PubTalk 10/2019 — Remembering Loma Prieta Earthquake 30 Years Later

Video Transcript
Download Video
Right-click and save to download

Detailed Description

Remembering the Loma Prieta Earthquake 30 Years Late: Looking back to see how far we've come.
By: Tom Holzer, David Schwartz, Jessica Murray, and Annemarie Baltay

  • The Loma Prieta earthquake ended decades of seismic tranquility in the Bay Area - Is the earthquake threat in the Bay Area real and imminent?
  • Up to $80 billion in earthquake mitigation investments have been made since the Loma Prieta earthquake - are we safer?
  • Learn how Loma Prieta has led to improvements to building codes, land use, ground motion simulations, and earthquake early warning.


Date Taken:

Length: 01:48:10

Location Taken: Menlo Park, CA, US


[ Please stand by for realtime  captions ] [ Indiscernible - Low  Volume ] how many of you  remember that. My recollection was  that  it had [ Indiscernible ].  That's what  I remember. That's  actually  pretty good you  know if I  can find. 
     Probably many of you recall these  iconic  images from Loma Prieta  .   In Oakland. This is one of the collapsed  buildings in the  Marina district. The one that  burnt. And the Bay  Bridge that David McClanahan  was  talking about. And what we  did was pair them  with those sites  today.  The structures, part the  Marina district has been substantially  retrofitted and rebuilt. The Bay  Bridge has been replaced with  this beautiful cable bridge. So  just the  pictures alone tell the story that  we are going to try and tell tonight  in a more sophisticated way in terms  of science. How far we have  come in the 30 years since the Loma  Prieta  earthquake. Which will  have its anniversary  next Thursday. >> The  Loma Prieta   earthquake  occurred  about 60 miles south of San Francisco.  So those images we are looking at  work actually unusual damage in  terms of the distance from  the earthquake it ruptured about  11 miles down about 25 miles  of fault. So those  three sites the viaduct and  the Dave were Bay Bridge and the  Marina district were really pretty  unlucky. Because  3 things happened during the earthquake  one of them was there  was a bilateral  rupture that propagated off  in both directions I will trace  the cell with the arrow.  It sent a lot of seismic energy  rather than dispersing it to  San Francisco. The other problem  is that a lot of them are built on  soft soil. It was a  little unusual it was a  little unheard-of.  It  amplified themselves. There's  this trifecta in  the end people were killed when  the viaduct structure  collapsed. About 3800 people injured  and about 12,000 people  were displaced. Increasingly  realizing that  the major problems with earthquakes  is people being displaced from their  homes. It becomes  a challenging responsibility.  For earthquake mitigation.  Loma Prieta  left us   with quite a few legacies. Our speakers  tonight will be  talking about some  of these. Probably the most important  one was Loma Prieta  was the first  large earthquake since the 1906  earthquake to strike  the bay area.  So what it does  was, it educated if you  generations to the reality of the  earthquake. We sort of knew about  it from  1906. There were many disasters  as time passed I didn't seem  as serious. But Loma Prieta   educated us. The second thing that  happened, Dave will be talking  about this in the end. It would  begin to develop ways  of estimating probabilities of  earthquakes. And we had  a substantial  probability in the bay area. Not  only did we see  in earthquake. But there was a sign  telling us that this is actually  a  serious problem. It became a  poster child  for that. And the second  thing convoluted that we talk about  was seismic quake. It really  educated us to  the complexity of the earthquake  and Anne-Marie is going to talk  about that. Finally Loma Prieta  led  to improved building codes because  we  had not really had, other than the  San Fernando Valley earthquake in  Southern California, we hadn't  had a good urban  earthquake to shake us up, so to  speak, and  last thing that's important is one  thing that happened in the Marina  district was I  can't think of how many times I've  heard people say why weren't we  warned about this. So  Willie Brown, who  was the representative of the Marina  district at the time introduced  legislation called the  seismic hazard matched. One thing  that happened was we had learned  how to map  these areas for seismic landslides this  is highway 17 that was closed after  the earthquake it  was legislative that the geological  survey do  this but the urban areas in California.  If you buy property  with one of these hazards it will  be disclosed. This finished  the land had been  met starting in 1972. All the things  we can  really destroy which was actually  the  most serious this shows  that  tremendous mitigation about  70 say strengthening  structures and if you  look at pie chart you can  see about half of it is lifelines,  transportation, water.  And there's  some electric and gas. The other  half was critical facilities like  hospitals and schools. And the other  interesting thing is when you  look  at where this money is being spent  you can see that most of it is being  spent in the northern bay area.  Most of these counties  you know that it's kind of  interesting here. The earthquake  was down near Santa Cruz  and yet most of the  medication money was actually spent  in the northern bay  area. Did neglect to  mention that Santa  Cruz was an interesting situation  after the earthquake because  it predated cell phones and the  intranet. So there was little  attention paid to Santa Cruz. They  had a little trouble getting the  word out so it took a few  days before emergency response focused  on  Santa Cruz. I think  were going to talk a little bit  more about what we have done to  address that  problem. Let me introduce our three  speakers. Dave Schwartz is  a geologist and he'll be talking  about the evolving geology and  the probabilities. Annamarie is  a seismologist who will talk about  ground shaking and Jessica Murray  will talk  about the advances in terms  of measuring the crustal movement.  In a sense all of  you are geologist now. Because if  you open up your cell phone you've  got a GPS  in it. You can look at your cell  phone almost as well as Jessica  can. So she can do  a little better job. But it's become  a very important tool in terms of  looking at that.  So he go Dave. Take  the helm. >> >>Thank you  Tom and  welcome everybody. I am a geologist  and I  was around for Loma Prieta . But  I'd like  to do tonight is tell  you a little bit about what  happened geologically. What we've  done in these 30 years since Loma  Prieta  and what  we have learned.  Talk a little bit  about earthquake probabilities.  We live in the  boundaries between the North American  plate  and those two plates are sliding  past each other at about 40 millimeters  per year that would doesn't sound  like very much. Just a real  balance here. What  goes in in  terms of plate boundaries stress.  It comes out  as co-seismic slip  on fault.  Sometimes faults creep and you'll  hear about that from Jessica  meaning  that they move slowly all the time  and sometimes the fault never reaches  the surface. It might fold  the crust so you.all of these  different things going on. One  of her goals is to solve the mystery  of the plate boundaries. We want  to know the things of current clusters  are there periods when  it's quiet. How regular and variable  is this behavior. Is it mostly  random or is there something more  systematic. So, we really haven't  learned everything in  the last 30 years. There were  geologists way back  who appreciated falls. The map on  the left is from the report is published  after the 1906 earthquake. At  that time the San Andreas  Fault was known. The  Hayward fault was known a part of  the central Calaveras  was on the map and a part of the  green valley faults was on the map.  On the right is  just a fault map showing the  major structures. We've learned  a lot we've added more fault. But  they didn't know these  things back in 1906. So when  the  earthquake occurred. I left early to  go watch the game in Walnut Creek  where we were living at  the time. You know I put my feet  up on the kitchen table and then  you saw what happened with  the TV screen. The game was called  off and I was like Dang but what  I was really curious about was what  fault had produced  this earthquake. I was sitting there  and I felt the  house move it was the P wave coming  in. I knew the earthquake  was large and somewhere to  the south but I didn't really know  where. It was hard to communicate  with Menlo Park because the bridges  were closed for inspection so I  couldn't make down the  next day. Heavy traffic you know  so it's like 2 1/2  days before I made it  down here. I just wanted to go out  in the field and look for the rupture  on the San Andreas Fault. So in  the  interim geologists  here we went out we went to the  favorite area where we expected  to see rupture and  fault seizures. In 1906 in Woodside  the sent had  been offset. You can see  what the rupture looked like on  the ground and then  this was the rupture from 1906.  The reports that came back in the  field where  we got out there but where's the  rupture.  What they did see. When I finally  got here Tom was sort of  running the show. He said were getting  reports on  Summit Road the Santa  Cruz Mountains. So  we sent people up and this is what  they saw they  saw  crack and this is a map showing  the distribution of these cracks.  But this is not Sandra's fault.  Fault  is this redline.  So what was going  on. In addition to  not seeing fault. Some of the  features  were observed. This  is the  San Andreas Fault service and these  redlines are  the zone of cracking probably  resulting from shaking. And the  rupture at  some point had about 18 or 20  kilometers below the surface to  about seven kilometers  and it didn't get any higher. So  this is one this lined  fault. And Barb McLaughlin and  his  group spent a lot of time out here  this is one of their cross-sections.  You can  see the whole series of stress faults  extending out to the east side of  the  San Andreas. Along those the Montevista  Shannon served a  series  of missile faults.  But this represented  was triggered  so it.  It was shaking that came through  that triggered near  surface movement. The trigger  switch is turning out to be a very  important component and a  hazard analysis. This was an interesting  earthquake completely different  from anything that anybody ask Becton  to have in the Santa  Cruz Mountains. And  this is, all talk a  little bit about Paley of  seismology that we have done since.  That's the study of past earthquakes.  We tried  to develop information  on currents of ruptures  we tried to pull that  information on how much the fault  moved when it slipped in the past.  We do this  largely by  excavating trenches. Mapping the  walls and from this we can  estimate past magnitude and rupture  links. We can use this to  estimate earthquake probabilities  which  I will talk about. We are thinking  seriously not  temporal variations. Over time,  over 200 years in the  region how does  earthquake activity very.  This information  provides sources for the ground  solution that Annamarie will  talk  about. So in 1906 faults  that had been mapped.  It  wasn't until the late 1960s that  the first trenches  were excavated. They were excavated  just  to define where the fault was. Because  that's all that people  were thinking about at the time.  There may have been a few trenches  between 70 and 80 but the real work  started as we approached Loma Prieta  . And  when  I was hired in 1985. They said to  me David,  we want you to work on the San Andreas  Fault and I said that's great but  he said in Southern California.  So Southern California was  the place that had significant earthquakes  and that's where all of this Kaleo  seismic and earthquake  geology work was going. We  did start a little bit of  work then Loma Prieta   occurred.  We can see the number  of locations that have  been developed. There were more  that I didn't put on the slide.  But Loma Prieta  was a catalyst  for changing the direction of where  we were doing a lot of  this work. Focusing here  in northern  California. So to sharing these  examples of what  we do. This is a  trench that has been excavated  right here this  is in Fremont. There were  two  trenches this is extended to Warm  Springs across the fall and this  is some detail of  the trench. You can see these different  layers represent  the positive in the pond  over time. These  redlines are trenches of the fault  and they come up to  a certain horizon.  They are varied as the  fault gets higher and in doing this  we can actually identify  and  date the timing of these past earthquakes.  At this location we found evidence  of 12  earthquakes on the Hayward fault  and then last 1800 or so years.  An average time of  about 150 160 plus or minus  80 years.  The last major earthquake was in  1868. So  this is the type of  work that we do I can show you one  more site this  is the San Andreas, north of San  Francisco. This  is a large wall of  the trench. These are  trenches of the faults coming up  to different levels in the  faults breaking hire. And  this year is the San  Andreas fault. No dinosaurs jumping  out. No  houses falling  this typifies many active faults.  So with this type  of approach we've looked at some  major fault in the  Bay Area and I think we've got  a pretty complete record of  what happened going back to about  1600 A.D..  These are three different models  that we developed from  our observations. In each one  you can see that somewhere  between the late 1600s and  1700s. All  the faults in the Bay  Area had services ruptured on them  and produced some kind of  large earthquake. We tried  to estimate the magnitude and some  people refer to this  as an earthquake cluster  but silent storm is what I  heard mentioned.  And then in the 1800s things were  relatively quiet. 1868  1830 earthquake then we had  1906. So at least for 400 years  you can see there's some variability.  We'd like to go  much further back with this level  of detail.  Just for additional  comparison. This is  the timeline. Showing these  are cracks are as  to to. In  1906 it occurred right here.  You can see in the 56 years before  that there were 39 earthquakes  that were 5 1/2 or larger.  In the 113  years  following, so in 1906 that was  a very important effect. It  released the stress in the entire  region and relax the fault in the  crust and we  can put this  together and again we are seeing  variability in the rate of  earthquakes occurrence. So the  question is. What happens in the  past. One of those things we have  done is to try and model some of  this. We have learned that faults  interact with each other. When  one fault moves it can increase  stress or increase the  stress  and using the tiniest of earthquakes  in  the cluster. When  we started in  1690 with a certain level of stress  the cluster occurred in  1980 all the faults except for the  San Andreas had turned blue meaning  they were relaxed  I had  failed. And at 1900  you can see that  the San Andreas has turned  red hot.  In 2010 the Hayward and  Rogers Creek are heating  up and we  started in 1988 to  do earthquake probabilities. We  broke the fault into  sections  and assigned magnitudes to them.  We calculated 30 year  earthquake probabilities. We needed  to know how much time had elapsed  since the last big  earthquake.  In 1988  we went back to the drawing board  and we  really didn't we added Rogers Creek  and the probability  changed from 50%  to 67%. The most recent  we had was published  in 2015 you  can see we have a much more  complex system that we are working  with and we started with  in 1988.  Is what we had to look forward to.  So just giving you  some thought. There was  a significantly improved general  understanding of the location and  style and rate of rupture on components  of the regions fault system that  had emerged. Questions  remained on  how long ruptures can grow to  be. Longer Kaleo seismic history  can help spread light on this as  well as questions  of  earthquake fluctuating. What  I think we need is a four dimensional  model of the bay area fault system  including all the physics and  time. So I will leave it here and  turned over to Anne-Marie.  Thank you. [  Applause ]  >> Thank you everyone for being here  and thank you David for getting  us acquainted I'm going to talk  a little bit about what happens  once we understand  the earthquakes. So I think about  ground motion. The  ground shaking.  We want to understand because 90%  of the losses were due directly  to ground shaking. So I like to  ask questions on how we can best  model that ground shaking. So here's  a  quick caveat. I was six years old  growing up in New  York State  in 1989. But and watch the World  Series and I really don't care for  baseball [ laughter ] but I'm going  to show you my looking back on  where field what were looking  forward to. This is a reconstruction  of what we think happened based  on  the physics of the earthquake and  the data we  collected. To orient you. This  is  the epicenter of the  fault  that ruptured.  The duration of the actual earthquake  is about 10 seconds.  The waves continued  to propagate for quite  a while. The oranges  the Reds and the dark colors are  stronger  shaking intensity. Takes a few  seconds. Then you  can see another earthquake has stopped  occurring  but the waves are traveling away  from  the epicenter. You can see quite  a few observations here. You  can see heterogeneous pattern  with these large  ground motions. Is the waves continued you can  see as Tom mentioned, all around  the bay there are  these man-made fields where we  see application of ground  motion. We didn't have those  capabilities at the time but it  gave us an idea of what happened  during the event.  This was developed after  the  fact. This was a service  snapshot of the  largest ground motion  that occurred. You can see the same  sort of features we have large  ground motion here but also in  these other locations  and along the bay. So in  particular as Tom mentioned there  was  strong shaking on  the east side of the Bay  Bridge about 60 miles away. It  was unusual to have such  large damage. This shows the actual  ground motion recorded at the site.  This  is the bedrock at the viaduct and  you can see these fairly  small motions a  much larger ground motion. These  amplified sections here caused the  structure to fall.  We see these large ground  motions in the basin structure you  can get increased  shaking  and duration.  This is very different than if  you do you know  sugar out.  We collected love these observations  and solidify the understanding of  the heterogeneity effect of these  soft soils.  This  is  USGS paver product I think of this  as the times the population and  the vulnerability came. It gives  an estimated model impact of the  tallies and economic loss.  This  is one way we can get a quick estimate  of the impact of  the earthquake. We didn't have a  lot of information about what  happened in Santa Cruz and we didn't  have cell service  or Internet. 
     You know we didn't have these products  or information so rapidly but now  it's available in just a  few minutes. So looking forward  in the future we continue to develop  products putting in  better models  to get a more accurate estimate  of all  these numbers.  >>I will  talk about is  developing models. We sort  of know this is what  happened it's a reconstruction of  what happened during the earthquake  and it captures  the heterogeneity. This is one single  earthquake we can treat as our observation.  What we want to be able to do is  take that information that  David presented and  then estimate what the ground motion  would be for a future earthquake  that has not occurred. So we  are  not predicting occurrence but once  we understand that  can happen. It's based on all  of our collections  of observations for  ground shaking for all magnitudes  of earthquakes. These are  our observations. This  would be the predicted  shaking intensity we've simply  got this one curve  that looks like this  epicenter for strong  ground shaking. That certainly does  not reflect what  we actually observed  which  is first modern  functional GMP it hasn't  changed much we are still dealing  with basically  a  single curve. We can put that complex  soil structure in there but what  we like to  do is be location specific for  individualized models that can account  for the  spatial heterogeneity. 's were going to get you  a more accurate description of  that. So taking  that ground model and combining  it with  thought probability. We can  combine that to go. So  if there's an earthquake that  is on the Hayward  fault and if we do that  for every earthquake that has a  recurrence rate we get a map  looks  like this.  This is a  2% chance this is  the shaking intensity for over  50 years. In 1976 this was the  first shake map  that  David described. Combining  use in building codes with logic  taken into [  Indiscernible ] In the future  we are moving more towards the  specific path  site. That information  is also put into building codes.  It's improved  by incorporating information about  these  soft soil sites. Current  building  codes do a good job. What we  are looking for in the future is  to improve our resilience to  minimizing impact. Thinking about  this  sort of  timeline going from operational  meaning you can we inhabit  it today. Thinking about this  as more of the spectrum performance  base design thinking our building  should perform under certain  conditions. So in  the hospital do you really want  to prevent any kind of damage  at all. But for Joe's bar you can  have a little bit of  damage. So I  will touch base  on this. We can quickly  detect this earthquake with our  sensors and get ahead of  those waves. This gives  us some lead time before the shaking  arise that  your location. So relearning everyone  as fast and accurately as possible.  So how much  ground shaking us  at  one location. We  had the first operational temporary  early warning system. This is  a map showing the central  region  that  occurred. We knew that we could  expect stronger ground motion at  that location,  receivers know the  source new they would  detect earthquakes . Back to the  current project. Project in  development for Oregon  and Washington there's already some  apps in  Los Angeles. Looking  at automated this depends on your level of ground  motion are  interested in. 
     I'd like to see some  specific alerts for  every location. 
     We  knew what we were expecting the  earthquake we were able to  build this in a  specific system. [  Applause ] See Jessica Murray talking  about how you think the  data in earthquake hazard assessments  for  quick responses. You're probably  already familiar with at least some  forms of GPS all talk a  lot about that as well. I'm going  to start  by saying where this data is important  for the context here. The first  of which is in  her seismic strain accumulation  and how we  use that, were using the  boundary here where the  North Americans plates pass each  other. Although the plates are always  moving we don't always have motion  at the  plate boundary. It only happens  occasionally so for either side  of  the fault on each side of the diagram.  It can be  worked. Because plate is cooling  on it but there's no  motion happening. We can measure  that strain in the earth crust  with technique, it's important to  note that if  we want to understand  what the probability of  earthquakes is, we've already seen  this image a couple of times. Tonight  it shows what the probabilities  are in  the area. Another thing that is  exciting is to  measure motion  that  is what seismometers are blind  to. A lot of the faults in  this area a lot of this  at a slow rate can be measured and  we can observe it visually if we  wait long  enough. So a series of photographs  has occurred  in Hayward the fault  over time. We can cause visible  tracking and pavement because some  of  those stress to the plate  motions actually believe a slow  steady motion.  Is a hazard assessment. Another  example of those  are observable  to techniques so continued motions  after an earthquake happens.  These photographs show  what happens in the  24 hours following  that event. This continues  to show motion. That's  problematic  for immediate  reconstruction efforts. To continue  to move ahead to keep  coming back it's difficult for homeowners  who want to start repairing the  property.  And finally it enables us  to measure the  motions on the fault,  the larger the clip the larger the  magnitude of the earthquake. Would  like to have an accurate estimate  for number  of purposes.  [  Indiscernible ] It can help with  proper  magnitude estimate on the fault.  So at the time of Loma Prieta ,  what  did  we do, we didn't  really have  it yet they had really pioneered  a program of bilateral  ration measurements 
     what's involved. Basically were  able to measure the  distances between stations that were being measured  and the data was  very precise. There were  some drawbacks  for change.  If you can get an idea from this  picture is fairly  labor intensive. As I mentioned  a moment ago GPS was  just getting started in  the mid-1980s. Those  were actually  very useful it's what  was collected after Loma Prieta  
     . So 
     were  generally speaking it's become a  major source  of data 
     -- because  the data is continuously recorded  you cannot  only measure what was measured  earlier. But what happens when earthquake  occurs. This time variable motion  that can continue after  an earthquake. One  example of a data set that is  derived is this is for  the bay area. This shows  how  fast that moved into the stable  interior of the continent. The  blue ones are from the permanent  stations, the green ones  are for where we make occasional  measurements. Relative to the interior  of North  America increasingly fast it's  moving to get closer to the  Pacific plates. This kind  of information was the  first one to formally and  indirectly include data that was  derived from genetic observations.  I wanted to  touch upon some technical advances.  These are imaging techniques  that are now very important for  the type of work that we do. There's  two I mentioned here.  In most cases  they are using sensors to transmit  and receive radar or reflect light  to measure  the topography. If you do  that repeatedly for the same location  you can detect changes that have  occurred during two  observation times. Changes would  be ground movement due to  earthquakes or a post  seismic motion. On the left is an  example of the  recent Ridgecrest earthquake. What  your seeing  here is the movement of the ground surface  due to that earthquake.  Each cycle color represents  14 centimeters. The direction from  the ground to the satellite on which  the sensor was  mounted in. It's something  new and benefiting us  a lot. You can get more information  about time  varying processes. This'll be the  first U.S. participation in  the satellite mission for those  purposes. And on the  right here is an image that shows  an example of data.  It can  be either airborne or ground-based  in this case is from a ground-based  sensor. It's showing you  a bridge that you can see here in  the image that's  built across the San Andreas  fault. Can  see that the fault closer to us bridge  is this green color. On the opposite  side you can see the color  change. It's showing that there's  actually movement along  the fault. 
     I mentioned this data can  be collected in airborne  or ground-based platform. Also in  a  mobile platform. Mobile laser  scanning abilities in this photo  from the recent Ridgecrest quake  is here. Going out to take images  of the fault rupture  and measuring the continued motion  there. This last advance I want  to talk  about is  real time  measurement when I talk about high  rate in the  content of that it's one of five  samples  per second this enables us to directly measure  ground incitement during  an earthquake. When this data  is available in real time that can  be a very powerful tool. So this  shows an example of data  from a high  rate system . It's happening in the halls to  orient you  in this  region here. The traces at the bottom  and from the station closest to  the rupture. And if you go to  the  top moving further away you can  see that the signal  arrived later in  the smaller and tiered of stations.  We can use this as a  functioning distance for what  Anne-Marie described using ground  motions to  infer  what the ground weapon  is. We know that for very large  earthquakes from seismic data alone  in real time it  can  be problematic. The  magnitude that's  becoming  quite useful another  way that we can use  that is estimating all  the  faults that's happening. This figure  is from a retrospective study from  the 2011  earthquake in Japan which many  of you remember caused a  very devastating salami. For each  of these it shows GPS station and  how it moved in the colored shading  background is the  estimated clip from the  data at a certain time  this case it's 180 seconds after  the earthquake started. On  the right you're seeing the  magnitude estimates compared to  what is determined from the  seismic data. So this is showing  the  estimates from the data. You can  see that this data gives you a  more accurate estimate. We can do  that rapidly enough and it becomes  very important for improved  early warning.  This is something we are currently  testing for  possible inclusion. I will leave it at this in turn  it over  to Tom.  >> Let's go back to cruising altitude  for a couple minutes then open it  up  for questions. 
     So how far  have we come in 30 years.  Just  summarizing briefly ,  the scientific advances in the 30  years since Loma Prieta  and many  of them  triggered by the Loma Prieta  have been pretty  impressive. We  learned a lot more about  Hall full behave and interact. I  think it's improved  our ability  estimates the  probability. At the same time it's  turning out there's a lot of complexity  to it. So we probably  won't come out with  a precise probability will come  up with the uncertainty  or range. We've looked at it  by doing a better job of predict  ground shake. It's a complex subject  when you get down to the details.  You need  better tools. We are getting there  in part of what because what  was created,  the technological advances  are very impressive in terms of  helping us  locate ourselves better in both  the time  and space we can make measurements  that we couldn't even dream of at  the time of Loma Prieta .  And as  our speakers pointed out these end  up in  USGS products. Loma Prieta   served  a very critical role as  I mentioned in the beginning in  terms of alerting us to the earthquake  hazards in the bay area. It made  it very real and led to very  practical things. Construction  practices gotten a lot better and  we are no zoning land in  terms of the hazards and coming  back to this you just drive  around and look at home  being built these days.  The amount of steel that goes into  a residence, is  very impressive. And they wanted  to close on this note because I  think it's an important change in  the way we are approaching natural  disasters and how they reduce the  impact of that. The conversation  started with Loma Prieta .  At the  time  of that we tried  to build  we want our community to be able  to recover quickly from  the earthquake. There's  hundreds of seniors in this  housing in you've  always noticed it  really creates a  crisis situation. Improving the seismic resistance  of  the building. So  chatting house a large underground  parking garage. , This basically stops the  shaking from going into  the building. It's  been tested in several earthquakes,  you can actually  see evidence from that if you happen  to walk by the charming house  take  a peek. What you see is  this steel  freight. Covers an open trench.  When you stop to think  about it with that underground parking  structure there's  still walls and supporting columns  touching that  external soil. 
     The ground can move but it's not  going to come in contact with  the building. Palo Alto  utility system provides that piece  of the resilient  community but nonetheless what it  does is  it takes home to hundreds  of  seniors, albeit a problem that one  would have if one didn't do this.  And I wanted to close  this slide. It summarizes what Dave  was showing  you earlier and put some probabilities  on the major fault in  the area. You can see the total  probability for the bay area is  about  72%, that's a  pretty high probability it certainly  drives the resilience effort you'll  see in the  bay area. With that I think you  for  your attention. I  forgot to note that this  week or will be talking about sea  level rise and  coastal erosion. How bad can it  possibly be. A nice change  of topic.  >> [ Indiscernible - Low  Volume ] maybe what you  can do [  Indiscernible ]  >>Any  other questions? Anybody who worked in the city  and lived down here had a  problem. Difficult  to see. The  transportation system.  >>People work here and live  up there. After Loma Prieta  it prompted the  largest investment  of  any earthquake in California  Street. So probably our transportation  system is pretty robust. It's just  going  to be the problem of moving a lot  of people in a  long time. Foster City as far as  I know it's a different kind  of soil. Most  of that occurs where they  don't stand into the bay. It's usually  pretty old Phil  area. And the other place  of course. Channels. San  Jose has  several creeks. The Foster City  question comes up quite a  bit. Because it's a clay soil it's  not prone to look  at vacation. >>I wonder how  were doing at the  public infrastructure. Our all of  our bridges  >> And know  the  large utilities  invested  large efforts  responses into the system. I  think Caltrans  rated about 20 years for  the  money dedicated  for  retrofitting. I think with the public  infrastructure we are doing  fairly well. But having said that  if still got  vulnerable areas. >>Thank you enjoyed your presentation  very much. Having grown up  in California and  experiencing earthquakes. My question  is the early warning  system. New have 10 or  20 seconds  to react. Getting an indication  of what to do  you know my  question is  public education in terms of what  they need to do in that short amount  of time. >> Certainly public education is  a large component of  that project. So that  need is recognized  and as the system rolls out to the  public that's going to be  a part  of it. I think the  thing is to be prepared to matter  where you are. For most  of us, you know were not always  in the hospital setting world in  one place in out your house, your  office, your gym, or your car. So  think about like right now where  you  would go. Stop drop and cover hold  on just stop and wait for  the shaking. If it's still shaking  without warning then you should  get out.  >>Of our seismologists  gave it talk and she pointed out  you don't need a cell phone  for  early warning. It's going to be  larger than anything you probably  felt before. You get  the alert then the shaking comes  after that. I think it's  important because of where we live.  Earthquakes are to be  expect to. 
     What is the  ground shaking.  Moderate immediate. The acceleration of the ground  and how that's  actually  measured. What  with  the missing from  your research.  >> We  typically measure velocity. We've shown  a lot of the projects of  intensity and we found that  that combination,  best represents  people perceive shaking  and we determined  that  velocity matches.  >> Would like to  have measurements everywhere all  the time which unfortunately cannot  do. But one  thing you may remember  having seen in a obviously there  were fault offshore  under the ocean. We like to be able  to  get data  there. For warning and whether  it be earthquake. The  Pacific Northwest is a place where  there's  huge hazards. We can get more observation offshore  and get  more insight. It's  really exciting extremely expensive  if I had a lot of money I would  put the money  towards that. >> I see an imaginary  horizontal line, that was one the  major effort 
     in can  design pipes that  allow movement and that sort  of thing. Investing a fair amount  of money. I think  the major , >> [ Indiscernible - Low  Volume ] moving  in Europe and the  earthquake hit. My question for  David is on  the one diagram we are  the multiple thresholds piled  up to the right of the  local data. There is curious  when you said they all started shaking.  With that immediately when the fault  it or  was it >> Being clear , to  growing up [  Indiscernible ] That's  an unsubstantiated observation.  The  question is talking about a  warning and things of  that nature. What you're really  asking is [ Indiscernible ] And  the answer  is no. Looking at  temperature and humidity and all  those things. They  can't  predict earthquakes. It  can be more sensitive to the  hyper frequency. If  there's a tiny  earthquake know a dog could have  felt those waves  coming in maybe you were jogging  or washing the dishes or watching  something good on TV and you weren't  paying attention. >>People  have actually  address that due to changes in the  hydrology. So  it's not completely off-the-wall  to think about that.  It  not something >> I  understand that , some  amazing studies you can take this and study  it over the years. Contact  the local  community in terms of letting them  know that your  doing that. So forth and so  on.  >>There's different ways  to measure that. It's  very traumatic when you go to  the curve partners read part  is clear it's disheartening to  go back.  In the and the fault  always wins.  I think if  you want downtown that's sort  of a nice urban  setting for it  to see the fault. You  can see where it  forms buildings and words  pulled buildings apart. You can  see where the sidewalks used to  be offset. The city  has a program of trying to  repair things.  We asked if they'd reset that  famous curve they said they  were sorry they should've asked  about it this was  there plan. I think it goes with  the territory.  The critical places to  measure don't have to be we don't  really need that  curve offset. This much force Isaac  measurement and other places  >>That being said. If you  remember back especially  looking back  over decades. The places where he  made measurements were ones  that were specific.  It's important to have  the longevity of measurements over  a decade. It's unfortunate when we lose the  data points. It's a  permanently involved station with  sensing  around it.  It's an  ongoing  problem for construction  or any number  of reasons. >> Thanks you did a good job of explaining  race  new technologies. My question is,  have  not had  any  effect  on accuracy. For insurance, the California  earthquake Authority actually uses  the models that Dave  will  show. So whatever the  survey issues were for these  probabilities will update the insurance  to reflect those estimates.  That's why one of the important  pieces of this data. And  I  don't think .  >>The most recent  study called  user 3.  That was funded to a  very  large degree. They used the probability  and all the  data calculating and  reevaluating insurance rates  for  specific different 
     companies. The commercial  sector is taken care of by  those companies. There are going  to be  adopting  that science. This stuff gets used  it to quickly,  probably faster than it  does in the  code  business.  >>Same question. What do you see  as an impact for early  development land-use. I will make an  observation from Loma Prieta . I  was surprised when  walking the  streets  this happened I remember talking  to a realtor about the impact of  housing prices. They couldn't detect  an impact of the  earthquake. We were in a little  bit of  a recession. My  guess is it's such a  small consideration in  the purpose that it just  doesn't impact housing values.  The added  cost of making a  home stronger, that  does  add a few percents to the construction  cost. So that way it  will have an effect. The hazard  itself doesn't seem to have that  much impact.  >> I  think the state zones them as you  have to do a special study  in  that area. I think they're  old enough to  logic departments. >>There's almost no mention of  fire hazard [  Indiscernible ] Are they doing  any studies for fire prevention  after  an earthquake?  >>I'm not  sure which chapter that is.  The  issue is huge it was  carefully looked at and there were  some very stunning  statistics developed. I don't know  where that particular  chapter is in  terms  of circulation. >> It depends on the time of the day  and the time of wins. Of course  the 1906 earthquake there was  a lot of  fire damage more damaging than the  earthquake itself. It's a very  important situation taken  into account more  resilience like  firehouse doors  opening. These are the kind of things  that are being done.  It's discussed in  that scenario.  >> And lived in  a  Hild area, 
     in the 89 earthquake  I was  home watching a ballgame. I look  out my  window and I  can see  this huge kind of pine  trees that looks like it's coming  up out of the ground.  This huge oak tree  it's the same thing but  wasn't quite  as tall. The daughter had a bunch of  trophies and these fell down but  these remained  standing. So is that shaking or  I mean  is it  I mean.  >> The shaking way of coming through.  They  have different waves that come in  this way it could've affected, it's  very common to be  completely shattered. Uprooted and  turned over .  >>Was in Tokyo when  the quake hit. We didn't  feel much but my question is how  much to interact with colleagues  around  the world and how well are we prepared  versus the Japanese or  other personal countries with  earthquake issues.  >>One  of the nice  things so do  interact internationally. But the  second part of  your questions .  >>I will make a comment from  the  perspective there's some  very extensive  work. For earlier that we  did  here. They really put  priority  on getting the collections that  they could. I would  say where. during the  so good.  >>Our building  codes similar.  >>Acting we  communicate with foreign colleagues  at  all levels. I know in my  career I did a tremendous amount  of work with the  Italian equivalent of the  geological  survey. With the Mongolian  equivalents and working together in  different places was one of the  best things that we could  be doing. It's absolutely critical  to improving our science  and our work  for you. That often the engineers have  their own  design bias.  In Chile they tend  to build  different [ Indiscernible - Low  Volume ] we learn from each  other. There's a lot of  exchange and it was interesting  this approach you've  got sometimes not so subtle differences  in  terms.  >>First we've got to switch to  metric [  laughter ] you mentioned your network  capturing  your radar. Are there  any other subjects and stay on the  ground or  in orbit I don't have a lot of information  and if  I'd did, going through a  lot of this GSS data with the  leaders getting data on  the  ground. The other time we have a  major earthquake it's an opportunity  to  get unprecedented to use  that data. There's  very high in detail it gives us  new insight it's very  important for us to understand the  hazard when you go back to look  at earlier ruptures you want  to understand what  this represents >> I heard that  this never  really happened.  So we did have an  earthquake there in 2004 that  was after a lot  of the instrumentation had been  put in we were  able to answer some of the questions  that we posed leading up  to that.  That in and of  itself is interesting because a  lot of the models pointed to the  similarities among the previous  magnitude of the  earthquake there  for  understanding earthquakes. We  continue to have  orientation there. >>Maybe a  couple hundred. The  ones that you see.  There's a  few examples >>It's not the  most  typical way there's other ways  of measuring having  to distribute a lot  of  the areas. .  >> What were  the marshmallows made  out of.  >>It's  a  polymer made for  base isolators. It can be huge depends on  the structure. >>I look at those maps sometime  and it seems like there's one  every day. >> One of the key points for establishing  long-term horizontal movement on  the San Andreas. The  other half is  the volcanic's down by the  great fine  area. And this  was  my volcanic complex that  got split by the San Andreas. So  you've got the San Andreas it's  self just a kilometer or  so from the east and as usual there's  more than one  fault there. So  no  problem accounting for that. There  will  be others.  Well thank  you all [  Applause ] [ Event  concluded ]