NORTHRIDGE EARTHQUAKE of January 17, 1994 Magnitude 6.6 4:00 p.m. January 18, 1994 Information from the Caltech Seismological Laboratory and the U. S. Geological Survey SUMMARY An earthquake struck the San Fernando Valley yesterday at 4:30 am. Pacific Standard Time. The epicenter is located at 34deg 13U North, 118deg 33U West at a depth of 14.6 km. The surface wave magnitude from the National Earthquake Information Center is 6.6. The local magnitude is 6.4. FAULT ORIENTATION The two planes of the focal mechanism of the earthquake both strike slightly north of east and show almost pure reverse motion. One dips 60deg to the south, and the other dips 30deg to the north. Which is the causal fault is still uncertain. The mainshock's epicenter is several kilometers south of the southern end of the rupture zone of the 1971 San Fernando earthquake (magnitude 6.6). Most of the aftershocks of the Northridge earthquake are located to the north of the mainshock with relatively unconstrained depths. Because of the lack of depth control, we cannot at this point discriminate between a shallowly-north-dipping plane with the aftershocks downdip from the mainshock and a steeply-south-dipping plane with the aftershocks updip from the mainshock. The mainshock occurred at the approximate depth predicted for the Santa Monica Mountains extension of the Elysian Park fold and thrust belt. A small part of the Elysian Park fold and thrust belt broke in the 1987 Whittier Narrows earthquake. Portable seismographs have been deployed to improve the depth control of the aftershocks and help resolve which fault caused this earthquake. SURFACE RUPTURE In the past thirty-six hours, we have been gathering geological information relevant to understanding the earthquake that struck the Los Angeles area yesterday. From our initial field reconnaissance yesterday, we concluded that no major surficial fault rupture occurred. We have received no reports today from other geologists in the field that contradict this initial assessment. This lack of obvious expression contrasts sharply with the 15-km-long surficial rupture that accompanied the 1971 San Fernando earthquake. In this respect it is ,however, very much like the 1987 Whittier Narrows, 1989 Loma Prieta and 1991 Sierra Madre earthquakes. One small zone of disturbance that was verified by USGS and Caltech geologists on the ground today does appear to be the result of motion along a fault. This few- hundred-meter-long system of cracks runs nearly east- west across Balboa Blvd, north of the Simi Valley freeway. Its location and orientation suggest that it is the result of minor movement along the Mission Hills fault. Movement of the fault resulted in major disruptions of water and gas pipelines. Seimographic information indicates that the fault that produced the earthquake lies many kilometers beneath the northern San Fernando Valley. It.is either a steeply inclined structure that dips south or a shallowly inclined structure that dips north. Aftershock depths are still too poorly constrained to conclude with certainty that the steep south-dipping plane is the fault that produced the earthquake, but at this time, we favor this interpretation. A south-dipping plane would be consistent with south-dipping faults known from oilfields in the northern San Fernando Valley. These faults appear to be part of a system of south-dipping faults that we are calling the Oakridge fault system. Well data indicate, however, that these faults have not moved at near the surface for at least the past half million years. This does not tell us whether they broke at depth. It is consistent with the lack of surface rupture in this event. AFTERSHOCK PROBABILITIES See accompanying discussion of aftershocks. ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Aftershocks -- What to Expect January 18, 1994 4:00 p.m. Lucy Jones Seismologist, U. S. Geological Survey Visiting Research Assoc., Caltech Earthquakes occur in clusters. One of the first things recognized about earthquakes is that large events hardly ever occur alone. When one earthquake happens, we usually see another at a nearby or identical location. To be able to talk about this phenomena, seismologists coined three terms -- "foreshock," "mainshock," and "aftershock." In any cluster of earthquakes, the one with the largest magnitude is called the mainshock; anything before it is called a foreshock and anything after it is called an aftershock. A mainshock can turn into a foreshock if a subsequent event comes along with a larger magnitude. Earthquakes happen over an area of a fault, called the rupture surface. Because of friction, when the rocks on each side of the fault are pushed sideways, they do not slip immediately. Eventually enough slip is built up and the rocks slip suddenly, releasing energy in the form of sound waves and shear waves that travel through the rock to cause the shaking that we feel as the earthquakes. Forget what your high school science books said about earthquakes happening at a "focus." The focus (seismologists use the term hypocenter, epicenter is the point on the Earth's surface above the hypocenter) is only the point where the earthquake starts. The rupture begins at that point and then spreads down the fault. It keeps moving down the fault until it runs into something that stops it (exactly how this happens is one of the hot research topics in seismology). Because each point on that surface radiates energy, the bigger the fault's rupture surface, the bigger the earthquake. Clustering of earthquakes usually occurs only very near the location of the mainshock. The rupture surface that moves in the mainshock experiences a great redistribution of the stress on it during the mainshock and it is that disrupted surface that produces most of the aftershocks. Sometimes the change in stress in the mainshock is great enough to trigger aftershocks on nearby faults. However, the stress change dies off quickly with distance from the fault so we rarely see aftershocks more than a few kilometers from the main fault. As a rule of thumb, we say that aftershocks are other earthquakes triggered at a distance from the mainshock fault no greater than the length of that fault. Bigger earthquakes have more and larger aftershocks. As the magnitude of the mainshock increases, the magnitude of the largest aftershock, on average, increases as well. The difference in magnitude between the mainshock and largest aftershock can range from 0.1 to 3 or more, but averages 1.2 (a M5.5 after-shock to a M6.6 mainshock for example). Below that, the number of aftershocks at each magnitude level goes up as the magnitude of the aftershock goes down. On average, for each magnitude 5 aftershock in a sequence, we will see 10 magnitude 4 aftershocks, 100 magnitude 3 aftershocks, 1000 magnitude 2 aftershocks, etc. In general, an earthquake large enough to cause damage will produce several felt aftershocks within the first hour. The rate of aftershocks dies off quickly with time so even the second day will have many fewer aftershocks than the first. The daily rate of aftershocks is proportional to the inverse of time since the mainshock. Thus the tenth day after the mainshock will have approximately 1/10 the number of aftershocks that the first day had. We call an earthquake an aftershock as long as the rate at which earthquakes are occurring in that region is greater than the rate we saw before the mainshock. How long that will be depends on the size of the mainshock (bigger earthquakes have a higher rate of aftershocks so it stays above background longer) and how active the region was before the mainshock (if it was quiet, the aftershocks stay noticeable longer). The relative number of small to large aftershocks does not appear to change with time. However, since the overall rate dies off, all magnitudes become less common with time. Since small magnitudes happen much more often, we have more of them later. But all magnitudes dies off at the same rate -- magnitude 5Us are 1/10 as common on day 10 as day 1, and magnitudes are 1/10 as common on day 10 as day 1. It is also possible that the first earthquake will turn out to be a foreshock to an even larger event (6% of the time in California). Like aftershocks, the chances of this happening also die off quickly. The most likely time for a mainshock is within the first hour (one- quarter of all mainshocks happen within an hour of their foreshock) and after three days the risk is almost all gone. Aftershocks to the Northridge Earthquake Please read the preceding description of aftershocks to understand the general patterns of aftershocks. The aftershocks to the Northridge earthquake are following a very regular pattern -- almost exactly average for what we expect in a California aftershock sequence. As of 3:30 p.m. on Wednesday, January 19, 1994, we recorded 4 aftershocks of magnitude 5.0-5.3, 23 aftershocks between magnitude 4.0-4.9 and 202 aftershocks between magnitude 3.0-3.9. The magnitude distribution is normal. The aftershocks are decaying as expected. The chance of another magnitude 5 in the next week has decreased as time has passed to about 1 in 3. The chance of a more damaging magnitude 6 is lower. More felt aftershocks are sure to happen but will decrease with time. Numbers Technically, the rate of aftershocks is rate= 10^(a-b(M-Mm) * (t+c)^-p where M is magnitude of the aftershock, Mm is magnitude of the mainshock, t is time since the mainshock and a, b, c, and p are constants. Once you determine those constants you have a rate and with a rate, you have a probability. The constants in California have been about normally distributed and the average values are the "generic" aftershock sequence. Right after an earthquake, we use the generic earthquake to estimate probabilities. Notice the role of the mainshock magnitude -- bigger earthquakes have many more aftershocks. As we record the sequence we can estimate the paramters for that sequence. We find that the Northridge sequence is very close to average. p=1.07 (exactly generic), b=1.0 (generic b=.9), c=.07 (generic c=0.05) and a=-1.45 (generic a= -1.76). The most noticeable variation is a slightly high a although even that is not a big diffference. It is well within 1 S.D..