Tibet earthquake: 5.9 aftershock hits Baqen after deadly 7.0
Tibet earthquake update: a 5.9 aftershock hit Baqen on October 25. The latest toll, what seismologists expect next and why the plateau keeps cracking.
Why it matters
More than 12,000 people were evacuated on a cold, high plateau, and strong aftershocks can still bring down homes the main quake left cracked.
The strongest aftershock yet hit northern Tibet on Sunday, October 25, three days after a magnitude 7.0 earthquake killed 16 people in the same remote county, according to Chinese officials.
It struck Baqen County, in Nagqu city, at 11:25 GMT (7:25 p.m. local time). The U.S. Geological Survey (USGS) and the European-Mediterranean Seismological Centre (EMSC) measured it at 5.9; the China Earthquake Networks Center (CENC) said 5.7. Early reports mentioned no deaths, injuries or new damage. More than 12,000 people had already been evacuated, Xinhua said.
Four quakes in six minutes on Sunday evening
CENC logged a 4.7 at 7:24 p.m., the 5.7 a minute later and two smaller quakes by 7:30 p.m. Chinese media reports based on CENC data called the 5.7 the largest aftershock of Thursday’s quake so far, centered about 19 km from the original epicenter. USGS coordinates put them 27 km apart, well within a fault patch the USGS says is typically 50 km long for such a quake.
A 5.6 in neighboring Qinghai, reported by The Standard in Hong Kong from CENC data, does not appear in USGS or EMSC lists.
- 22 Oct, 14:33 GMT: A magnitude 7.0 quake strikes Baqen County at 10:33 p.m. local time, 18 km deep by USGS figures and 10 km by CENC’s.
- 23 Oct: China raises its national earthquake response to Level III and sets aside 80 million yuan (about $12 million) for relief and repairs.
- 24 Oct, 08:39 GMT: A 5.3 quake hits Zadoi County in Qinghai, which borders Baqen, CENC and the USGS report.
- 25 Oct, 11:24 to 11:30 GMT: Four quakes in six minutes, the largest 5.9 (USGS, EMSC) or 5.7 (CENC).
16 dead by China’s count, at least 20 by the exile government’s
Chinese officials say 16 people died. The Central Tibetan Administration, the exile government based in India, said on October 23 that at least 20 had died. We found no updated figures as of 19:00 GMT on October 25.
| Measure | Figure | Source, date |
|---|---|---|
| Deaths | 16 | Chinese officials via Xinhua, 23 Oct |
| Deaths (exile count) | at least 20 | Central Tibetan Administration, 23 Oct |
| Injured | 76 | Xinhua, 24 Oct |
| People evacuated | 12,489 | Xinhua, 23 Oct |
| Homes destroyed or damaged | 339, plus 20 livestock sheds | Nagqu city government, 23 Oct |
Sources: Xinhua, Central Tibetan Administration, CBS News/Associated Press.
Is a 5.9 the biggest aftershock to expect?
Båth’s law, named for the seismologist Markus Båth, says the largest aftershock is on average about 1.2 magnitude units smaller than the main shock, regardless of the main shock’s size, a rule of thumb seismologists still use. Peter Shearer of the University of California, San Diego, restated it in a 2012 paper in the Journal of Geophysical Research. Subtract 1.2 from 7.0 and you get 5.8. Sunday’s 5.9 sits almost exactly on the line.
That is an average, not a ceiling. Agnès Helmstetter and Didier Sornette argued in Geophysical Research Letters in 2003 that the 1.2 figure comes from the statistics of how mainshocks and aftershocks are defined, not from any special physics. The USGS says aftershock magnitudes do not shrink as a sequence ages; only their rate does.
The rate is falling. CENC had counted 413 aftershocks by 7 a.m. on Friday, about eight and a half hours after the main shock, and 949 by 2 p.m. on Saturday. That is roughly 49 an hour at first and 17 an hour over the next day, in line with the USGS guide that day one can bring about 10 times as many aftershocks as day ten. CENC experts said felt quakes remained possible in the coming days, People’s Daily reported.
Why one agency says 5.9 and another 5.7
The USGS says gaps of 0.2 or 0.3 units between agencies are common even for well-recorded quakes, since each uses its own stations, formulas and magnitude types. A 2007 study by Peter Bormann and colleagues in the Bulletin of the Seismological Society of America traced the systematic differences between China’s national network and the USGS to instrument response, measurement time windows and calibration.
Baqen shows it. The USGS rated two aftershocks in the first four hours at 5.2 and 5.0 on its short-period body-wave scale; by CENC’s count, none had reached 5.0 by noon on Friday.
Was it felt in Nepal or India?
We found no reports that either quake was felt there. By USGS coordinates the epicenter is about 960 km from Kathmandu and 750 km from Gangtok, in India’s Sikkim state, and the USGS shaking model put everyone exposed inside China. The January 2025 magnitude 7.1 in Dingri County, about 80 km north of Mount Everest, did shake buildings in Nepal, Bhutan and India.
A plateau pulling itself apart
The textbook story of Tibet is a collision. India converges with Eurasia at about 45 mm a year, the USGS says, and that collision raised the Himalayas and the plateau, as Peter Molnar and Paul Tapponnier argued in Science in 1975. Thrust quakes on the Himalayan front, like Nepal’s 7.8 in 2015, come from that squeeze.
Thursday’s quake came from something else. The plateau is so high that it now spreads east and west under its own weight, which the USGS compares to honey spreading after it is poured. That makes normal faults, where the crust pulls apart and one block drops. The USGS and CENC experts both describe Baqen as a normal-fault quake, near the Baqen-Leiwuqi fault, and the USGS says the same of Dingri in 2025. Those are the only two quakes of magnitude 7 or more on the plateau since 2025 in the USGS catalog. Both came from the stretching, not the head-on collision.
Research suggests that stretching is easy to underestimate. In a 2010 study in Geophysical Journal International, John Elliott and colleagues found normal-fault quakes account for only 3 to 4 mm a year of east-west extension, 15 to 20 percent of what GPS measures across the plateau. Only five of the eight quakes they modeled left a clear mark at the surface, so surface mapping alone understates the stretching. Before Thursday, CENC experts said, no quake of magnitude 7 had been recorded within 200 km of Baqen, though nine reached 6 or more.
Distance from towns decides the toll: the 2010 Yushu quake, 272 km east-northeast of Baqen, was a smaller 6.9 but killed more than 2,500 people, the USGS says.
Watch for: CENC’s aftershock counts. A falling hourly rate with nothing above 5.9 is what a normal sequence looks like; a quake larger than Thursday’s 7.0 would mean it was a foreshock.
