Can We Predict Earthquakes?

Not Geography Geography Lesson 3 

The Dodgy Science of Earthquake Prediction

In slight contradiction to the 'not-geography' theme, this week's focus is somewhat stereotypical for the subject - Earthquakes (EQs)! And the all important question of can we predict earthquakes?

The aftermaths of earthquakes can be absolutely devastating - in the past thousands of people have died, and millions of dollars have been lost in single events, and they receive global attention for the scale of rescue and repair efforts required. Entire cities collapse, enormous landslides are triggered, fires consume the landscape, and the ground can even liquify underneath you (known as liquifaction). They are bad mojo. Hence, governments and scientific organisations have invested big bucks and decades of work into understanding their mechanisms, with a view to predicting them and preparing accordingly. 

The EQ Prediction Race 

Although the modern seismograph was invented in the late 1880's (and some forms of the device have been around in China since 132 AD!), they were less reliable than a Northern Rail train until around 1980. From this point though, an innovation race was on between China, Russia and the US to find a pattern in seismic waves that could reliably be seen to 'predict' EQs by way of precursors. Millions upon millions of dollars were invested into research programs, and some early successes heightened the scientific community's optimism that within the next decade, we could successfully predict EQs. 

It All Goes A Bit Downhill 

Despite the previous optimism, research began to stale and plateau, and slowly scientists and governments lost hope. A series of severe failures in prediction (and not prediction!) led to the closure of all bar one Japanese program. And you don't have to be a MENSA member to see why....

A Series Of Incredibly Unfortunate Events 

• 1976 Tangshan EQ, China - In the previous year, Chinese scientists had 'successfully' predicted the occurrence of a 7.3M EQ in Haicheng, and saved 1000's of lives through their evacuation. Full of hope and confidence from this, they went on to highlight several other areas ready for a big event - Tangshan was not one of these. On the 28th July, a 7.8M EQ tore through the region, and left between 250'000 and 700'000 people dead (there are some dodge figure misreporting issues, but that's another story!). The fact that China failed to foresee this event, the second most deadly natural disaster in modern history, off the back of their previous success, went a serious distance to discrediting the science of EQ prediction. 

• 1990 New Madrid EQ, Missouri - Whilst not being known for having regular EQ's (as there are no tectonic plate boundaries close to the area), the US's mid-west is actually a seismic zone; this is known as intraplate seismicity. There is reliable evidence that big EQ's have occurred in this area in the past. And so in 1990, a man by the name of Dr. Iben Browning declared that the coming alignment of the moon was going to trigger an enormous and devastating event in the very near future. Whilst every seismologist and his dog rubbished the idea, the local media seized it with both hands, and generated absolute pandemonium. People were evacuated, buildings reinforced, and for the very reasonable price of $39 you could purchase a 30-minute recording of Browning describing how he had rendered his wonderous prediction. When the 3rd December rolled around, and the earth didn't so much as hiccup, people realised that perhaps they should have been slightly more skeptical. And with that Dr. Iben Browning made his fortune in pseudo-science recordings and 'I survived the 1990 New Madrid Earthquake' t-shirts. And the world lost another piece of it's dwindling hope in EQ prediction. 

• 1985 - 1993 Parkfield Experiment, California - The final nail in this sorry coffin, is the slightly funny but mainly sad failure of the San Andreas Parkfield experiment. With government funding, the USGS (US Geological Society) set up this experiment, with hopes of proving their accuracy in predicting EQs on a small segment of the infamous San Andreas fault, based on the number of years between events seeming to be highly regular (known as interval times, or Seismic Gap Theory). By their reckoning, the next big one was coming in 1993, so they set up $1000's of dollars of equipment and the whole world sat and waited. 1993 came and went... as did 1994.... and 1995.... By 2004 I imagine even the most dedicated seismologist had packed up his camping chair and gone home. An earthquake did not occur until that year - 11 years late. Rude. Unfortunately, this was the final nail in the EQ prediction coffin. 

But We Do Still Get EQ Warnings..?

Whilst we no longer believe we can predict EQs - which is on a short-term basis of hours to weeks - we do still produce reasonably reliable EQ forecasts. These run on much longer timescales, and are probabilistic in nature. Based on seismological, historical and geological records, seismologists build a timeline of previous EQs as accurately as possible, and use that to determine the probability of an EQ happening within a given time-frame, using recurrence intervals and maths. In all likelihood, these are the best estimates we will ever be able to produce. 

Unfortunately we will most likely never be able to predict earthquakes, but there is a moral to the whole sorry affair - if Dr. Iben Browning could have the nerve to fool an entire nation with a fictional apocalypse, you can talk to that girl you like.