 Good morning, my name is Jonostor Snapsson and I come from Reykjavik University in Iceland. And my idea was to present to you some of the... how does this work? It doesn't work? No, it's working. Okay. So, my idea was to present to you what I... well, monitoring systems in Iceland monitoring seismic hazard and discuss some of the values that I feel are involved in such monitoring. This is just a review of the key issues that I'm going to address. But basically, I'm going to introduce to you the current seismic networks in Iceland and there's a significance for society and industry. So, seismic networks, they differ from traditional structural health monitoring networks in the sense that they are aimed at mapping gear hazard and earthquake action and they have usually either both a wide focus on national or regional level or they can have a narrower focus on specific towns or sites or single structures through arrays or structural monitoring setups. So, these networks, they serve a multitude of stakeholders. Unfortunately, the stakeholders usually don't realize their benefits fully and therefore it's often difficult to fund the initiation and operation of these types of networks. But the information they provide is, in my opinion, essential to be able to build a safe and resilient society. So, if we just list up some of the key stakeholders as I see them, then of course we have the general population and house owners which depend on the safety and resilience of their buildings and infrastructure and also on disaster preparedness which relies on information about where the weak spots are where the hazard is most critical and how it is distributed throughout a country or a region. The municipalities, in the same way, they use the information for planning decisions and for setting building regulations similar to the state on a slightly different level. They also have planning requirements and building regulations and design requirements. And then, in the case of Iceland, the state operates insurance and reinsurance for public and private real estates. And then, of course, the major industry needs this information both for sensible site selection and design specifications and also for their insurance and reinsurance of their assets. Now, if we look at the case of Iceland, Iceland is located in the middle of the Atlantic Ocean on the Mid-Atlantic Ridge and basically the Mid-Atlantic Ridge crosses through Iceland and the tectonics are sort of controlled by the combination of plate and plume tectonics. So we have Iceland is located on a spreading zone and you see the ages in millions of annums. So in Iceland, some parts of the Iceland are zero year old and for then some parts are 10 million, 20 million. And then the age increases towards Greenland and also, of course, towards Europe. But we have, through this continuous spreading, we have, of course, both volcanic and earthquake activity. So we are being pulled apart by the American plate and the Eurasian plate by a rate of about two centimeters per year. Now Iceland and earthquakes have been recorded since around the 1900s. So we have some sporadic records from 1886 to 1927, mostly recorded, of course, in Europe or US. And then since 1927 there have been more or less continuous seismological observations. Now we have basically two sets of networks. One is, that is seismological networks based on seismometers and a strong motion accelerometer network. The earlier part of the seismological network was basically built or is composed of analog seismographs and they were installed in the 60s and 70s. And then digital seismometers were installed in the 90s. And then in 85 and we started installing strong motion accelerometers. There had been one or two instruments installed in 70, two or three, but sort of systematic network was not initiated until 1985. And all these networks were initiated by scientists, by scientists at the University of Iceland and scientists at the McGraw offices. And the initial funding came from research funds and from companies like, for instance, the strong motion accelerometer, the meteor network was supported by the national power company. But the initiation was not really a governmental initiative, although they have supported the operation to some extent. But there is these networks differ in the sense that seismological networks are basically aimed at mapping seismicity for geological purposes and they are focusing on very small magnitude earthquakes and use velocity sensors for the most part, monitoring specific frequencies to get the greatest sensitivity. Whereas strong motion networks use acceleration meters and or acceleration sensors. And the aim is to monitor strong motion for design of buildings and structures. We also have continuous GPS monitoring networks and both through some fixed stations. And then there have been regular campaigns where mobile stations have been placed on fixed points. And this is a result of two sorts campaigns, one in 1994, 2004. So 10 years apart and based on these data we can sort of map the present day geodynamics of Iceland. So what we see is horizontal velocities. So we have the north-south component, the green one, and the red one is the east-west component. The red one is the west component, the blue one is the north-east component, but the resultant component. So you see this part of Iceland is moving in this way and this part is moving this way. And we have the rift axis going through here in between. And therefore of course the seismometers are distributed more or less along this line. And I'm not going to discuss this system in much more detail, but this is what comes out of a seismological system. You have mapping of epicenters of small earthquakes which show sort of where the seismicity is the greatest. And then basically we have the Mid-Atlantic Ridge coming here into Reykjanes Peninsula. And then we have the rift zone traveling through the volcanic zones in Iceland. And then again out to the ridge which continues here. But in between we have transfer zones in the south and we have transfer zones in the north. And in these transfer zones we have strike-slip earthquakes which are the biggest earthquakes that we get. And if we reduce the amount of events by limiting the size. So this figure here shows some magnitudes of 3.5 and greater. And then we see that there are not that many big earthquakes. And the biggest ones are in the south where the red dots are and in the north on these two transfer zones. And that's why the focus of the strong motion network is in the south lowland and in the north. In the north the faulting is mostly out in the ocean. And there is just a few of the faults that actually reach the shore. So there is a different faulting mechanism basically here in the north and in the south. But there is about 40 strong motion stations operating now providing ground motion. And then we have some office buildings and power stations monitored for structural responses as well and bridges. Now this shows and then I forgot to mention that in two towns in Credit Gerdie here and in Pusavik here in the north we have Aris where we have a dense accelerometer network within an area of 2 or 3 kilometers in each direction. And this is from an earthquake in June 2008. And you see that the acceleration levels recorded in this town were quite significant or about 80% of G. But the fault was only about 2 kilometers from the town. And in the north we have recently installed a second RA in Pusavik. And this is actually where we expect the nest earthquake to occur and the town is located here. And the fault that we see, the activity here on the black dots and the green dots. That's actually running here just slightly north of the town, these two major faults. Now we have of course used the data that we have recorded for Mario's purposes. Such as strong ground motion modeling, seismic hazard assessment. We have been evaluated vulnerability profiles for the building population. We've used the structural data for system identification. We have observed very significant side effects at many places and etc. And of course for engineering education, not the least. And this shows the ground motion modeling results. And what we see is that we have considerably more rapid attenuation of motion than is usual for other areas. And that is, we believe, related to a young and fractured crust in Iceland. And if we look at this in a linear scale, then we see that if we are about 20 kilometers away from the fault, then we are quite safe. And we also have seen that the duration of strong motion is relatively short within the critical area. So we are talking about usually less than a 10 second duration within the 20 kilometers zone. So this is the hazard map that was created or proposed in 2004. And we see the south Iceland lowland here is the highest ground acceleration and then the northern coast here. And the biggest population is in Reykjavík, which is here. And there we have an acceleration levels of between 10 and 20 percent, or yeah, mostly less than 15 actually. And most of the data comes from these three big earthquakes that we had to win the year 2000 in the south Iceland lowland. And then in 2008, year between Kveragér and SELFOS. And in this area, there is about 10, 15,000 people. So the density of population is much greater for this earthquake here. The density of population affected by this earthquake was much greater than the density of population affected by these two in 2000. And all these three earthquakes were of similar magnitude, 6.5, 6.4, 6.3. And the first two had about 20 kilometers surface rupture. The surface rupture was not well visible for the third one. But this is a sort of a traditional pattern for seismicity in the south lowland that we get a series of events with relatively short time lapse in between. And these series of events occur approximately every 100 years. The sort of most interesting data has been available through this internet site for European strong motion data that was established in 2002, I think, by several partners through a European project. And at least this site is still active. And it contains about 3,000 processed strong motion records from all of Europe and parts of Turkey. Now in Iceland, we have a catastrophic insurance fund. It was funded in 1975 after a volcanic eruption in the Westman Islands. And it acts as an insurance company. And the purpose is to operate a catastrophic insurance which should cover damages from earthquakes, volcanic eruptions, avalanches, rock slides and floods. And basically all buildings are insured according to the valuation for fire. And we pay a fee that is 0.02% of the insured value each year. And so fire insurance is compulsory. More or less every property is basically insured by this except properties or assets owned by major industry which often choose to have a separate insurance or self insurance. But part of the infrastructure and lifelines is insured by the ICI. And in 2014, the ICI covered assets of about 60 billion euro. And of course the assistance of a fund like this increases the resilience of the society. And we have of course had, we regularly have some catastrophes in a small scale. The Westman Islands was the trigger for this, for initiating this fund. And that of course we lost a whole town of 5,000 people which has been resettled now. But by 4,000 people not 5,000, but still flourishing. But of course that was a big blow. It was about 13% of the gross domestic product. And then we have of course we have earthquakes in the north in 76. Modern ice flows, wind driven ocean flooding, storms, avalanches which are actually the events that have killed people. Other events have not been really, people have not died in the other events except the volcanic eruption. There was one person who died there. But there were about 30 people who died in this avalanche in 95. Glacier Alps was flooding earthquakes in 2,000, volcanic eruption in Eyjafjallajökull just so you know how to pronounce it. And one year later in Greenfoot. But basically these events are smaller in terms of gross domestic product. And even though the expense of the earthquakes comes close to the expense of the volcanic eruption in Westman Islands. Because of better economics it does not affect our national budget as strongly. But this graph shows sort of the payment that had been cast out in the past 10 years. And then it is because this starts in 2004 that earthquakes in 2000 or not. Or just vaguely included here. But it's the earthquake in 2008 that has been the most expensive event. But basically the fund is supposed to be able to cover a loss of 100 billion Icelandic Kronos or about 500 million euros. And it's divided into sort of own liability of 10 million. And then you have reinsurance. There is a 20 group of 20 reinsurance that each covers 1 billion Icelandic Kronos. And then the fund has collected money. So it has some capital of its own. Now it's 17.8 billion Icelandic Kronos. And then it is possible to take a loan with a government guarantee to cover the rest if needed. But as you see we have not really needed all this funds so far. And the reinsurers have not done so badly. We have been paying premiums of 25 million euros for the past 10 or 4-10 year period here, 2000 to 2011. And we only claimed 23 million back. So they made profit of 2 million roughly. And of course there is a regular re-estimation of hazard and risk when the premiums are determined and the self liability. And then foreign experts come to evaluate the hazard in Iceland. And they make the right reports with risk scenarios and hazard scenarios. And often they are way off target. And Icelandic experts have had to fight with the insurance fund to correct some misunderstandings and exaggerations. And just to, as an example, not a direct example of this, but as a sort of a site example that actually has influence is that in 2000 there was a project, the SHIR project, seismic hazard harmonization in Europe operating in recently. And they produced a European seismic hazard model which is probably a very good model. But then they decided to apply it to the whole of Europe. And to do that you need to simplify the seismic sonation. And for Iceland this has catastrophic consequences because suddenly the risk in many areas is already hazard is greatly exaggerated for large parts of Iceland. We have here 50% of G and now all of Reykjavik has acceleration levels up to 50% of G and so does a larger portion of the country. And this basically comes from oversimplification of and maybe to some extent misinterpretation of the seismic sonation, appropriate seismic sonation for Iceland. And it's important to, sorry for this. And this is a comparison of two maps, one from the previous one that I show you based on Solna Zetal in 2004 which is based on Icelandic data and only considering the Icelandic tectonics. And then this is from this year, one of the sheer projects hazard figures. And then the Reykjavik peninsula is the critical one. The South Lowland is not so critical. And then they overestimate the seismicity around the glaciers which is primarily volcanic seismicity where you don't really have this type of acceleration levels and you don't have very high magnitude earthquakes. And it is particularly important to have correct seismic hazard defined for the area around Reykjavik because this is supposed to show you the population density in Iceland. And so that's the black dots basically. And about two thirds of the population live in the southwest corner of Iceland. And if we just look at the Reykjavik area then that's probably close to half of the population. And this shows the sum of insured aggregates by region. And as you see the sum is 54 for the Reykjavik area whereas it is 10 times less for most other regions. And therefore it is important to realize that in Reykjavik peninsula we have a different type of faulting than in the South Lowland. We have magnitudes of earthquake that are less than six. We have normal faulting and we have a very narrow seismic zone with shallow focus earthquakes here. And so this would be a more appropriate seismic source formation where you basically have a line here that has not experienced any earthquakes greater than four as far as we know. And the red dots are the earthquakes of six plus and in this area we have three earthquakes of six plus for the last 300 years. The green dots are the distribution of smaller magnitude earthquakes four and five, green are five, the yellow are four and those are monitored in the past 30, 40 years. So this is of importance. And it's also of importance when you are trying to attract new business into the country. We have a lot to offer. We have relatively cheap energy compared to other places. This is statistics from Verne Global which is a data center company. And we have hydroelectric and geothermal energy so we have a fairly low carbon footprint compared to many other places. But if we have unrealistically high hazard then of course new industry is not likely to be interested in settling in Iceland. But so to summarize based on the recordings that we have and experience that we have had with the strong ocean monitoring in Iceland, we can say that the earthquake hazard can be quantified as moderate on an international scale. It's fairly localized and primarily limited to the transfer zones in the south and north and the immediate vicinity. And in the south Iceland earthquakes in 2000 and 2008 all showed similar characteristics. We had considerable spatial variation in ground motion peak parameters partly because of specific side effects where we have lava flowing over ocean sediments, creating a soft layer under a stiff lava layer. We had high peak accelerations which were partly also due to these side conditions. We had fairly short duration of motion and rapid attenuation of motion which all made life which both made life easier for the structures in the area. And even though we had considerable damage to buildings and their contents and perhaps especially utility systems in the ground, these recent events demonstrated that buildings and installed equipment can reliably be designed to withstand these expected actions. And electrical and telephone utilities operated more or less uninterrupted during and after the earthquakes. So we feel that we have sampled valuable earthquake induced accelerometric data and we have applied the data for various engineering purposes which will in the future have further implications for the understanding of earthquakes in Iceland and structural design. We feel that strong motion recordings provide indispensable information for structural design and codification. And the data and the local expertise and experience developed through data monitoring, data analysis and related studies are important for the future development of infrastructure and industry through realistic assessment of the relevant geohasert and the related risk. Thank you.