PubTalk 5/2019 - Rodgers Creek Fault

Video Transcript
Download Video
Right-click and save to download

Detailed Description

Title: New Mapping of the Rodgers Creek Fault: It's longer and more complex than we thought

  • Remote sensing technology reveals surface traces of the fault covered by trees and buildings.
  • Greater fault length and linkages to neighboring faults make larger earthquakes possible. 
  • Fault complexity has implications for hazards from ground shaking and surface fault rupture. 


Image Dimensions: 926 x 604

Date Taken:

Length: 01:04:35

Location Taken: Menlo Park, CA, US

Video Credits

Amelia Redhill -


>> Good  evening and welcome to the  May public evening lecture . Before I say anything else, next June will be held June 27. There  is no title yet. It will be on USGS   wildland  research. 


At my name is David Schwartz.  Over the years you may have seen  me up here giving a public lecture . In October, of 2018, after 33  years here I decided  commuting from Danville to Menlo  Park was no longer viable . I retired. Put  that in quotations. I'm an emeritus scientist. I'm  involved with ongoing projects it sort of like the  Eagles Hotel California. You can  check at any time you like you can  never leave. Tonight, I'm very  pleased to be able to introduce Suzanne Hecker. An earth  quake  geologist, she loves studying at the faults and will hear about a lot of the Bay Area active faults tonight.  She graduated from Bucknell University

     in Pennsylvania. She got her masters  at the University of Arizona where  she studied the 1954 Dixie Valley Fairview peak earthquake  rupture in Nevada.

     I had the good fortune to meet her in 1988 which was working for  the Utah geological survey. In 1992, we extracted her  from the Utah survey and hired her  here at the USGS. Tonight,  she will  talk about the Rodgers Creek fault.  Before I allow her to come up and  start speaking, I would like to  put Rodgers Creek into  historical  perspective. If you  were living in the Bay Area at the  time of  the 1906 earthquake and looked at  a geologic map,  you would see faults on the map.  The San Andreas Fault was known. Part of the fault was on the  map. The Hayward fault was mapped. The northern Calaveras which was called the Sonoma fault was on the map.

     The Greenville, the Concord and  the Rogers Creek had yet to be identified. In 1908, a very well-known USGS geologist hired a Berkeley professor

     named Harry over. I asked him to develop a  fault map for northern California. We went  to work on that. In 1960 he published  a paper in the bulletin of the seismological  Society of America. It was called  California earthquakes, synthetic  study of recorded shocks. What would tried to do was take what  was known about historical seismicity and associated with known or  suspected faults. He was very bright  and knew that slip on faults  produced earthquakes.



There are two sections out of  the paper. Would like to read them  to you. The  Hayward fault stretches south easterly from San Pablo Bay along the western margin of the  hills, East of Berkeley and on for  distance of 50 miles Shirley , possibly 100 miles or more. It  is probable that it may extend northwesterly beyond San Pablo  Bay. It's complete extent  has not been worked out. There is  a footnote number 10. Work in  the field in Sonoma and  Mendocino counties , has brought to  light abundant evidence of youthful faults trace phenomena lying in a line along a  natural extension in this direction. These features undoubtedly expend extend many miles , their course has yet to be  completely delineated.



This is part of the plate on  the paper.



You see this line going up  through Santa Rosa. There was a linear band  of earthquakes that extend from  the Eureka area down to San Pablo  Bay. Along this band, there is a vault  or group as yet  not discovered. He considered that active  faults and he drew this line. He called it  the Eureka, Ukiah, San  Pablo line. It was where he suspected active faults exists.  This line , basically defines

     our currently known locations of  Rogers Creek and the faults. He was very prescient, very intuitive . He took a small amount of information,  put together a bigger picture. Now,  Hunter three years later, he will  have the opportunity to hear from  Suzanne about what we have learned  regarding the  Rodgers Creek fault.  [ Applause ]  >>> Thank you. Can everyone hear me,  in the back? And we should turn  down the house lights.  David's introduction may be a difficult  act to follow.  What I will do, I will start off  by introducing the Rodgers Creek fault and get  background information  and context as to why improved  Mac in mapping is important  for accurate for accurate characterization . 


Before I go further, I will set  my pointer.



This is an earthquake forecast  map showing no active faults in  the area. The major branches  of the faults the system are shown  in different colors of the lesser-known  smaller faults are in yellow. The  Rogers Creek  fault extends from San Pablo Bay through Santa Rosa to North  of Healdsburg. Healdsburg is right  about here. Analyses have indicated that  the combined Rodgers Creek  fault  branch has the highest likelihood , 33% chance  of producing a large regionally  damaging earthquake, magnitude 6.7  or greater by the year  2043. There's a 72% chance of one  or more of these earthquakes happening  anywhere in the region.



Here is a simplified map of late  boundary showing the direction of  motion between Pacific and North  American tectonic plates. The Pacific  plate is moving northwestward

     slowly and continually at a rate  of 1.5 inches per year past the  North American plate. Quarter of  this motion is being stored  and occasionally released on the  Hayward and Rogers Creek fault.  As a results, the dominant sense of movement across the fault that make up the plate boundary  is right lateral strike slip.

     These block diagrams illustrate  this kind of movement. T-1 would  be the time before an earthquake,  you can think of right lateral strike  slip you're standing on one side  of the fault looking across the  other side, when an earthquake happens,  the other side will moves your right.



Fortunately, only large earthquakes  are energetic enough to rupture  up to the earth surface.



This is illustrative by this  pair of locked diagrams. Small earthquake  rupture is remain buried in the  subsurface whereas in a large earthquake  rupture can grow to encompass

     pretty much the entire width of  the fault plane and breakup to the  earth surface. Where the fault meets  the surface of the earth, this is  called the faults trace, the trace of  the fall and this is what we map  when we map active faults. 


How large is large? What size  earthquakes can produce surface  rupture? This is a plot of historical earthquakes  worldwide that have ruptured to  the surface showing the relation  between the length of rupture and  the magnitude of the earthquake.  The symbols represent types of faults movements. IRD  told you about strike slip which  is a horizontal or lateral movement , reverse the normal faults are  different kinds of vertical slip. 


 You see there is a clear trend that  longer ruptures produce larger earthquakes.  Although there is some scatter.  In the smallest earthquakes that  produce surface rupture are about  magnitude 5.5 to 6.? Of  the size and larger can  happen on certain kinds of faults  and not write to the surface. This is true  of some reverse faults in certain  settings. 


The longest rupture in the catalog is right here. This is a 1906  magnitude 7.9 earthquake. It  was almost 300 miles long.  The photo in the upper is of a fence near  Bolinas offset 8.5 feet by  the earthquake. In the lower right, there is an  image of a woman standing next to  the 1906 rupture. I want to point  out that on this hillside, the fault  rupture follows a pre-existing bench or slight reduction in  the slope. 


This is a scar  in the landscape and results from recurrent fault ruptures over the course of thousands of  years. This is the kind of landform  feature that geologist look for  to map active faults. 


It was once conventional  wisdom that the length of an individual  fault constrains how long an earthquake  rupture can grow to be . We know larger earthquakes can  occur on a single fault , like on the San Andreas Fault  we have come to the can also occur  by breaking across multiple faults.  Here is an example of a complex multi-fault rupture  that occurred in the Mojave Desert

     in 1992. It's one of the first?  It really made scientists aware  earthquakes are not always confined  to a single fault. 


This earthquake broke portions  of five faults and most had  been mapped before the earthquake  but there was an important connecting  fault right there called the Kickapoo fault or lenders  fault that was not identified before  the earthquake. Later, geologic  studies found evidence that it had  a history of surface rupture. Just hadn't  been noticed before. 


 The Kickapoo fault allow the earthquake  which started at the south end of  the Johnson Valley fault to propagate across of  the Homestead Valley fault and from  there onto several other fault allowing  the rupture to grow to 50 miles  in length. In comparison, the Johnson  Valley fault is only 35  miles. 50  miles long and the Johnson Valley  fault is 35 miles long including  a portion of that that didn't  rupture in this earthquake. 


Here's a photo of  the house  builds on the Kickapoo fault and  it was damaged by the 1992 rupture.  It illustrates one  reason why it's important to know  where active faults are located  before an earth take earthquake  occurs. 


And earlier

     earthquake, the 1970 magnitude 6.6  San Fernando earthquake near Los  Angeles had extensive surface rupture the damaged numerous  homes and commercial buildings built  across the faults. This earthquake  led to passage of an earthquake zoning act named for two state senators who  introduced the legislation. The  law prohibits the building of structures , most for human occupancy  across traces of active faults.

     This led to a program of active  faults mapping by the California  geologic survey order to establish  regulatory zone. Geologic  investigations have to be done before  building is built to make sure  it won't be constructed across an  active fall. And active fault is defined for  this purpose as having ruptured,  usually repeatedly in the last 11,000  years which is the time period  of the most recent geologic  epic.  >> Here's another example of a complex rupture,  one that occurred not long ago,  2014 magnitude 6 South never earthquake. Some of these strands that  ruptured had not been mapped  and none had been zoned as active.  It was after the fact that the states in 2018 the state established  regulatory zones along these ruptures. A little late  for that earthquake. The problem  is these smaller faults produce  smaller displacements at the surface. They may move less frequently so  they are less well expressed  in landscape and therefore are harder  to map. 


The largest displacement in this  earthquake was 46 centimeters. At 18  inches and that's only one 20th  of the largest slip that occurs  on the 1906 earthquake rupture. 


 Now we will take a look at how recent  advances in remote sensing technology  have made possible higher resolution  faults mapping. 


First I  will give you a short introduction  into the type of landform features the geomorphology that forms a  lot active strikes . It's illustrated here by this  classic stretch of the San Andreas  Fault an aerial view. This is in the Caruso plane in Southern  California. The vault here is defined  as a sharp line and marked by faults  scarp which are  steps in the landscape. Also my  offsets of drainage channels . In particular, the largest channel  here is Wallace Creek. You can see  where Wallace Creek takes a sharp  right hand and is offset , 430 feet along the fault . When the channel 1st cut , it flowed straight across the fault so has accumulated this  offset over the course of many large  earthquakes that happen about once  every few hundred years. Here is older abandoned  channel offset more than twice as  far. 


Other features common along strike  set faults shown in this diagram  are benches which I pointed out in the photo  the 1906 earthquake, linear valleys,  linear ridges, scarp, sag poems which  are water-filled depressions. Historically, aerial photographs have been  the basis for mapping. Overlapping  photography provides a shift in  perspective that enables depth perception . This method works well for the  ground surface is visible from above  but less so were vegetation obscures  the grounds. Unfortunately, in the last couple  decades, there is a new surveying  method that has become available  that overcomes this problem.  That is a Lidar, light detection  and ranging  is a three  dimensional wagers laser scanning method used  to make higher resolution topographic  maps. It doesn't depend on the clear  view of the grounds. This technology  involves sending pulse laser light  from the source mounted on an aircraft.  You measure the reflected pulse  with a sensor. Differences in  laser return wavelengths and times give information on distance to the objects below. 


Some of the laser light  almost always reaches the ground even where you have thick vegetation.  It reflects back up to  the sensor. These last returns can be isolated and used to make  digital models of topography.  >> The survey of the Bay Area major  faults was flown in 2007 . Here is the footprint of the survey  flown along the Rogers Creek faults. 


 A higher resolution, higher density  survey was flown for all of Sonoma  County in 2013 and it was this data  set that I principally used for  the new mapping.  >> Here is an example from the northern  San Andreas Fault of how it can  be used to see many dance forest  canopy. On the left is  aerial photograph and on the right  is image produced from the model

     of the full data sets. What you're  seeing on the tops of trees. You can filter the point cloud  to remove all but the ground returns  from essentially removing the trees  and leaving a bare earth image. Even though this is an older  vintage data set, it's not the highest  of resolutions, you can clearly  see strong liniment a defined surface  trace of the San Andreas.  >> Let's consider  the Rodgers Creek fault  in context  of rupture complexity. Many of the  smaller faults in the region shown  in yellow lie close to the major  faults but little is known about  how these interact. In particular,  fault number 4 right there called the Bennett Valley fault  lies less than two miles east of  the Rogers Creek faults. It extends  to the north and projects toward  the my Acoma fault which  is another major plate boundary  and it lies about five miles East and overlaps much. 


We will look later  at how these may interact in their mapped relationships. 


I want to point out an important  example of interlinking faults and  that is between the Rogers Creek  in Hayward faults. It had been suspected these two faults might be connected  the need San Pablo Bay and this  possibility was accounted for in  hazard models but only recently  was direct evidence found of faulting beneath the bay. 


This is work

     done by USGS scientist Janet Watt  and her colleagues. Janet gave a  great public lecture on this about  a year ago so I encourage you to  look for that online. 


For this study they used a seismic  reflection method image shallow  sediments layers beneath the ocean  floor of San Pablo Bay. The method  illustrated with that cartoon on the left  involves a meeting high-frequency  acoustic waves from boat towed source  and recording the signal that reflects  off of layers of sediment that have  contrasting acoustic impedance properties.  >> The survey track lines are shown  in yellow on this map and the day  they were processed to produce two  dimensional cross-sections from  which researchers  could examine and look for evidence  where the layers of sediment have  been broken and offset. In three examples are shown on  the right. 


Places where they identified  evidence of faulting our mapped  by exes on this map. If you connect the dots , you  can map the Hayward fault were trends  beneath the bay. To the  north end, it bends slightly to the north  and projects directly into a branch  of the Creek faults that has been  mapped as active using  a Lidar. We also found evidence  of active faulting  in the Sears  point area on the Sonoma Raceway.  This provides linking structure  between the Rogers Creek and extension of the Hayward  faults beneath the bay. 


Now we  know these appear  to be connected. The question is,  how they produced earthquakes together , a single earthquake in  the past.



To evaluate this possibility,  we need to turn to the science of Haley oh seismology. This is the  geologic study of large prehistoric  earthquakes. Geologists X a great excavate trenches to uncover evidence of  old earthquakes. This example shown  here on the left on  the Hayward fault. We tend to target sites that have active fine-grained  deposition because these kinds of  sites have the best opportunity  of preserving continuous record  of service halting. 


Here's an example  of the Rodgers Creek fault. This is a  log what we mapped out the relationships  . What we do is strive  to identify the ground surface present  at the time of an earthquake. An example here is this redline  which represents the ground surface  that was present at the time  of the most recent large earthquake  on the fault. Then we collect material from sediments beneath  that horizon and above in  order to bracket the time of the  earthquake. In this particular trench  we used radiocarbon dating of charcoal  to date the sediment below the earthquake event horizon. Then  we use the arrival  of non-native pollen associated  with European settlement in the  1800s to date above this horizon. 


 It turns out the timing of the most  recent earthquake on the Rodgers  Creek Fault is similar  to the timing  of earthquakes on the Hayward fault allowing for the possibility  of combined rupture sometime between  1715 and 1776 which is the  beginning of the historical period. Rupture that involves both  faults could reduce an earthquake  as large as magnitude 7.5.  However, because of the inherent  uncertainties in the dating methods  we cannot know for sure whether  these rupture together. If they occurred  a separate earthquake, they probably  occurred within a few decades of  one another which also has implications  for earthquake hazard in the region. 


Now we will take a look at the  new mapping of the Rodgers Creek  Fault.  Using a Lidar based image of identified  geomorphic landform evidence of  active faulting along the entire  Rodgers Creek Fault including the  entire  Healdsburg section  north of Santa Rosa. This increases the known length  of active faulting by 10  miles for a total fault length of  at least 45 miles. This increases  the length-based estimates of magnitude on the fault by about a 10th  of that magnitude units. More importantly,  especially for the residents of  Healdsburg increases the proximity  of strong shaking and brings service  faulting through the town. 


We found the zone of  faulting is also broader and more  complex than previously known.  Faults strands outside the main  zone tend to be less well expressed  and so we categorize these as part  of the distributed sound.  We showed those in Orange. Faults  strands whose origins are uncertain  are in yellow. Some of these could  be landslide related or bedrock  features. 


For the first  time we have been able to map the  active trace of the through central  Santa Rosa. We will take a look  at that in a few minutes. 


As mentioned already, is evidence  of active faulting that branches  to the south at the south end of the Rogers  Creek fault appears to connect with  faulting at Sears point and with  the continuation of the Hayward  fault beneath San Pablo Bay.  >> Here is a shaded relief image showing  the main trace of the fault and  how it relates to topography. Most  of the fault is associated with  positive relief, hilly and mountainous  train for  this is evidence of compressional  forces caused by a slight misalignment  of the fault with respect to plate  boundary motion. Some of the relief  is locally generated by bends in  the faults. Notice the topographic low as  Santa Rosa where it takes a right  bend. I will show you this diagram  that shows how it  topography is created on a right  lateral strike slip at bends in the faults. At a right  bend you get release and as result  you get subsidence . At left, these  are restraining and you get crowding and uplift.  An example of the left bend is  at Sonoma Mountain, here you can  see there is a broad left bend and  here you have the mountain forward. Another example  is toward the north end here, again  you have another left bend.  >> Before I leave this slide I want  to point out the Bennett Valley  which I mentioned earlier , lying proximity to Rogers Creek.



Next I'll take you on a short  Google Earth aerial tour of the  southern part of the faults. We  will start in the mud flats at the south end  of the faults and then we will fly  to the north around the left end  on the southwest flank of Sonoma  Mountain.  >> Although the fine scale features  won't be visible, see if you can  pick out where the fault follows  a broad bench and linear Valley. Along the flank of Sonoma Mountain.  These features are larger scale  and the results of long-term  repeated fault movement.  >> I will  show you the main zone of faulting  in yellow, usually it's in red.  I have marked locations of SAG ponds  , those water-filled depressions  and there are lots along this part of the faults.  Also I will mention the relief is  exaggerated two times in this  view.  >> Here we are, starting at the south  end of the mud flats and flying  along the fault and you can see  it's crossing a few SAG ponds. As  we approach Sonoma Mountain how  the fault takes a left  bend. See here if you can see that bench I was referring to on the  side of the mountain and there is  a slight Valley formed along it.  >> It  goes long east side of  Taylor Mountain and beyond that  you can see the topographic lobar  Santa Rosa is located and  you can see there's a right bend  where the faults goes across  Santa Rosa . That is the releasing bend that  we talked about. 


 Here is a static aerial image the  south end of the fault. The location  is marked by the Star. This is a  place where the fault geomorphology  is particularly well developed. Now I'll show you what it  looks like on a Lidar shaded relief image.  Here, artificial illumination is  from the Northeast. The fault is  marked by strong tonal  lineaments  through here. It pops out. The others  you see, here is a landslide,  those are, and along  the faults. There is the interpretation of the faults.



This is an example from the north  end, just south of the Russian River. Here  is no bleak aerial view, the pushpin  marks the location. Here is Rich Mountain and part of the town  of Healdsburg to the left. 


This is a part that wasn't mapped  as active prior to this study. 


 Here, with the imager you may be  able to pick out a sharp tonal lineaments. It continues here informs an uphill  facing scarp or step so stream alluvium has funded  against here. This would  make a good trench  site and we are looking to get access on private ranch land.  >> In the upper right is the interpretation  of the faulting. Here is the oblique image showing

     the matrix in red and you can see where it is wrapping around the  southwest flank. This left bend  restraining band you have considerable local uplift  in relief that has developed over  a long period of time. 


Next, I will talk about another line of evidence the Rodgers Creek Fault between  Santa Rosa and Healdsburg is currently  active and that is  fault creep. Creep is slow slipping at  or near the surface of a faults.  This corroborates the conclusion this part is active . First I will show you examples  of creep from the Hayward fault  which is a famous fault that creeps  famously. There's lots built across, lots of roads and  buildings that have been damaged  by continual creep. 


You can  see those assets. Only  recently has creep been recognized  along the the Rodgers Creek Fault.  Probably because  rates are slower and the fault is less urbanized. 


This is a study done by researchers  from UC Riverside and a documented  creep on and northern Rodgers Creek  Fault . Using  InSAR. This generates maps of surface  deformation over time. This map shows the pattern of faults parallel  surface motion where the velocities are positive,  the blues and purples, that  means movement of the ground is  toward the southeast. Where there  is an abrupt stepped in velocities from west to  east, that is consistent with right  lateral creep. The fault appears  to be creeping to the north of  Santa Rosa. It's absence to the south of  the city. 


Here is a graph showing how measured  creep rate  varies along the fault. The vertical  bars of 95% confidence intervals  of these triangles are creep measured using alignment  arrays closer into the faults. The average creep rates along  this portion is about two millimeters  a year which is less than a 10th  of an inch a year.  Not very fast. It appears to increase  quite a bit at Santa Rosa. It may  be as high as 68 millimeters a year  at Santa Rosa. To the south, there  need, there seems to be no creep . The interpretation is the area to the south maybe fully locked and storing all of the energy  we will talk more about that in  a few minutes.  >> We have seen  ground evidence of service creep  at a couple of locations. One in  North Santa Rosa and the other in  Healdsburg. Where a strand of the  fault crosses Santa Rosa Memorial  Park, curbs and sidewalks are deflected. About two inches or  five centimeters. This cemetery  is 50 years old so that indicates  a creep rate here of about one millimeter  a year. To the north where the fault  crosses at old Road in the open space preserve, creep  is expresses these fractures in  pavements. The pattern of left stepping  fractures is consistent with right  lateral slip.  >> Now we will take a look  aware the vault passes through the  topographic low. You will recall Santa Rosa lies within a right  releasing bend and geometry that  results in a component of extension  and land substance. Prior to the Lidar mapping the fault was believed  to be buried  by young alluvium  stream deposits where crosses the  flood plane a Santa Rosa Creek. 's location was inferred by projecting  location of map strands to the south  and north of the foot plane. 


Here is an elevation map created from the data . Red to green will go from lower  to higher elevations. There is a 30 foot change in relief across  the width of this survey. Which  is about three quarters of a mile  wide. This indicates the average  slope is only about a half percent.  It's pretty low  gradients. 


If you look closely you  might be able to see as it too shows  there ain't fault liniment set trend Northwest, across the floodplain. The grid pattern you can see our  city streets. Interpretation is  on the right. These strands define  the outline of what is known as Apollo parts  basin. Pull apart lessons commonly  develop to accommodate extension  that occurs at Ben's. Here is our little cartoon of restraining  and releasing bend showing where  pull apart reason might develop. 


The basin beneath has a  complex configuration, the  main here is three quarters of a  mile long and 2/10 of a mile wide. It bifurcate at the south end and  there's a separate smaller basin  to the Southeast. 


Now I'm showing Lidar based mapping on the Google  Earth  base and the inferred trace  is dotted in yellow. I  have also plotted the locations  of schools and hospital. It  seems like all  too often, we find these critical  facilities near active faults. 


This is a topographic profile  created from Lidar elevation data across the  most  prominent scarp on the east  side of pull apart basin. This white line shows the  location. I have plotted this a  25 times vertical exaggeration so  you can clearly see the scarp. You  will notice the slope below is more  gentle than above and this is likely  to to ponding and deposition of  sediment within that whole part  basin.  >> Here's a Grandview looking at the  scarp. It is expressed as a subtle rise in the stream  and what otherwise is pretty flat  foot plane topography. 


 An important question is how the  expression of faulting at the surface  may relate to underlying geology. Geophysical measurements of fluctuations  in the Earth's magnetic and gravitational  fields can be used to draw conclusions  about rock properties at depth. These are geophysical maps  and they reveal in East central  Santa Rosa lies a small but prominent body of rock  defined by  high gravity and strong magnetic , susceptibility. 


In detail, the Western margin

     as defined by these geophysical  edges or boundaries coincide in map view with eastern  side of the pole part basin. Shown here. 


This suggests to us this complex  geometry may be related to properties of faulted rock at depth. We  suspect there may be an increase  in frictional resistance along this  part of the that has led to this  geometry. This rock body in combination  with larger Bennett Valley high  to the south may be  a stuck patch . It has remained a lot and is accumulating  stress. If this  is the case, if this part of the  fault were to break in a large earthquake,  this could release a large meta-seismic energy beneath Santa Rosa. 


You may recall this is also the  south end of where creep has been  identified. This indicates a change  in slip behaviors hear from partially  creeping to the north to fully  locked to the south. This is consistent  with the idea this is a  stuck patch. 


It could also be this difference  in slip behavior

     could have led over time to bending  of the fault in creation of this  right releasing geometry. This is  a topic that needs further investigation  because it has implications  for shaking hazard and Santa Rosa. 


This October, is the  50th anniversary of a pair of moderate  size but damaging earthquakes that  occurred on the north side  of Santa Rosa. On the north side  of this dense  magnetic rock body here. This is  consistent with the idea that stress  may be concentrated in this area. Earthquakes were highly damaging  even though they were moderate in  size. Round shaking was strong in  the downtown area and quite a bit  of damage . In the next slide will take a  look at results of the study that  examined why this might have happens.  I will show you a map of sedimentary  basin the is that is derived from  this gravity map. Particular I will  talk about the Qatari basin.  >> The Qatari basin is right here . It is filled with low density  sentiment that thickens deepens to more  than a mile beneath the surface  southwest of Santa Rosa. I'm showing  you on this map , areas of concentrated earthquake  damage in Santa Rosa that occurred  in the 1969 earthquakes , this dark black circle and also in the 1906 earthquake shown by the pink oval. Even though the  1906 earthquake occurred on the  San Andreas fault more than 20 miles  away it pretty much destroyed downtown  Santa Rosa. They slow down and increase attitude  and thereby increase the shaking  at the surface. Downtown Santa Rosa  lies in the Northeast protruding  edge of the basin . The thinking is, energy for the 1906 in  1969 earthquakes was focused into  the area causing the strong  shaking that was experience. The  lower pair of maps show results of ground motion simulations that  tested the effect of the basin on  shaken. The strongest is  and read. On the left is the  model run with basin structure and  on the right, is without. You can see how  strong the shaking is if you consider this basin. 


These results support the inference  that the basin plays a significant  role  in the strength of shaking in and  around Santa Rosa.  >> Lastly, we will take  a look preliminary mapping of the  zone of active traces along the  Bennett Valley fault. The Bennett  Valley fault crosses the  force of flanks of Sonoma Mountain  until the Lidar data became available,  the forest canopy pretty much hid  evidence that the Bennett Valley  fault is active. Except for the  very north and,  this north trending strand here  has really well expressed  geomorphology and has been trenched  and shown to be active. 


The Lidar data  for Sonoma County reveals clear  evidence of recent vaulting all  along the Bennett Valley fault and  along faults in  the area between the Bennett Valley  fault and the Rogers Creek. In these shaded relief images,  arrows point to alignments of faulted  features. They are sharp tonal lineaments, scarp that faced East and West, offset landforms,  drainage , these are all classic expressions of active faulting.  >>'s  youthful geomorphology isn't apparent on the Lidar where  the fault crosses  the urbanized floodplain a Santa Rosa  Creek. There uplifted landforms  that mark the location of the fault.  I have inspected the Lidar to the  left  and found a few  discontinuous fault liniment here  the project northward which is  the next major plate boundary to  the east and north. 


 Here we have zoomed out to show the fault.  Here, the mapping comes from USGS online fault database  were color indicates age of recency of rupture. Oranges  strands, which there are some at  the north end here are active. Yellow have less age certainty so they can be regarded as potentially  active. If these faults are all active and they  are connected, earthquake cascades or multiple  ruptures are possible. Ruptured  begin on one vault and propagate  onto another. I will show an example of how this  could happen rupture starts,  continues onto the Bennett Valley  fault then on to Rodgers Creek Fault  and  there possibly onto Hayward. This connectivity opens the possibility  of larger than anticipated earthquakes . It would expose Eastern Santa  Rosa to greater hazard from shaking and surface rupture.  >> In summary,  new mapping of the Rodgers Creek Fault reveals it  is longer, broader and more connected  to its neighbors  than previously  known. Implying multiply fault ruptures  and large earthquakes are possible . Greater rupture complexity which in some locations such as  Santa Rosa may have applications  for patterns of energy release and  ground shaking hazard. And, a longer,  wider zone of faulting means more  areas are exposed to surface rupture.  >> Thank you, I will leave you with  some resources to help you prepare  for the next earthquake. 


[ Applause ] 


Any questions?



There's hydrothermal activity north of Santa Rosa along  Highway 101. That's most associated  with some sort of lifting type of fault, so any relationship to the  Rodgers Creek Fault North?  


 I'm not familiar with exactly were  you are talking about. Are you talking  about risers? 


North of Calistoga. 


It's the geysers, it's not on Rodgers Creek Fault, it's a separate  area  probably related  to general fault activity but is  not specifically related to a  major faults. I don't know  the answer exactly but I'm not completely  smile with where you are referring  to. Is not uncommon to have hydrothermal  activity related to faulting, lots  of springs are long faults a lot  of heat is generated at depth. 


I can't identify its relationship to the faults you  are describing but I can tell you  it's to the east of Highway 101. 


Thank you. Anybody else?  >> Can you say in layman's terms, a little bit  about how model the behavior of these plates?  >>



Specifically in terms of  earthquake forecast? Or  what? 


Had to approximate the behavior of the material? Is  that solid?

     >> There's a lot of data sets that  go into knowing about structure  of the earth and in terms of the plea motion,  we have geodetic information that  tells us about the motion , the broad regional motions, that factors into knowing about the plates and how they are moving.  There's lots of different data  sets that go into looking at the  structure. I'm  not entirely sure. My expertise is in one  particular area but there are certainly  lots of different geophysics and various data sets that go into  knowing the broad picture of how  the plate boundary works.  >>



You showed some intuitive features as plates move past each other.

     In one case I get a mountain, in  another, I get a lower points.

     It's not a whole. 


For that, that has  to do with the relative motion of  two sides of the else. When you have a right  lateral strike slip  fault and a band that goes to the  right you end up creating a small  amount relatively small amount of  extension  so it doesn't create a large hole.  Most of the motion is still strike  slip, it's horizontal . Because of the band there is a  slight amount of extension. Over  time, that can lead to a Valley. At the  same time you get materials sediment  deposited so it kind of keeps up  with it. It's a  gradual process but the amount of  extension is relatively small.  Compared to the faults movement  which is mostly lateral strike slip. It is a local  phenomenon. Along the fault but over time you  get landforms developed. 


Thank you. 


Just to add  to that,  when you have a single faults and it has a large earthquake rupture, you go out and measure it, there  is a variability that he mentions.  Some places, the slip is low and  in others it's higher. What happens over time? That's  one of the things we try to study  but it could be during the next  earthquake, the area of low slip is higher an area of high slip  has lower so over time, things balance  out. Or if you have a system of  faults, you can have  low slip and adjacent fault picks  up the additional slip. There is  a balance across the plate boundary , the Bay Area has the highest density of active  faults per square mile of any  urban center in the country. The earth itself is extremely complex . There is a  huge variability in physical process and properties as you go down deeper, as you go  along the length of these faults,  it's a very complex interaction. From geology we can actually measure the amount of slip and an  individual earthquake or over time to build up a rate of movement  in all these different faults and that is a primary focus of  our research to understand how these  complex boundary zones work and  the better we understand the physics of that, the better  estimates we can make of what will  happen in the future. How large earthquakes will be and  where they are going to occur.  >> Thank  you for your talk. I feel it was  well presented. 


You showed a couple examples  of surface ruptures. Are  those selected just to be the most  dramatic, clean structures or as it moves down the fault, does  it change from well contained two  more vague occurring over a  larger area across the fall. I would  think there would be a lot of  homogeneity in the rock as you move  along and sometimes it would be  breaking very closely and sometimes  more squishy. Is that true that at various or is it always  quite sharp along the entire fall. 


It can vary. I showed a couple examples of complex ruptures, the  letters earthquake rupture is quite  complex. Depends on what scale you're looking  at. Sometimes, you can get a very sharp single strand of  the fault, in other places,  there's quite a bit of complexity . It's hard to see in detail. As the rupture goes to the north  you can see how it is stepping from  one to another and in between those  individual strands you get a network.  That is a map scale  standing back. When mapping the Rogers Creek faults  we found zones of faults that were hundreds  of meters wide, half a mile . Some of that complexity could  be the results of multiple earthquakes  over time, rupturing overbroad zone . Certainly that kind  of complexity were you can get so  wide zone in a single earthquake  is something we see. Typically the  main zone of slip is fairly narrow in terms of  where most of the motion is. You  get a broader zone of deformation  in places . It varies, depends on where you  are on the faults.  >>



When you show  the gravity highs and lows in those  different basins and the particular one  near Santa Rosa, should we give  those as  rock that is  more solid, denser or not so much sediments but more  bedrock that forms those chunks that are more gravity high and the ones that are low gravity or fill and sediment? 


For the density, it's  exactly that. The low density which  would be the basin, they are fairly  unconsolidated. There still somewhat  consolidated but not as dense or  heavy as the rock that has the high  gravity signal. That rock body but he Santa  Rosa although we don't see it at  the surface, it's down at least  half a mile. That has been seen on  other faults associated with high frictional properties.  That may be . There are various  rock types but we have that to go on in terms  of knowing what it might be. Is  consistent with the idea this could  be an area where frictional resistance  is higher. This is work done  by a colleague of mine and part  of a paper I put together so I'm  not expert on that but that is my  understanding of how works. 


I'm intrigued by the thing  you mentioned a moment ago about as the fault occurs, their  areas of substance caused by stretching . I am wondering on a higher scale if that  phenomenon might be related to  the fact there is a break in  the Diablos and that's the Sacramento  River comes through. Is there any  correlation or could you talk about  that more?  >> You talking about the Diablo? David  is from Danville and knows that  area.



There is a phenomenon of the earth stretching due to curves in  the faults. I am curious about that on a slightly  broader scale if that might, that  might explain to some degree why  the Sacramento River  exits the central valley at that  particular point.



I don't know  what controls the Sacramento River.

     The west side of the central valley is really  characterized by folding  by compression, by broad warping like this.  It's a very different style of deformation. Different from what Suzanne has  been talking about.  How that has affected river flow  over time is a really  good question. I will just leave that  there.  >>



Not directly related to your  researcher presentation here, it  seems faulting is very three-dimensional situation, you're studying  the fault basin which are basically  surface traces. There's talking  about fault planes and other things . Are all these other things that go down to depth, it's not a simple  plain would be my guess.



What is the state-of-the-art  in terms of mapping the three-dimensional  nature, how much understanding do  we have of that, how much is speculation  and how much is measured ? 


The  geometry of faulting at depth  is something I don't work on directly.  Scientists use seismicity , if you have  well located earthquakes , that can actually delete fault  planes so a lot of the Bay Area  faults are pretty much lit up by  seismicity. The northern part of Rodgers Creek Fault has a lot of  small earthquakes and shows  it's  dipping towards the east . That's one way to get a handle  on the geometry. I showed you the geophysical maps . That helps constrain it in a general way. The  long-term geometry of the fault  can be shown by these kinds of data. It's difficult  to get a handle on the complexity  until there's an earthquake and  then we have a lot more information.  As often as faults come up, at  depth they are pretty much constrained

     to more or less a single plan. As  you get closer to the surface, they  tend to branch out on different  scales. The South Napa earthquake, show the array of faults  over quite a broad area. That is an example  people have model that geometry  in different ways. It looks like  the fault is coming up in, and branching.

     There's quite a bit of complexity. Is not easy to get a handle  on all of that before an earthquake  happens. 


Thank you. 


Any  other questions? Let's think Suzanne  Hecker  again. Fabric [  Applause ]  [ Event Concluded ]