When secret government talks are hacked it shows no one is secure in the connected age

The end of privacy? Shutterstock

Hotel rooms aren’t as private as they used to be. Recent reports suggest luxury hotels may have been targeted by national intelligence services trying to spy on negotiations over Iran’s nuclear programme.

The talks weren’t bugged in the traditional way of hiding microphones in the room. Instead, hackers infected hotel computers with a computer virus that its discoverers say may have been used to gather information from the hotels' security cameras and phones.

The virus was discovered by cyber-security firm Kaspersky Labs when the company itself was infected by a sophisticated worm known as Duqu2. Kaspersky went about investigating which other systems around the world might have been attacked. Among the huge range of systems they checked, thousands of hotel systems were analysed. Most of these had not been subjected to an attack, but three luxury European hotels had also been hit by Duqu2.

Each was compromised before hosting key negotiations between Iran and world leaders regarding the country’s nuclear programme. Having previously been accused by the US of spying on the talks, Israel – which was not involved in the discussions – is now under suspicion of (and denies) deploying the virus.

Hacking a hotel room

Of course, full details of exactly what information has been leaked will take some time to understand. As we saw when Sony was hacked, further revelations are likely to emerge over time. What is apparent is that parts of the worm were designed to compress video, and others to collect communications data from phones and Wi-Fi networks.

Many hotels, especially luxury ones, use computerised camera surveillance and have many other sensor devices collecting and transmitting data, such as smart TVs. The fact that these three hotels were all scheduled to hold very sensitive talks before being attacked by highly sophisticated malware is unlikely to be a coincidence.

Hacking target Laurent Gillieron/EPA

There are a number of ways the worm could have been spread to the hotel computer systems. Viruses can, of course, be sent as attachments to emails and often spread in this way. Up-to-date security software can stop most known viruses. But in cases such as this, where the malware and the vulnerability it exploits were previously unknown, the virus is not detected and so can infect the machine.

Another possibility is that an employee or contractor or someone masquerading as such could have infected a machine at the hotels. Duqu2 is thought to be related to the virus Stuxnet, which brought down Iranian nuclear facilities and was spread, at least in part, through USB drives used by people working in the nuclear industry. Coincidentally, it is thought the infected USBs were likely to have been picked up from in hotels in India and Iran.

We are now living in a highly connected world that is increasingly dominated by smart devices and the so-called internet of things, where many objects and appliances gather data and are connected to the internet. These devices have all types of sensor and actuators and can be controlled remotely and without human intervention.

No escape

If these devices are controlled by someone other than the owner, they can be used to pass interesting information to the person in control. Last year, a Russian website streamed data from over 500 internet-connected video devices, including baby monitors. Accessing these devices didn’t even require advanced malware. Instead, hackers abused the failure of the devices' owners to set a complex password in order to gain control.

Numerous actors, from terrorists to cyber-criminals have an interest in accessing information from governments, companies and individuals. But Edward Snowden, who leaked details of the US and UK’s official data-capture programme, revealed just how much nation states also have a thirst for information, using both targeted and more blanket attacks to provide intelligence.

Clearly, in such a “smart” world, we need to get better at protecting access to our systems and devices, and that includes ensuring that the users smarten up too. This means not only ensuring our anti-virus software, firmware on our hardware, and operating systems are fully up-to-date, but also that we take care ourselves using USB devices or opening unknown attachments.

We are seeing an increase in political groups compromising the systems of companies, governments and individuals, as well as attacks for notoriety or financial gain. No system is beyond being a target, no matter how small or large.

Most of us don't read the social media small print – and it's a data goldmine for third parties

You may read paper, online is no different. Signing by Shutterstock

The history of human experiments often focuses on biomedical research and the gradual changes in acceptable practice and ethical considerations. But another class of human experiments that has had its own share of controversies is the study of human behaviour.

Internet Mediate Human Behaviour Research (IMHBR) is primarily defined by its use of the internet to obtain data about participants. While some of the research involves active participation with research subjects directly engaging with the research, for example through online surveys or experimental tasks, many studies take advantage of “found text” in blogs, discussion forums or other online spaces, analyses of hits on websites, or observation of other types of online activity such as search engine histories or logs of actions in online games.

It’s big business and the pervasive use of these methodologies is not only by academics but also corporations and governments seeking to support evidence-based policy decisions or to nudge societal behaviour.

Even though the basic principles of “respect for the autonomy and dignity of persons”, “scientific value”, “social responsibility” and “maximising benefits and minimising harm” are the same for this type of research method as for any other, the following issues often pose particular challenges for internet-mediated research: the distinction between public and private information, confidentiality, and informed consent. There is an urgently need to establish clear codes of ethical conduct for IMHBR.

Whose information is it?

The distinction between public and private domains is vitally important since this greatly affects the level of responsibility and obligation of the researcher. For human behaviour research online, however, it is often difficult to determine if participants perceive an online forum as “private” or “public”. While almost all internet communication is recorded and accessible to the mediating platform, such as Facebook and Twitter, and much of it even publicly accessible, users of these platforms may nevertheless consider those communications to be private, despite click-signing the terms and conditions of the service provider.

To quote professor John Preston’s testimony to the House of Commons science and technology committee on responsible use of data:

People treat social media a bit like they treat the pub. They feel that if they go into a pub and have a private conversation, it does not belong to the pub; it is their conversation. They interpret Twitter or Facebook in the same way – as a place to have a conversation.

This was also one of the contributing factors in the Samaritans' radar debacle where they proposed an alert system to flag when people were tweeting potential distress and suicidal messages. In its post-investigation communication by the Information Commissioner’s Office to the Samaritans, the ICO stated:

On your website you [Samaritans] say that ‘all the data is public, so user privacy is not an issue. Samaritans Radar analyses the tweets of people you follow, which are public tweets. It does not look at private tweets.’ It is our view that if organisations collect information from the internet and use it in a way that’s unfair, they could still breach the data protection principles even though the information was obtained from a publicly available source.

Read the small print. Terms and conditions by Shutterstock

Confidentiality

Anonymisation is one of the most basic steps for maintaining confidentiality and showing respect for the dignity of research participants. It is also a requirement imposed by the Data Protection Act 1998 when dealing with personal data. The need to protect the anonymity of participants is even more pressing when the research uses data from online sources where access to the raw data cannot be controlled by the researcher.

At the same time, the wealth of secondary information sources that can be mined in connection to any hint at the identity of a participant is making it increasingly easy to de-anonymise data. This was publicly shown by journalists for the New York Times who followed the web tail of user No. 4417749 in the AOL Search Log in 2006 and were able to identify her – and also by the lawsuit against Netflix for insufficient anonymisation of information disclosed in a prize competition database.

Terms and conditions that no one reads

In order for informed consent to take place, it is vital that the participant is fully aware of what is being consented to. Unfortunately, current online business practice has heavily eroded the concept of informed consent by habituating people to click-sign terms and conditions forms that are too long and unintelligible to understand.

Sometimes driven by social pressure to join the network their peers are using, people readily skip over the details and give their consent for allowing corporations to access their data for a wide range of purposes. A hint at the dangers of normalising such attitudes towards the concept of informed consent was given by the statement in the controversial 2014 “Facebook news feed manipulation experiment” – a secret study on “emotional contagion” that involved changing what 689,000 users saw from their friends' feeds to see if it influenced mood.

One of the researchers attempted to defend the study, saying that participants had provided consent because “it was consistent with Facebook’s data use policy, to which all users agree prior to creating an account on Facebook, constituting informed consent for this research”. The data use policy, however, does not provide any information about the nature of that specific study, instead speaking only of “research” in general terms.

Various organisations and learned societies, such as the British Psychological Society, the Association of Internet Researchers, the British Association for Applied Linguistics, the Information Commissioner’s Office, as well as our own research group at Nottingham University and many others are currently actively engaged in formulating and improving the guidelines for internet-mediated research.

As part of this work we are currently running a survey to ask citizens which conditions they would like to impose on researchers for making their social media data available to research studies. Ultimately, without clear guidelines and transparency, we’re hiving out decisions about us and our information to companies, governments and researchers, without us knowing what it will be used for.

Philae phones home -- but the mission is about to get riskier

Philae is awake again but a cometary tail threatens its future. ESA

I had planned to spend the day with my mother – nothing special, just a quiet Sunday with family. The plan lasted until midday, when the announcement that the comet lander Philae was finally awake made the twittersphere explode with excitement. Wonderful news – but it spelled the end of my day of rest. As I write this it’s midnight here in Milton Keynes and I’m putting the events into perspective.

On November 12 last year, the Philae spacecraft was released from its Rosetta mothership, travelling for seven hours to the surface of comet 67P/Churyumov-Gerasiemko to carry out a series of observations and analyses. The landing was not straightforward, and failure of the harpoons designed to anchor Philae did not work as planned.

The spacecraft bounced back into space, eventually settling several tens (or possibly hundreds) of metres away from its designated landing site. Unfortunately, the final landing site was in shadow and Philae could not use its solar panels to charge its second set of batteries. After around 60 hours of experiments, the primary battery was exhausted and Philae and its instruments went into hibernation.

Philae as it separated from Rosetta and headed for the comet. ESA

But the scientists and engineers who had worked so hard to build and operate Rosetta and Philae did not despair – Plan B was put into action. As the comet approached the sun, they anticipated that the sunlight falling on the solar panels would be sufficient to charge up the batteries. So Rosetta started a listening campaign, waiting to hear from its erstwhile flight companion. And just before midnight on June 13 – more than 7 months since Philae separated from Rosetta – ESA received a signal. Philae had phoned home!

Troublesome tail

Keeping the connection alive, however, might not be entirely straightforward. Philae has been in hibernation, in the cold, at temperatures well below the -45°C limit for the electronics to work. The first priority has been to warm Philae up – and news from ESA suggests that this has been successful.

Onboard data imply that Philae might have been awake for a couple of days, but unable to communicate with Rosetta until systems were warmer. The solar panels now seem to be generating power but ESA has to redesign Rosetta’s orbit so that Philae and Rosetta can be in better contact.

However, this is more complicated than it was last November, because the comet is much more active. As it moves closer to the Sun, jets and plumes of dust and gas are escaping from the surface – producing a cometary tail. Unfortunately this confuses Rosetta’s pointing device. If care is not taken, Rosetta might lose contact with mission control, which would almost certainly be the end of the mission. A regular Rosetta-Philae communications connection must be re-established, which will allow reliable transmission of data when Rosetta is above Philae’s horizon.

The Philae instrument teams have not been idle over the past seven months. They have had data to reduce and interpret, papers to write and lectures to give. They have also – as optimists with faith in the hardiness of Philae – been planning experiments for when Philae wakes up. Now that day has arrived, and they must return to the fray, ready for a much longer run of observations and analyses.

Over the next two months, Rosetta will fly even nearer to the sun, reaching its closest approach in mid-August. As the nucleus of the comet becomes more active, Philae’s observations and analyses will become more exciting, as fresh parts of the comet are revealed. But the mission also becomes more perilous, increasing surface activity could damage Philae. A worst-case scenario is that a jet blows the spacecraft off the surface and back into space.

Philae has woken up, and has work to do. I wish the instrument teams good luck and will keep my fingers crossed that Philae achieves its experimental goals. I also hope that next weekend is much quieter than this one!

Disclosure

Monica Grady receives funding from STFC and is a Trustee for Lunar Mission One. She is participating scientist with the Ptolemy Instrument team on Philae

Snoopers' Charter plans under fire from UK terror watchdog

Not dancing in the aisles Facundo Arrizabalaga/EPA

The author of the latest report on government surveillance practices is probably right when he says the document “won’t please everybody (indeed it may not please anybody)”. The Independent Reviewer of Terrorism Legislation, David Anderson QC, has laid out his proposals for reforming the way intelligence services collect and use data about our online activities. And the prime minister has reportedly already rejected one of the key recommendations that judges – not ministers – should authorise the interception of communications.

The report is a substantive piece of work and deserves careful reading and consideration in full. It argues that people’s internet searches should only be captured where a rigorous assessment proves there is a strong case for doing so. This is not in line with government plans in the so-called Snoopers' Charter.

It also emphatically rejects the idea that government should have a backdoor into all encrypted communications. And it highlights the need to replace the current patchwork of communications laws with a single framework and a new oversight body.

It’s becoming clear why the government delayed the report’s publication. The opposition of David Cameron and Home Secretary Theresa May to judicial rather than ministerial authorisation of warrants was predictable. You can also understand why they might be a bit miffed that Anderson disapproves of blanket encryption backdoors. He argues the agencies don’t want it and that it would undermine security for everyone.

Government disappointment

You would expect the government to be positively dancing in the aisles for the report’s apparent support for the bulk data collection and retention. And its support for the notion that UK law in this area should have global reach.

However, their joy will be tempered by Anderson’s qualification that he is not offering a legal opinion that these practices are proportionate. On the contrary, he notes: “A number of such questions are currently before the courts.” He also insists bulk collection and retention would have to comply with the European Convention on Human Rights and the ruling by the European Court of Justice last year that outlawed indiscriminate data retention.

And on international communications, Anderson says capturing this data without the official agreement of other countries is an “unsatisfactory substitute for a multilateral arrangement … which must surely be the long-term goal.”

So Anderson’s report has turned out to be nothing like the useful excuse for pushing through the Snoopers’ Charter that the home secretary must have hoped it would be.

Bulk data. Shutterstock

On the down side, I have to admit I share Privacy International’s disappointment that Anderson did not condemn bulk data collection. The justification for this is rather weak and not in line with the deeper consideration of the rest of the report.

Anderson links the issue to the principle of minimising areas – in both the physical and digital worlds – where law enforcement doesn’t apply. He supports this position by reference to six real bulk data case studies in Annex 9 of the report. None of these cases definitively demonstrate that bulk collection was the main reason those involved were identified in the first place.

Information overload

If the police or intelligence services have just cause to suspect someone of criminal activity, then being able to access and search the suspect’s data could reveal significant information about them. But authorities simply do not have the resources to sift through data about the lives of everyone in the country.

Time and again since the turn of the century, from 9/11 to the murder of Lee Rigby, authorities have failed to prevent terrorist attacks by known dangerous individuals after being overloaded with information from bulk collection. Additionally, it is simply not proportionate to engage in bulk data collection, in the hope that it will be useful when the authorities decide to look into someone they disapprove of. It actually impedes already over-stretched authorities, who would be better off recruiting more and better-trained investigators and analysts.

The government would do well to note that the opportunity costs of bulk data collection and retention make the jobs of those tasked with protecting us more difficult, while simultaneously denying them the resources to be more effective. This undermines security for everyone.

How to sail through space on sunbeams – solar satellite leads the way

The future of space travel? Planetary Society

A new kind of spacecraft that sails on sunlight has just been successfully deployed above Earth. The LightSail satellite yesterday overcame a series of glitches to unfurl its solar sails, a propulsion system that’s entirely different to traditional rockets, thrusters or even solar panels.

The prototype craft, built by the US Planetary Society and funded by public donations, is testing the sail deployment mechanism. Another mission due to be launched next year will trial the full propulsion system. And it’s not the first spacecraft to fly using solar sailing. That honour went to the Japanese craft Ikaros in 2010. But the growing interest in the technology creates some exciting possibilities for the future of space travel.

Solar sailing is the space equivalent of boat sailing. A boat is pushed along by the air particles that strike its sail when the wind blows. The idea behind solar sailing is the same except, instead of wind, particles of light emitted by the Sun drive the craft forward.

Light is made up of moving particles known as photons that can be reflected by a mirror. When this happens, the mirror experiences a pressure across its surface called solar radiation pressure. This pressure is very small, so you couldn’t feel it just by holding up a normal mirror in sunlight while on Earth. But in the vaccuum of space it’s enough to move a very light mirrored sail.

To understand why solar sailing is important, we need to go back to the emptiness of space. Propulsion is the ability to accelerate a vehicle. Vehicles on Earth accelerate by interacting with something around them. For example, a car accelerates by producing a force on the tarmac.

In space, however, there is no medium to interact with. This is where rocket propulsion comes in. By ejecting propellant in one direction, the rocket is pushed in the opposite direction. This is because of the scientific law (Newton’s Third Law of Motion) that says every time an object exerts a force on another object it experiences an equal and opposite force.

The catch is the propellant on a spacecraft is limited, and once it is all used up, no further no acceleration is possible. Because of this, some distant targets in space are impossible to reach with pure rocketry. Solar sailing, instead, goes elegantly around this this problem by not requiring any propellant mass at all. It is like having a rocket that can thrust for an indefinite amount of time, for free.

A solar sail produces thrust without propellant mass by simply reflecting incident photons from the sun. from www.deorbitsail.com

The concept of solar sailing is very simple, almost romantic in a sense – and definitely appealing for space exploration. Why then has it not been more widely exploited since its first conception in the 1920s? There are a number of reasons, which are all related to engineering challenges.

The main one is the fact that the solar radiation pressure is very small (hence why you cannot feel any force if you hold a mirror up to the sun). This means solar sails have to be large in order to reflect more light but also lightweight so less force is needed to move them.

And in order to be safely launched, the sail must be folded away and then deployed in weightlessness. The main body of the LightSail craft containing the folded sail is approximately 30cm across at its widest point but the sail deploys with a total area of 32 m². This deployment mechanism must be very reliable because if it fails, the whole mission is wasted (hence LightSail’s practice launch).

Even with large and lightweight sails, the acceleration can still be small. A solar sail covering two tennis courts would experience a force equivalent to the weight of 1g on Earth. If a spacecraft has a mass of 30 kg, its acceleration will be just 0.0003 m/s². This sounds extremely small for space standards: even a 1980 Fiat Panda does much better, at about 1 m/s².

However, a solar sail can orbit the sun over and over, slowly but continuously accelerating (or braking) towards its target. This is why sails are suited for long, interplanetary trips in the solar system. But while the sail can theoretically accelerate forever, practically there are time constraints. Radiation, impacts with micrometeoroids and outgassing (the release of trapped gas from within electronic instruments) can damage the sail material, making it less reflective with time.

The advantage of projects such as LightSail, even though it is a small prototype device, are that they demonstrate the feasibility of the technology and create greater interest and enthusiasm in this mode of flights. LightSail shows that, even a hundred years after being first proposed, solar sailing is still a very popular idea – and scientists (such as ourselves at Space Glasgow) still consider it promising for space exploration.

Artist impression of LightSail in Earth orbit The Planetary Society/Rick Sternbach

The hope is that it will attract further substantial investment to the sector in order to develop the technologies required to make solar sailing the main propulsion system of future spacecraft. We are almost there!

AI 'cheating' scandal makes machine learning sound like a sport – it isn't

Under an uncomfortable spotlight. Baidu image via Gil C/Shutterstock.com

News that Baidu, the Google of China, cheated to take the lead in an international competition for artificial intelligence technology has caused a storm among computer science researchers. It has been called machine learning’s “first cheating scandal” by MIT Technology Review and Baidu is now barred from the competition.

The Imagenet Challenge is a competition run by a group of American computer scientists which involves recognising and classifying a series of objects in digital images. The competition itself is no Turing test, but it is an important challenge, and one of commercial importance to many firms.

The cheating by Baidu was nothing sophisticated, more akin to an initial stolen glimpse at the answers, which was followed by more of the same when it went unnoticed. Even that makes it sound worse than it was. Part of the competition involved looking at the answers anyway: someone in the Baidu team simply did it more than they were officially allowed to.

In their paper about the submission, Baidu themselves weren’t claiming anything more than an engineering advance: they built a large supercomputer that could handle more data than previous implementations. A necessary advance, but very much a “scaling up” of existing solutions – one that would be financially outside the reach of a typical academic research group. They participated in the competition as an attempt to demonstrate that, after such significant investment in hardware, their new supercomputer was able to perform. They have since apologised for breaking the rules of the competition.

In any case, the significant breakthrough in the area had already been achieved by Geoff Hinton’s group at the University of Toronto. They produced the machine learning equivalent of the high jump’s “Frosbury Flop” to win the 2012 version of the competition with such a significant improvement that all leading entries are now derived from their model. That model itself also built on a two-decade-long program of research by Yann LeCun, then of New York University.

Blown out of proportion

The result of Baidu’s entry into the competition was posted as an “e-print” publication. E-prints are articles that are unreviewed. They are a slightly more formal versions of a “technical blog post”. The problem was identified by the community quickly, within three weeks, and a corrected version was published. This is science in action.

The “cheating scandal” was labelled as such by the very same prestigious technical publication that broadcast the initial results to its readers within two days of the e-print’s publication: MIT Technology Review.

Singling out MIT Technology Review in this case may be a little unfair, because this is part of a wider phenomenon where technical results are trumpeted in the press before they are fully tasted (let alone digested) by the scientific community. E-print publication is a good thing, it allows ideas to be spread quickly. However, the implications of those ideas need to be understood before they are presented as scientific fact.

Ideally knowledge moves forward through academic consensus, but in practice that consensus itself is swayed by outside forces. This raises questions about who is the ultimate arbiter of academic quality. One answer is opinion: the opinion of those that matter, such as governments, businesses, other scientists or even the press. Success in machine learning has meant it is attracting such attention.

Getting on with it for decades

Ironically, the developments that enabled recent breakthroughs in AI all took place outside of such close scrutiny. In 2004 the Canadian Institute for Advanced Research (CIFAR) funded a far-sighted program of research. An international collaboration of researchers was given the time, intellectual space and money that they needed to make these significant breakthroughs. This collaboration was led by Geoff Hinton, the same researcher whose team achieved the 2012 breakthrough result.

This breakthrough led to all the major internet giants fighting for their pound of academic flesh. Of those researchers involved in CIFAR, Hinton has been hired by Google, Yann LeCun leads Facebook’s AI Research team, Andrew Ng heads up research at Baidu and Nando de Freitas was recently recruited to Google DeepMind, the London start-up that Google lavished £400m on acquiring.

The Baidu cheating case is symptomatic of a big change in the landscape for those who work in machine learning and who drove these advances in AI. Until 2012, ideas from researchers in machine learning were under the radar. They were widely adopted commercially by companies like Microsoft and Google, but they did not impinge much on public consciousness. But two breakthrough results brought these ideas to the fore in the the public mind. The Imagenet result by Hinton’s team was one. The other was a program that could learn to play Atari video games. It was created by DeepMind, triggering their purchase by Google.

However, just as Deep Blue’s defeat of Kasparov didn’t herald a dawn in the age of the super-intelligent computer, neither will either of these significant accomplishments. They have not arrived through better understanding of the fundamentals of intelligence, but through more data, more computing power and more experience.

Who follows in whose wake?

These apparent breakthroughs have whetted the appetite. The technical press is becoming susceptible to tabloid sensationalism in this area, but who can blame them as companies and universities ramp up their claims of scientific advance? The advances are somewhat of an illusion, they are the march of technologists following in a scientific wake.

The wake-generators are a much harder to identify or track, even for their fellow scientists. But the very real danger is that expectations of significant advance or misunderstanding of the underlying phenomenon will bring about an AI bubble of the type we saw 30 years ago. Such bubbles are very damaging. When high expectations aren’t immediately fulfilled then entire academic domains can be dismissed and far-seeing proposals like CIFAR’s go unfunded.

Academics make those first waves. Boat wake via Dennis Tokarzewski/www.shutterstock.com

Even if Baidu’s result were valid, it would have been just the type of workaday scientific development that most of us spend most of our time trying to cook up. It did not merit a pre-publication announcement in MIT Technology Review and the pre-publication withdrawal should have been just a footnote to add to the diverse collection that keep all astute academics scientifically wary. Rather boringly, the only true marker of scientific advance is repeatability. Whether that is within the scientific community or by transfer of ideas to the commercial world.

When reporting on the scandal MIT Technology Review refers to participation in these competitions as a “sport”. I feel sporting analogies give a wrong idea of the spectacle of scientific progress. It is more like watching a painter at work. It is very rare that any single brushstroke reveals the entire picture. And even when the picture is complete, it may only tell us a limited amount about what the next creation will be.

Fracking technique raises prospect of better prediction of mega-earthquakes

Could fracking save lives? Joshua Doubek/wikimedia, CC BY-SA

An experiment has shed light on how controversial human activities such as fracking can cause small earthquakes. This may enable better controls of hazardous techniques. But it could have even greater relevance as a way to predict big earthquakes and other natural hazards.

The past decade has seen a significant growth in so-called “fluid-injection techniques". In the most common one – fracking – shale gas is released by the drilling of a high-pressure water mixture into the earth.

The environmental implications of such practices are highly debated. One of the problems is that these injections produce small earthquakes where there is no historical seismic record, meaning we don’t have a good idea about the hazards and vulnerability of these places.

An oil refinery on fire following the 2011 Tohoku Earthquake in Chiba city, Japan. EPA

However, the study is a perfect frame for understanding how fault slip works in the presence of fluid flows. This is extremely relevant at large scales, because we know that nature’s own hot fluid injection from the mantle and in subduction processes, when tectonic plates meet, may be a factor in the development of mega-quakes – albeit at a much greater depth. The better we can understand injection, the better the chances that we can one day use it in forecasting quakes.

Quakes: from micro to mega

The experiment looked at fluid injection at a depth of 282 metres in a well-known setting: the cretaceous limestone of the south-east France sedimentary basin. What the scientists did was to activate a natural but inactive fault with a high-pressure injection of water and monitor the consequences with advanced instruments.

The measurements enabled the team to get a complete 3D image of the dynamic of an injected fault in real time, explaining why injection-induced techniques work. They discovered that there was a lack of audible small earthquakes during most of the injection, with a 14-fold increase of “permeability” – the Earth’s ability to allow gas and light fluids to move upward due to deep pressure – before any seismic event occurred. This sets an important time limit to when the injection should stop before causing hazard.

The study is also important to seismologists because the energy that is typically measured from small earthquakes – and used to try to predict bigger ones – is three order of magnitude lower than the effective energy released from the small earthquakes in the experiment. This is because just measuring the energy of the actual quake doesn’t take into account energy released before the eruption from the movement of the fault. Therefore, the “seismic magnitude” technique that we use for predicting hazards from earthquakes strongly underestimates the true energy release of the fault slip, making it an imperfect and late sign of hazard.

The experiment shows that a better alternative for monitoring injection-induced fault slip is “aseismic creeping” – which is the displacement along a fault in the absence of earthquakes – which starts soon after injection. This creeping may be produced by friction of the uneven surfaces along the fault interfaces and is a much earlier symptom of the fault’s movement than small earthquakes.

When there is subduction or rifting, the processes where two tectonic plates meet or separate, creeping results in shallow tremors known as slow earthquakes. These have been recorded just before destructive mega-quakes, as the Tohoku earthquake in Japan, possibly as a consequence of a fluid injection from depth. More recently, along the Kyushu Palau Ridge in the Philippine Sea, this creeping has been detected and monitored, which could be a sign that Japan faces another mega-quake.

These results are a critical advancement for understanding the dynamic of a fault at shallow depths. They can be used by the industry in order to improve methodology and better account for seismic hazard during fracking.

The study shows that creeping can provide us with forecasting information. If scientists can scale up these results to the depths and structures that produce mega-quakes, fluid-injection technology will have granted a value far greater than the economic benefits of fracking.