Early on Saturday, a 7.8-magnitude earthquake struck an area of central Nepal between Kathmandu and the Pokhara, killing more than 3,000 people. The tremor also unleashed avalanches on Mount Everest. A powerful aftershock was felt on Sunday in Nepal, India and Bangladesh, and more avalanches were reported near Everest.
Our colleagues at the UK and Australian Science Media Centres gathered the following commentary:
Professor John Wilson, Executive Dean of the Faculty of Science and Engineering at Swinburne University of Technology, comments:
“Unfortunately, this was a tragedy waiting to happen – a 7.8 magnitude earthquake in a highly seismic zone striking a low socio-economic country with local construction that is highly vulnerable – a blend of abode construction, unreinforced masonry and poorly detailed reinforced concrete structures. Reviewing the damage, there were no surprises and very sadly the death toll will rise.”
Dr Gary Gibson, Principal Research Fellow in the School of Earth Sciences at The University of Melbourne, comments:
“The devastating Nepal earthquake illustrates the cause of sad frustration for seismologists. We can learn much from past earthquakes, but cannot predict future earthquakes. In 1991 I was working on an earthquake hazard study for Kathmandu. The previous large earthquake in the area was in 1934, and before that 1833. It was common local belief that a major earthquake would not occur for about 40 years.
“Little seismological data was available, and most of that was only since improved seismograph coverage in 1960. Nepal has a written history dating back about 1,000 years, mainly relating to royalty, but including references to major earthquakes. Collating data from several sources, local seismologist M.R. Pandey and I found devastating earthquakes with magnitudes greater than about 7.5 had occurred on ten occasions, on average about every 100 years! However the interval between them varied from 23 years (1810 to 1833) to 273 years (1408 to 1681). This is a very short recurrence interval for a devastating earthquake. We can say that there will be another earthquake in Kathmandu, we cannot say when, but on a scale of tens to hundreds of years it will be sooner rather than later.
“A large earthquake in the Himalaya region occurs at relatively shallow depth, so can cause more damage than the more common large offshore earthquakes on faults that dip deep underneath the adjacent continent, as in recent Japan, Chile or Indonesia earthquakes. In mountainous areas, widespread landslides, rockfalls and creation of weak temporary natural dams that then fail catastrophically can be a major problem. Unfavourable soil conditions in the Kathmandu region give increased likelihood of damage. The Nepal earthquake will be more similar to the 8 May 2008 magnitude 7.9 Wenchuan earthquake in Sichuan, China, which caused 69,000 fatalities and caused many landslides, or the October 2005 Pakistan earthquake.”
Dr Behzad Fatahi, Senior Lecturer of Geotechnical and Earthquake Engineering and a Chartered Professional Engineer at the Centre for Built Infrastructure Research, University of Technology Sydney, comments:
“A strong earthquake with magnitude of 7.8 struck Nepal on 25 April 2015 (06:11:26 – UTC) and the epicentre of the earthquake was 34km away from city of Lamjung and 77km away from Kathmandu. Several aftershocks have occurred since, with the largest having magnitude of 6.7 occurred 25hrs after the main earthquake and 19km away from City of Kodari (67km east of Kathmandu).
“The population growth rate in Kathmandu has been more than 3 per cent per year in the last 30 years. Kathmandu is located in a synclinorium tectonic valley controlled by faults in the central zone. Typical stratigraphy of the area shows deep layers of alluvium consisting of sand and clay deposits extended more than 100m below the ground surface. Thus it is expected that high rise buildings with high natural vibration period would be more influenced in the region.
“More than 80 per cent of Nepal has hilly/mountainous terrain, leaving less than 20 per cent in the Terai marshy lands and forests. Since the region is one of the most activate tectonic areas with the India plate converging with Eurasia plate (at a rate of more than 40 mm/yr on average) and intense rainfalls, large landslides are expected. The intense rainfalls occur typically in monsoon season in Nepal between June and September each year. Areas in the proximity of Tamakoshi River in central Nepal may be susceptible to large landslides after this recent earthquake and aftershocks. Areas close to natural slopes with intense precipitation rate should receive particular attention now.
“Nepal has more than eight major dams and hydropower stations; large embankment dams such as Kulekhani Dam (114m high) located 40km away from Kathmandu should be inspected immediately for possible cracks or flow slides. Large earthquakes can change the stress state in the embankment dams increasing tensile stresses and in some occasions decreasing the shear strength due to the increase in the pore water pressure.”
Commentary gathered by the UK Science Media Centre
Q1: Is there any more risk of aftershocks?
Prof Ian Main FRSE, Professor of Seismology and Rock Physics at the University of Edinburgh:
“Yes, almost all earthquakes are followed by aftershocks due to events being triggered by permanent redistribution of stress and strain associated with the source slip itself, and by a delayed response to the passage of the dynamic waves. Aftershocks tend to decrease both in numbers and amplitude with time, but could be large enough to be felt for a few weeks.”
Prof Sandy Steacy, Head of the School of Physical Sciences at the University of Adelaide, said:
“There will definitely be more aftershocks across the 100 km or so of the fault that slipped in the earthquake. The rule of thumb is that the largest is about one order of magnitude smaller than the mainshock and on Sunday [26 April] there was one of magnitude 6.7. However, it’s likely that there will be several more aftershocks of magnitude 5 or more, and probably a few of 6 or more.”
Prof David Rothery, Professor of Planetary Geosciences at The Open University, said:
“Yes. In the 24 hrs prior to 11:40 GMT on Monday 27 April there were 7 aftershocks greater than magnitude 4 on the Richter scale; one was 5.3. They were all shallow, mostly about 10 km deep. These generated ground shaking strong enough to feel in Kathmandu, and the M5.3 quake was felt on the Ganges plain in northern India. The frequency and strength of aftershocks will probably decrease over the next few days.”
Prof Mark Allen, Department of Earth Sciences at the University of Durham, said:
“Yes, undoubtedly. The slip on the April 25th earthquake will have increased the stress on other segments of the same fault, or of adjacent faults, such that they are more likely to rupture themselves. There may be earthquakes for months in the region, which likely relate to the main event.”
Q2: Why are aftershocks so deadly after a big earthquake?
Prof Main: “Generally aftershocks don’t have as big an effect as the mainshock. For example here the largest aftershock (Magnitude 6.7) contained 30 times less energy than the mainshock (M7.8). However, aftershocks can affect buildings that have already been weakened in the mainshock. They can also affect rescue operations, including triggering further landslides and blocking access roads, and of course their effects are more likely to be reported by journalists.”
Prof Steacy: “Aftershocks cause disproportionate damage because they affect buildings that have already been weakened in the mainshock. So a building that might have easily withstood a stand-alone M=5 earthquake may fall down when that event is an aftershock of a larger event that has caused earlier damage.”
Prof Rothery: “While mostly too weak to damage sturdy structures, aftershocks can topple structures that were weakened by the initial earthquake (putting rescuers at risk), or even cause additional landslides.”
Prof Allen: “The greater energy released in a big event like on April 25th in Nepal leads to a proportionally greater loading of adjacent faults. Two of the aftershocks themselves have magnitudes of 6.6 and 6.7 – significant events in their own right. The 2003 earthquake which devastated Bam, Iran, in 2003 had a magnitude of 6.6. It killed over 20,000 people.”
Q3: How much did the earth move?
Prof Main: “The maximum slip estimated by the USGS was 4m or so, or around 2m on average across a very shallow-dipping thrust fault on the order of 100 km long, originating at shallow depth (10-24km). The shallow depth and dip has led to strong ground shaking being experienced over a very wide area. The ultimate cause of thrusting in the area is the convergence of the hard Indian plate with the softer Eurasian plate, leading to shortening in the NNE-SSW direction and vertical thickening behind the suture. The Himalayas have been built up slowly by countless such earthquakes over millennia.”
Prof Steacy: “It’s likely that the earthquake occurred on the Himalayan Thrust fault, a plate boundary that separates the northern moving Indian sub-continent from Eurasia. The fault dips about 10 degrees to the north-northeast. The relative movement across the fault zone was on the order of 3m at its greatest, just north of Kathmandu.”
Prof Allen: “The fault plane is estimated to dip at 10 degrees northwards, i.e. towards Tibet. It is a thrust, meaning that the rocks on top of the fault moved southwards over the rocks underneath it, causing overall shortening of the Earth’s crust in the region. This is part of the plate-scale convergence of the Indian and Eurasian plates, which has been going on for 10s of millions of years. The thrust faults within the Himalayas are one of the main mechanisms by which this plate convergence is achieved.
“More motion is taken up by faults in Central Asia, in ranges like the Tian Shan and at mountain ranges at other margins of the Tibetan Plateau. These ranges include the Longmenshan, which had the Wenchuan earthquake in 2008 that killed over 60,000 people. Note that the magnitudes of the Wenchuan and Nepal earthquakes are similar – 7.9 and 7.8 respectively. It would be simplistic to say that Kathmandu was relocated by 3 metres: there may have been ~3 metres of slip on the fault at the earthquake nucleation point, at ~15 km depth (the hypocentre), but this dies out in all directions, including upwards to the surface.
“As the fault dip is only ~10 degrees, the line where it intersects the surface is tens of kilometres south of the epicentre. In the coming months people will no doubt study signs of this fault trace and rupture pattern. The US Geological Survey have produced a map that shows the earthquake parameters and context.”
Q4: Was Everest’s height changed?
Prof Main: “In theory the height of Everest should increase, but it’s quite far from the epicentre, so the actual static height change might be very small. At this distance it’s likely the snow avalanches were triggered by dynamic wave propagation rather than the permanent stress and strain changes that would be picked up in any change in height. We will know soon how if any detectable height change has occurred, when those working on satellite-based ‘synthetic aperture radar’ look at the difference in snapshots of the ground surface position before and after the event.”
Prof Steacy: “It’s unlikely that the height of Everest was changed. The main slip was west of Everest; the mountain was not directly above the fault plane. In addition, the dip of the fault is very shallow so 3 metres in a horizontal direction doesn’t mean much vertically.”
Prof Rothery: “I expect so, but by a few mm at most.”
Prof Allen: “It’s unlikely that the height of Everest has changed to any detectable amount because of the movement along the fault during the earthquake – Everest is too far away from the epicentre. There may be an effect if an avalanche has dislodged some of the snow cover on the summit.”
Q5: Why can’t we predict earthquakes?
Prof Allen: “We can’t provide a useful prediction because we don’t have enough information about the state of stress on any part of a fault, and what stress is required to make it slip in an earthquake. We are building more and more detailed pictures of past earthquakes and present strain in earthquake-prone regions, but these are not nearly detailed enough to predict when and where the next event will be. Nor do there seem to be useful “precursor” events, even though everything from radon emissions to animal behaviour changes have been proposed.”
Issued on 25/26 April:
Prof David A Rothery, Professor of Planetary Geosciences at The Open University, said:
“This morning’s earthquake in Nepal was both large (magnitude 7.9 on the Richter scale) and shallow (with a source only 11 km below ground). The shallowness of the source made the ground-shaking at the surface worse than it would have been for a deeper earthquake.
“Clearly there will have been casualties. I’ve seen pictures of poorly-constructed old buildings destroyed in Kathmandu, and I’m concerned that in this mountainous region there could have been landslides that might have destroyed or cut-off various remote villages.
“One factor that mitigated the severity of the shaking in Nepal, is that most areas are on solid bedrock. The shaking intensified on the north Indian plans near the Nepalese border, where the surface sands and silts naturally shook more strongly than solid rock.
“The cause of this earthquake was India’s slow-motion (5 cm per year) northwards collision into central Asia. This results in thrust-faultung and has thrown up the Himalayan mountain range.
“Other significant earthquakes in this region include the 1934 M8.1 Bihar, the 1905 M7.5 Kangra and the 2005 M7.6 Kashmir earthquakes. The latter two resulted in the highest death tolls for Himalaya earthquakes seen to date, together killing over 100,000 people and leaving millions homeless.”
Updated comment from Prof David A Rothery:
“My previous statement about the Nepal earthquake was written at 09:15 BST.
“Now at 18:45 BST the situation is clearer:
“The magnitude of the quake was been revised downwards from 7.9 to 7.8, and the depth revised from 11 km to 15 km. These are fairly trivial changes. The earthquake was still large and with a shallow source.
“In Kathmandu the newer buildings tend to have survived relatively well. Maybe seismic-resilient building codes have been adhered to. It seems to be old buildings that fared worse. There are ways to ‘retrofit’ old buildings to make them withstand earthquakes, but these are unlikely to be affordable in a poor country such as Nepal. It is probably a good thing that this happened on a Saturday when Nepal’s schools are closed, because school buildings (which tend to have large areas of unsupported roof) often fare particularly badly. However there are reports of school-children hurt during today’s quake in W Bengal.
“Examples of major school trajedies: 2008 Sichuan quake http://en.wikipedia.org/wiki/Sichuan_schools_corruption_scandal
“In Venezuela two schools, built in contravention of the local seismic building code, collapsed as a result of a M 7.0 earthquake on 9 July 1997 killing 46 students. In Algeria a M 6.8 quake on 21 May 2003 left 122 schools in need of rebuilding. School house tragedy was avoided in this case because the earthquake happened after the end of the school day, though 2,287 other people died.
“However, a M 6.4 earthquake in eastern Turkey on 1 May 2003 brought down the roof of a school dormitory, killing many children as they slept. A subsequent survey found that none of the local schools accorded with the 1998 Turkish seismic code, and the blame for this was placed on lack of resources to supervise building works, and on poorly trained architects and engineers.
“There were nine aftershocks exceeding magnitude 5.0 in the six and a half hours after the initial quake. These are dangerous for survivors still trapped in collapsed buildings, and put rescuers at risk too. Smaller aftershocks will probably continue for several days, but it can be hoped that there will be few in any above magnitude 6.
“We have by now all heard of the avalanche near Everest base camp. The biggest issue that I am still wondering about is the extent to which landslides may have affected remote settlements. For example, the 2005 M 7.6 Kashmir earthquake caused several thousand landslides, one of which obliterated the village of Dandbeh killed a thousand people http://pubs.usgs.gov/of/2006/1052/pdf/ofr-2006-1052.pdf If this situation is repeated in Nepal, the death toll could be far higher that that reported from Kathmandu alone.”
Dr Ilan Kelman, UCL Institute for Risk and Disaster Reduction, said:
“This earthquake is the nightmare scenario which we have long discussed and wondered if we could make major improvements before a catastrophe. Nepal has some of the world’s best people and initiatives for community-based seismic risk reduction and earthquake education. They have taught me plenty. But the country has also suffered terrible conflicts, poor governance, and heart-wrenching poverty, all of which created and perpetuated the vulnerability which has been devastatingly exposed during the shaking. The pictures and reports emerging do not bode well for other earthquake-prone cities with similar vulnerabilities.”
Updated comment from Dr Ilan Kelman:
“The videos and photos of the devastation show the chilling truth that earthquakes don’t kill people, collapsing buildings do. We have long known how to construct buildings which do not fall down in earthquakes, but that knowledge is not always applied. Despite the heroic efforts of our Nepali friends and colleagues over the years, including retrofitting schools, political not technical factors slowed the work. As people are still being pulled from the rubble, let’s recognise that this was not a natural disaster in order to work towards overcoming the politics which causes disasters.”