In 1995, the Southern California Earthquake Center produced long term hazard maps for Southern California. For the first time, scientists from academic and government institutions pooled their knowledge of faults, earthquake repeat times, and ground motions caused by earthquakes to produce a map showing the long term odds of damaging shaking in this region. This sometimes contentious process has led to new insights on a number of issues and this will lead to improved hazards maps in the future.
It is important, however, to understand what the revised maps will, and will not, mean for seismic hazard in Southern California, and to give credit where credit is due. The latter issue requires, one hopes, the least discussion: the earthquake center hazard maps reflected an imperfect consensus but one that did succeed in gathering diverse and extensive knowledge and producing a bottom line for quantifying long-term seismic hazard. As scientists and lay people pick apart what was done wrong, they should remember that the alternative was to do nothing.
Understanding what the revised maps mean, however, is a little more complicated. The key characteristic of the new model is a consistency between the rate of earthquakes experienced by Southern California this century and the long-term expected rate. In other words, we have had about the number and magnitude of quakes that we expected to have.
But that doesn't mean that the greater Los Angeles region will continue to face seismic hazard only on a par with that experienced over the last 150 years. Suppose you deal 100 hands of poker and never see a royal flush: does that mean that a royal flush is impossible, or that you haven't dealt enough hands to expect to see such a low-probability outcome? Since a deck of cards is a known commodity, the answer is trivial; one could, in fact, work out how many hands one would have to deal to expect, on average, to see a royal flush. The properties of the earth and earthquake ruptures are somewhat more elusive and the "hands" are years that unfold at rate that human beings have no choice but to accept.
We do know some things, however, and the newest, best models for earthquake occurence include larger events for some regions than those that have been seen to date. In particular, something of a consensus has emerged in recent years, that earthquakes considerably larger than Northridge are both possible and expected over the long term, in the greater LA area. This forecast comes from lines of evidence ranging from a mathematical/ probabalistic evaluation of past seismicity, to an assessment of the size of regional fault systems, to actual field examination of the amount of slip that has occurred on faults in earthquakes that predate our historic record, but leave a prehistoric signature that can be understood today.
Perhaps the simplest argument for the plausibility of a M7.5 earthquake in the Los Angeles region is the 1952 Kern County earthquake near Bakersfield, which was of this magnitude: there is no reason why a comparable earthquake could occur not south of the San Andreas fault as well as north.
The remaining unresolved scientific issues include a refinement of the actual maximum magnitude expect for the LA region--7.5 or 7.8 or 8.2?--and the expected average rate for such events--300 or 500 or 700 years? Progress on these issues will alter the hazard maps and {\it could} reduce the long-term expected hazard for this region. If so, it would be appropriate to consider the revised results in societally relevant issues such as insurance rates. But no revision will alter the fundamental fact that virtually any site in Southern California faces appreciable risk of damaging earthquake shaking over any person's lifetime. The SCEC maps that use red to show high hazard cannot be expected to move much towards cooler colors.
And, when it comes to criticizing those who undertook the the enormous task of translating academic earth science knowledge into the usable public information that motivates funding of our research in the first place, please remember that nobody is perfect, but somebody had to go first.