Talk about big earthquakes, and people start thinking about California. Ask someone to name a major fault, chances are you'll hear the San Andreas mentioned.
The San Andreas is certainly a major earthquake-producing fault. It is a fault of a very specific type, and not the commonest type either. It's what seismologists call a strike-slip or transform fault, which means that movement on it is purely horizontal. The San Andreas Fault marks the boundary between two great tectonic plates that make up part of the earth's crust -- the Pacific Plate moving north slips past the North American Plate.
While Californians wait for the Big One -- the next earthquake similar in size to the famous San Francisco earthquake of 1906 -- people elsewhere in the world, living near similar faults, have their own Big Ones to worry about. And one such place is the Holy Land.
I use this term as a convenient geographical expression for the lands along the eastern shore of the Mediterranean -- Syria, Lebanon, Israel, the Palestinian territories and Jordan. From a geological viewpoint, this is an area dominated by another major strike-slip fault: the Dead Sea Transform (DST). This huge fault starts in the Gulf of Aqaba and runs northwards, up to the Dead Sea, up the Jordan Valley, through western Syria and eventually up into Turkey, where it merges with Turkish fault systems.
In fact, the earliest description of strike-slip faulting can be found in the Bible, in the book of Zechariah: "half of the mountain shall remove toward the north, and half of it toward the south" -- exactly what happens when a north-south trending fault like the DST moves in an earthquake.
So why isn't the DST named in the same breath as the San Andreas? The answer is historical. Although strong earthquakes happened along the DST at Aqaba in 1995 and near Jericho in 1927, to find a disastrous one you have to probe back much further in history -- such as the October 11th 1138 earthquake that struck the city of Aleppo, with casualties running into the hundreds of thousands. Or the May 20th 1202 earthquake, also in Syria, which at one time was believed to be the deadliest earthquake in history. (It was subsequently discovered that casualty figures had been inflated by including those who died in famine and plague in the next few years).
One thing the Holy Land has plenty of is history. As one Israeli friend of mine once put it, "every second stone is some sort of historical monument." So the history of earthquakes in the region can be traced back a long way, something that has been worked on for many decades by seismologists such as Nick Ambraseys in London, Iaakov Karcz in Jerusalem and Emanuela Guidoboni in Bologna, and their various collaborators.
Every so often someone rediscovers the curious fact that the seismicity of the DST in recent centuries is very much less than it was in ancient and medieval times and asks the question, why? Most recently, Shmulik Marco, a geologist from Tel Aviv University, has been widely reported in the press as warning that a major earthquake in the Holy Land is seriously overdue, even imminent.
How serious is this? Sadly, in seismology things are, to borrow from Oscar Wilde, "seldom plain and never simple." Let's go through the options.
The first possibility is the obvious one: strain has been building up all along the DST since the last major earthquakes, and in places it must be near breaking point. You can't keep bending rocks forever without them eventually snapping. The longer it's been since the last earthquake, it seems obvious that the more imminent must be the next. And it's been quite a long time.
Where is this strain coming from? Just as the San Andreas fault is a plate boundary, so is the DST. On the east, the Arabian Plate, moving north. On the west ... what? Well, technically, it's the African Plate, but things are not so simple. A sliver of the African Plate has broken off -- or is still breaking off -- making a small plate we call the Sinai Microplate. While this has a nice neat eastern edge along the DST, its other borders are uncertain. One possibility, admittedly rather speculative, is that the plate boundaries may be complex, and some of the strain caused by the Arabian Plate's northward movement is actually now being absorbed by scrunching up of the Sinai Microplate, reducing the need for earthquakes along the DST.
Next. Faults can sometimes come unstuck and slip smoothly without earthquakes (this is known as creep). Could this explain the lack of recent earthquakes along the DST? We can rule this option out. If the DST was creeping, this could be detected using GPS. It isn't.
I said it seems obvious that the longer the time since the last earthquake, the more imminent the next but, curiously, it's not necessarily true. Imagine you are waiting for a friend to turn up for dinner at seven. You expect him to be on time, but if he isn't, you expect he'll be along soon. Maybe the traffic was a little heavier than usual. But if by eight he still hasn't shown up, then maybe he's had a breakdown, in which case it's probably equally likely he'll be three hours late as two hours late. It's been suggested that this is true of earthquakes as well -- there is a typical interval between major events on a fault, but if an earthquake is late, it could be just as likely to be very, very late. If this is the case, the next Big One on the DST is still in the pipeline, but one should be cautious about using words like "imminent." Sure, it could happen tomorrow; but it would not be surprising if it's still decades away.
The last possibility relates to one of the most interesting discoveries made by seismologists in recent years, that large earthquakes seem often to occur in clusters. A fault will have a period of activity in which major earthquakes arrive every few hundreds of years, and then a protracted quiet period with no earthquakes over some thousands of years. The expected time to the next earthquake, then, depends on whether one is in the middle of a cluster or in a quiet period.
Recent work by a team of seismologists from Strasbourg concentrating on the Jordan Valley, and combining historical, archaeological and geological data, has demonstrated that this part of the DST does, indeed, show clustered behavior, with intervals between major earthquakes as short as 284 years and as long as 1508 years. Since the last major quake on this stretch was 979 years ago, we are clearly between clusters. The next cluster might well be imminent; but it would not be impossible for it still to be a few hundred years away.
Seismology continually throws up surprises. Earthquakes one is sure are imminent prove elusive, while others strike in places they weren't expected. A major earthquake in the Jordan Valley would be a humanitarian calamity, and whether it is imminent or not, it will happen eventually. What matters is to strengthen communities so as to minimize casualties when the inevitable occurs.