|
Here we discuss some methodes and approaches that scientists are using in order to estimate the occurrence of earthquakes in the future. Non of them can provide the exact moment in time or the magnitude of the earthquake, but they can do it in an approximate way. Thus they give a lot of useful information for precautions to be taken, such as building adequate structures. For example, if engineers are told that the building or bridge they are designing will have to withstand a shock of up to 0.5 g in the next 50 years then they can design the building to cope with this.
Some of these considerations are the following:
This interval tell us how frequently earthquakes tend to occur on a certain fault, and the maximum ground motions that eartquakes are likely to produce in a certain area over a certain period of time. It is worked out using information from several different sources: historical recordings of earthquakes, geological evidence (the tell-tale traces which earthquakes leave behind), and geodetic evidence (the amount of strain building in the rocks). Following the hypothesis that big earthquakes take place between similar periods of time, these can be used to predict future earthquakes in terms of probability. However, the accuracy of these long-term probabilities based on recurrence intervals is pretty limited because the story in a fault might change from period to period due to the existence of new forces.
Another way of predicting earthquakes is to measure displacement along in a fault. Harry F. Reid, a Californian seismologist, could predict that the next shock on the San Andreas Fault would occur about a century after the big quake on the fault in 1906. Survey measurements taken before the earthquake indicated that the ground was becoming strained, on average, by almost 0.65 m per decade. Reid figured that since the maximum displacement along the fault in the 1906 quake was 6.5 m, and therefore probably the result of a century of building strain, an earthquake of similar size probably occurred on the fault roughly every 100 years.
Nowadays, satellites capable of providing precise positioning information (GPSs) allow seismologists to determine which an incredible precision how much and where the Earth is deforming. Repeated measurments can determine if the fault is creeping. Thus, the rate and the strain of a fault region can be followed and similar predictions can be even better made.
Here the key assumption is that big earthquakes tend to happen in the same place each time. If you make a plot of all the big earthquakes on a plate boundary, you find they fill discrete, adjoining segments of the boundary. A seismic gap is a segment where there has not been an earthquake for a long time but where there is a historical record of an eartquake in that area in the past.