Earth & Sky

An earthquake in Pakistan in 2005 killed thousands

Peter Molnar explains why that earthquake was inevitable.

In 2001, Peter Molnar was co–author of an article in the journal Science, warning that “one or more great earthquakes may be overdue in a large fraction of the Himalayas, threatening millions in that region.”

In 2005, a devasting earthquake in India and Pakistan caused the deaths of at least 87,000 people and left millions homeless.

Earth & Sky’s Jorge Salazar spoke to Molnar in October 2005, a few days after the earthquake.

Salazar: Why was an earthquake was predicted for the Himalayan region?

Molnar: I urge you to be very careful with the word “predict.” For an earthquake scientist, that word has a lot of baggage. It implies more than specifying where an earthquake can occur. It requires specifying a time.

And, to be useful, the word “predict” has to do much better than what we wrote about in that little article in Science. People who use that word generally want it to mean much more precise in space and time than what we’re saying.

The article just called attention to the fact that earthquakes are inevitable. They are more common in some regions than others. It’s been a long time since large earthquakes in many areas, including this part of Pakistan. There’s no historic record of an earthquake this size, though we would argue that an earthquake of this size was inevitable. And I would go further and say that this earthquake is probably on the small side compared to what we might have expected.

Salazar: Why was this earthquake inevitable?

Molnar: We know India moves northward towards Siberia at about 40 millimeters a year, or 1 1/2 inches a year. And we know that roughly half of that northward movement is absorbed right there at the Himalayas. Essentially, India is sliding beneath the Himalayas.

We know that’s happening. It’s measured in a variety of ways, most recently with GPS measurements. And if you’re sliding India under the Himalayas –the rate there is 20 millimeters a year it means that you accumulate 2 meters of slip in a century, and 4 meters in 2 centuries.

And we have a historic record that’s adequate for the last 200 years. We have a good record for the last 100 years because there have been seismographs recording all over the world. For the 19th century, there’s enough written record that, I think, we know where earthquakes occurred along the Himalayan arc, along most of it at least. So we had the last 200 years covered.

So, where earthquakes have not occurred in that region, we know there has accumulated approximately 4 meters of potential slip. When you have four meters of slip, you get an earthquake with a magnitude of, roughly, 8. If you have more slip, you have an even bigger earthquake.

My guess, for the earthquake that occurred on October 8, 2005, is that the slip wasn’t even four meters, that it was a little bit less. I base that on what I’ve read from the USGS and the Harvard group that studies these earthquakes in a quantitative way.

It’s actually a rather simple calculation. Four meters of slip is a large earthquake, and it takes two centuries to accumulate four meters of slip. There have been earthquakes not far from this one in Pakistan. There was one in 1803 in northwest India, southeast of this earthquake. And there was another one in 1905 that was definitely in what we call India today, not near the border with Pakistan. That earthquake we know more about because it was recorded by seismographs, and it was studied immediately afterward by members of the Geological Survey of India, at that time largely British scientists. So, we know something about that earthquake, and I would say that it was comparable to the one that occurred on October 8.

We know that much. That’s easy to do. Going better than that, we know places that are overdue. What we don’t know is whether it’s tomorrow, whether its occurred and we haven’t recorded it yet in North America, which is unlikely. Or whether there’s another 10 to 20 years before that earthquake will occur. We just don’t know.

Salazar: Earlier you mentioned that the October 8 earthquake was relatively small. What do you mean by that?

Molnar: We can compare the size of this earthquake with others that have occurred in this region. There are three that we know something about, because they were recorded by seismographs and studied in the field.

One, in 1905, seems to have been the smallest of the three. The second, in 1934, which mostly affected Nepal, was not accessible to foreign visitors at that time. But the Nepalis themselves actually carried out quite a thorough study of the damage. That earthquake was substantially larger than the one occurred on October 8. The length of the fault that ruptured was 200 to 300 kilometers, compared with what appears to be at most 100 kilometers for the October 8 earthquake. So, if you like, the 1934 earthquake was 2 to 3 times bigger. The largest earthquake that we know of in this region occurred in 1950. We know well how large it was, because it was well–recorded by seismographs that existed at that time. And it was huge. I think it was the largest earthquake to have occurred within a continental region in recorded history. It occurred in a remote area, in 1950, at a time when India was preoccupied with developing independence, which it had had for only three years. So, the 1950 earthquake was not studied very well. There’s not very much known about what happened, and of course, being in a very remote area, they just couldn’t get to it. But it was very large.

What we know is that the Himalaya can sustain large earthquakes. I might say that those I’ve just mentioned are small compared to what the geologic record suggests. There are places where surface faults have been found, and, by dating these and studying them by geological methods, one can determine that it’s probably one event that caused the amount of displacement. The displacement can be measured, because you see offset layers of rock. As much as 15 meters of displacement has occurred on some occasions, in contrast to 4 meters of displacement for the October 8 earthquake.

There are suggestions, unpublished and incomplete, of extremely large earthquakes. One, known mostly from records in monasteries in Tibet, apparently occurred in 1505. The records suggest a very long rupture not 100 kilometers as for the October 8 earthquake but roughly 500 kilometers long.

So we know, from the basis of the other geologic work, and we suspect from these historic records, that even larger earthquakes than those known in the 20th century have occurred in this region in the last 1,000 years or so. And those earthquakes are much larger, 10 times larger than the one that occurred in Pakistan on October 8. They are really big. We know that the potential is there. That’s the main reason for saying that the one in Pakistan is relatively small.

Now, I would go further and say that looking on a map, there appears to be a gap between where the October 8 earthquake in Pakistan occurred and the one in 1803, a couple of hundred kilometers to the southeast. The segment between those two has not slipped in 200 years. So, the potential is there for another earthquake.

I certainly haven’t used the word “predict” here. But roughly half of the Himalaya has not slipped in the last 100 years. Roughly half could easily slip in a large earthquake, and part of the segment that hasn’t slipped seems to lie south east of where the October 8 earthquake occurred.

Salazar: Now about the risk posed to human centers of population is that increasing just because the population is constantly increasing?

Molnar: Of course, population is increasing. There are more people alive today than before. The standard argument is that there are more people alive today than have ever lived, and the risk goes up with the number of people.

In other areas, humans are having an impact on the surface of the Earth. But humans are not having any impact on the frequency or occurrence of earthquakes. The likelihood of an earthquake anywhere hasn’t been changed by any human intervention. In so far as risk to humans going up with time, that has nothing to do with the Earth. It has to do with where people choose to live and how many of them are living there.