The introduction talks about the structure and geology of earthquakes. The San Andreas, for example, only produces big quakes. The surfaces that are pulling past each other have been ground smooth enough that they stick together. This means it can't release pressure a little at a time; it waits for a sizable build-up, and releases the tension all at once. The magnitude of a quake is determined by how much of the fault releases at once. If it's a short distance, it produces a small quake. Quakes that release a few yards of pressure would be under 2.0. If the rupture goes for a mile and then stops, you get a magnitude 5. A 100 mile long break would produce a magnitude 7.5 quake. Since the rupture front on the San Andreas is pretty smooth, a quake on it will continue to propagate once started, and will cover most of the length of the fault. The built up stresses (at two inches a year) on the southern end of the fault have accumulated about 26 feet of differential since the last major release more than 300 years ago. The section in northern California has had more recent quakes. If two hundred miles of the fault give way, we're talking about 7.8, while 350 miles is conceivable, and would reach 8.2. The section around Paso Robles releases pressure gradually, and should stop further propagation.
I was already pretty aware of the big picture for a major earthquake, since I've been part of the earthquake response teams both at Google and for Mountain View. After a major quake, some roads will be out, and all the fire, police and hospitals will be busy, so no one will get the mutual assistance that they can usually count on. Jones led a team of more than 300 experts as they explored what an 8.2 in LA would be like. Even though building codes have been improving for several decades in California, not enough have been retrofitted to keep this from being a serious disaster. 1500 buildings are likely to collapse including possibly some high-rises. When we drill in Mountain View, or at Google, we always assume that we'll be on our own--no fire or medical help should be expected for a few days. Google (and most other large employers) have plans to be able to feed employees for a few days, and the earthquake team is trained in triage and first aid. But anyone needing attention from a doctor is unlikely to get it.
And all that was in the introduction. The next few chapters cover the volcanoes that buried Pompeii in C.E. 79 (there were early warnings, so there are eyewitness reports from people who fled days or hours before the final eruption) and Iceland in 1783, and the earthquake that shook Lisbon in 1755.
I want to spend more time on chapter 4, which covers the great flood of California's central valley in 1861-2. Just that description should make you suspect that it was bigger than you'd expect. This was a flood that filled the Central Valley, and the water didn't recede for 9 months. California had only been a state for about 10 years at the time, and the only thing that most Californians today have heard about this event is that Sacramento raised its street level by 10 feet in response.
Most people who are familiar with California weather know that most of it is basically a desert. It usually only rains in the winter, and most of the rain falls in the mountains. We only have enough to drink because we dam the rivers, and store water from rainy years in the reservoirs. If you live here for a while, you get used to the idea that some winters are pretty dry, and other years, we'll get a couple of storms that seem to get stuck here, and we can get rain that lasts for a week or two.
Starting in December 1861, the rain throughout much of the state was continuous for nearly 45 days. Other than the mountainous areas, normal rainfall is 12-18 inches, with 24 inches being heavy. That storm apparently dropped 5-6 feet of rain in many places. There were no dams at the time, so by January 9th, the water in Sacramento was 24 foot above its normal level. Most of the city was at 16 feet, so the water was 8 feet deep. The water was still there 3 months later. But this was only what was visible at Sacramento, which is pretty much the northern tip of the central valley. The entire central valley: 30 miles wide and 200 miles long was inundated to a depth of thirty feet. Innumerable cities and towns were completely washed away. All the cattle grazing there died.
Modern California has dams and reservoirs, but they wouldn't have been able to hold back this much water. It was only 150 years ago, and there's no reason to think that extreme variation in annual rainfall has abated. says that geologic records indicate we should expect this much rain "once every century or two", which is suitably vague, but scarily often. There's no way we're prepared for an event of this size. We now get decent alerts about rain two weeks ahead, but several recent winters have included anomalous weather patterns that persisted for longer than that, and the weather bureaus don't have much more to say than "we can't tell how long it'll last". If it starts raining and doesn't stop, we won't know until two weeks before all the dams are overtopped.
Later chapters cover flooding on the Mississippi and in New Orleans, tsunamis in the Indian Ocean, other disasters in Italy and China, and Japan's Fukushima, which combined earthquake, tsunami, and nuclear meltdown. She talks about emergency response, long-range preparedness, and our tendency to estimate the future based on past incidents we're familiar with.
