The original question was very broad. I answered it one way, other respondents answered it other ways. The question is similar to: are there MapInfo programs that can calculate the occurrence, location, timing, and effects of large hurricanes? Here is some background info (I hope I do not bore too many of you). There are several sides/parts to the original question: a) developing an understanding of the timing and type of shaking (ground motion) from large earthquakes, b) developing realistic (?) 3-D models of the subsurface surrounding the location of earthquakes, c) developing realistic models of the effects of earthquake shaking on structures (earthquakes do not kill people, buildings do), d) Deciding how and what type of documents/maps/reports/models/regulations/laws a government should create to inform/educate/control/restrict the public (both the general public and the technical public - engineers and geologists). Hundreds of geologists, engineers, and planners have been working on these problems since 1906. Understanding of the timing and type of shaking output of large earthquakes. Large earthquakes have reoccurrence intervals in the range of 100 to 1000's of years. One of the best ways to estimate the reoccurrence is to excavate trenches across the main fault trace. Examination of the trench walls allows a geologist to estimate the time between major earthquake events (over the past 500 to 1000 years or so) and a rough idea of the size (magnitude) of the earthquake. When enough trenches have been evaluated, one can get a reasonable idea of the average(!) time between earthquakes on that particular fault. Each fault has different parameters. This technique only works for faults that cut the surface of the earth with a distinct fault plane. You have to use different methods for large crustal earthquakes (such as those that hit the east coast every 200-400 years), thrust fault related earthquakes, and subduction type earthquakes (such as those off the northwest coast of the US). This question has also lead to the wide spread placement of seismic motion recorders. When an earthquake occurs, the ground motion is recorded to help improve the data/parameters for that fault. There are no for-sure answers to the timing and shaking of earthquakes, and only the well known active faults have been analyzed (it takes years to decades to do enough trenching to get sufficient data). This problem has lead to the development of mathematical models to estimate the physical effects of large earthquakes. There are two camps - the deterministic and probabilistic. The deterministic group evaluates the potential shaking effects by specifying a fault (and its parameters) and then evaluating the effects of that fault on the site in question. For example, the published information about a fault indicates that it will generate a magnitude 7.6 earthquake, has a north-south trending fault plane, and is 10 km from the site. Plugging this date in to various equations indicates that this fault would cause a ground acceleration of 0.6g at the site in question. Each fault in the surrounding area would then be evaluated, and the highest resulting site acceleration would then be used in the structural design. The fault parameters are based on actual analysis (trenching, etc.) of that fault (hence the term deterministic), and are used to come up with terms such as the MCE (the maximum credible earthquake) for each fault. (you could also call this a fault based analysis.) The probabilistic group recognized a major problem. What do you do about faults that you have no information about or that you do not even know exists? They studied the problem and came up with a different method. They do not care about specific faults. Instead they look at an overall region, evaluate it, and create a series of probabilistic equations. Note, the equation are not created out of thin air. They are based on the same data that the deterministic group uses. Using these equations one can estimate the scale of ground shaking (0.3g, 0.5g, etc) that will occur at a site over a specified time period (10 years, 50 years, 100 years, etc). The longer the time period, the greater the ground shaking (you start pulling in bigger and bigger earthquakes as the time increases). There is no such thing as a worse case earthquake in this methodology. However, the engineer/geologist has to choose the level of risk (not an easy task). A hospital is a critical structure and would be would be evaluated over 5,000 to 10,000 years (resulting in a large ground acceleration), whereas a house might only be evaluated over a 100 year time period (resulting in a lessor ground acceleration). (you could also call this a site based analysis) In general, a deterministic evaluation will give ground shaking values 10 to 15 percent less than a probabilistic evaluation. However, over long time periods, the deterministic values will equal probabilistic values. Developing realistic 3-D models of the subsurface surrounding the location of earthquakes The above is only concerned with the faults themselves. Seismic waves that occur during an earthquake travel through the ground to a site. The properties of the soil/rock through which the seismic waves travel have a large control on the resulting effects of the earthquake (called site effects). This is why someone standing on bedrock will barely feel an earthquake while someone a half a mile away on basin fill will become sea sick from the ground motion. In the simplest earthquake models, the ground is assumed to be a homogenous soil and everything reacts the same (this is what is assumed in many of the probabilistic models - one of the drawbacks to that method). The problem is in gathering enough data to develop a geologically reasonable 3-D subsurface model of an area. This includes the spatial variation in soil/rock types, as well as defining the geotechnical parameters of each of the soil/rock types. The work is on going, but will take a decade or more to complete. The fun thing is that when you have a good 3-D subsurface model, you can put the sub-surface data and earthquake parameters into large FEM/FDM codes and model the basin effects of individual earthquake events (the USGS has some neat videos of this type of computer output). This is called a strong ground motion analysis. When you put an earthquake into the geologic model, how much of the earthquake signal do you use? Now, only the main part of the signal is used, the Coda (the tail) is ignored. There is also the problem of the attenuation relationships for each of the soil/rock types. The attenuation factor indicates how much energy a soil/rock type absorbs - it fundamentally controls the look of any seismic shaking map. There are no easy answers, and as with any large-scale endeavor, there are competing groups/models/concepts. Developing realistic models of the effects of earthquake shaking on structures This may sound like basic engineering, but it can be difficult. In any area, there is a wide range and age of structural designs. Each will react slightly differently. To analyze this, you have to survey all of the buildings in an area. (a wood framed building is affected by high frequencies while a tall building is affected by high frequencies). How much ground movement can a building withstand? We generally talk about ground shaking, but what about permanent ground deformation (landslides, lateral spreading, liquefaction, etc)? There is also the fundamental design question. Do you design the building for life safety or for usability? Under life safety, the building has to hang together long enough for people to get out safely (the structural engineers have always assumed stable ground). You do not care if the building has to be torn down later. Over the past few years, governments and insurance companies have become aware of the infrastructure costs, and there is a move to design the buildings to a usability level. The people have to get out OK, and the building has to be serviceable (easily repaired) after the earthquake. This is a much more expensive design standard. On top of this, you also have to include all of the previous earthquake and geologic problems mentioned earlier. This is where the EQE model appears to fall. Their model appears to be designed to estimate the effect of a specific earthquake (a determinist model) on infrastructure. They have gathered data about the buildings in an area (from insurance companies). What is unknown is the level of their basin geology. Do they assume a 1-D soil model, or have they gone to a more realistic 3-D geologic model (basement depths, formation velocities, etc) that also includes liquefaction and other localized hazards? Deciding how and what type of documents/maps/reports/models/regulations/laws a government should create to inform/educate/control/restrict the public This seems simple, but is quite complex. What is the purpose of data released by a government? Are seismically related maps designed to reduce the damage from future earthquakes; are they intended to be used by engineers or the general public; are they intended to protect life or property (these are very different); are they designed to allow planners to include seismic information in regional plans? Is it possible that we will never sufficiently understand earthquake related hazards and are wasting our time? Some engineers have a very different view. They say, why waste the money on geologic/seismic evaluation? We know earthquakes will occur in certain areas, just design all of the buildings in that area to withstand a certain level of shaking. We would not need all these fancy codes and parameters. I have gone on to long, but it is a fun subject. s. figuers ---------------------------------------------------------------------- To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put "unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]
