Earthquake prediction: What we have learned and what we should do now
abstract Adopting out of necessity a definition of “earthquake prediction” which stresses short-term warnings (months, weeks, day, and hours), I must conclude that either we are doing the wrong things or such predictions may be impossible. Assuming the former, I recommend major emphasis on direct or nearly direct measurements of stress/strain change in or across fault zones in the immediate vicinity of strong points or “asperities” (how do you find asperities?). This recommendation presupposes essentially total emphasis of the United States' program in California. Even if this recommendation is unpalatable, it is certain that resources now available require a total reanalysis of the present program and a consequent reallocation of funds.