Why can't the exact time and location of an earthquake be predicted?

Answer 1

Good question. Let's back up a bit. Earthquakes are generally caused by shifts in the tectonic plates that comprise the earth's surface. These plates are massive, and they are veined with cracks called faults. The faults are "fractures" in the structure of the plates, and they appear because the plates themselves are large and brittle and are being subject to massive stresses. The plates are moving away, toward or along each other, and the resultant stress on the plates is ginormous. There's the problem. The plates that form the earth's crust respond to the monumental stress in different ways, but the key is that they cannot respond "smoothly" to relieve the stress. Take as an example the junction of the Pacific and the North American plates at the edge of California. The Pacific plate is moving "down" or "south" along the North American plate, with is inching "up" or "north" along the boundary. Though the plates are moving slowly, the "friction" at the boundary is so large that it cannot be calculated with any accuracy. We have a vague idea of how the plates are moving but no way to find out how "smoothly" the movement is actually occurring. It's all about the "slip" here. If the plates are "hanging up" on each other (which they always do to a greater or lesser degree), they cannot move. But the force to move them still exists, and it will continue to "drag" the two plates past each other. If they don't want to move, the stress will build up until "failure" occurs along the boundary and the plates "jump" along each other. The "jump" is what triggers a quake. How can we know when things will "break loose" along the plate boundaries? We've got great technology applied to the "earthquake problem" as you've asked about. We can "watch" the plates slip past each other, and we can know when they aren't "slipping" any more. We then know that pressure is building and that the possiblity of a quake is increasing. But that's all we can guestimate with the tools we have. There isn't any way to get hard numbers when we analyze the forces involved and the "friction" at the plate edges as well as the "weakness" in the structures that will be the points of failure when the plates "jump" along each other. Make sense? We can't measure all those things and crunch the numbers to predict a quake. We can only talk about an increasing probability of a quake when we see the plates aren't "slipping" like they are supposed to be. Plate tectonics is the key to understanding a lot about quakes and vulcanism. A link is provided to the Wikipedia article to get you started.