Geoscientist Online

John Bell believes that the accepted explanation for the destruction suffered by the Terracotta Army of China’s First Emperor is mistaken. Seismicity, not human agency, was to blame…

Geoscientist 20.4 April 2010

In 247 BC, 13 year-old Crown Prince Zheng became King of the state of Qin. Work on his necropolis began immediately, a city for him to inhabit in the next world. By 221 BC he had conquered six rival states, created a nation that became known as China, and declared himself Qin Shi Huang Di: The First Emperor of Qin. The necropolis, still incomplete when he died in 210 BC, had taken 38 years to build, and spread over more than 56 square kilometres. The First Emperor was entombed beneath a burial mound similar in size to the larger pyramids in Egypt. Originals or replicas of his most precious possessions were buried in over 600 pits surrounding the mound. Seven thousand life-size Terracotta Warriors stood guard in rank and file, 1500 metres to the east, five metres underground. The Emperor had created what are now the most widely recognised historical figures in the world.

Qin Shi Huang Di is a legendary figure. Stories of his life and deeds continue to fascinate and have spread across the world, from Sima Qian’s Shi Ji (Records of the Grand Historian) written c. 100 BC, to the movie The Mummy: Tomb of the Dragon Emperor and the Tan Dun opera The First Emperor.

Today’s visitors to the Museum of the Terracotta Warriors, Horses and Chariots (Lintong County near Xi’an) are given explanations of the near total destruction that has taken place since 210 BC. Sometimes, the devastation is passed over. The books and other documents produced by the British Museum for the 2008 exhibition The First Emperor: China’s Terracotta Army include no written references to the destruction of the figures, or the processes used to stick them back together for their 21st Century exhibitions.

However, I believe I have solid and convincing evidence that reveals the truth about these figures’ destruction for the first time. The answer lies in the geology and seismology of Shaanxi province and the remarkable earthquake records kept by Chinese officials since 1177 BC.

Discovery and excavation

When completed, the necropolis included a pyramid-shaped burial mound constructed from the area’s loess clay. It is claimed that it stood 115 metres high in 210 BC, though its current dimensions are 76 metres high on a base 355 x 345 metres. The necropolis had tombs, palaces and halls, inner and outer walls and subterranean pits containing items to make the Emperor feel at home in his next life. Many of the pits, including those containing the Terracotta Warriors and Horses, lie outside the walls. Farmers first unearthed fragments of the Warriors in 1974, leading to the discovery of thousands of figures and other items.

Excavation has continued, but has not been continuous. The most recent dig in Pit 1 commenced in June 2009. The front ranks from Pit 1 have been glued together from excavated pieces. Further back there is a clear view of fragments that have been fully or partially exposed by the removal of hard-packed covering earth. The pit is 230m x 62m, with lengthwise rammed-earth walls dividing it into 11 corridors.

Pit 2 has a different appearance. Substantial areas of timber support pillars and roofing have been uncovered. Terracotta figures lie in pieces, made visible by the partial removal of some timbers. Warriors weigh from 110 to 300kg; horses weigh 1000kg. These massive objects have been smashed to pieces. Horses are seen with limbs torn off. Warriors lie in every direction.

Finds from other areas of the necropolis include; two half-size decorated bronze chariots, each pulled by four bronze horses (1980); terracotta acrobats (1999) and 44 life-sized bronze water birds with 15 terracotta attendants, possibly dancers and musicians, (2001). The bronze chariots and horses were in a wooden box measuring 6.8m x 2.1mx 2.0m, eight metres below ground in a pit 20 metres east of the mound.

All of these artefacts were seriously damaged, bronze much less than terracotta. Museum displays show water birds, acrobats and musicians with images of repair processes. The bronze horses and chariots were found in 1555 pieces, taken to Beijing for restoration, and now occupy a central position in the museum.

Construction and destruction

An official notice reads:

This explanation raises many issues and questions. Pit 1 contains unburned combustible material, including chariot wheels and parts of weapons. Pit 2 contains large areas of unburned roof timbers. Pit 2 is divided into smaller ‘cells’ that would prevent fire from spreading. How could fires spread through pits with limited oxygen supply and egress for the products of combustion? There is damage and destruction in every underground part of the 56km2 necropolis. How long would it take to dig down to the pits to destroy the figures? How could the rebels survive the fires underground in pits up to 230m in length, 5m wide and containing Warriors, Horses and Chariots in close formation? Terracotta is fired clay, baked at 950°C to 1050°C. It is fire resistant - and strong in compression.

My first visit to see the Warriors was in February 2008 during the Chinese New Year holiday. My doubts about the underground fires were soon put aside, until events brought them back into focus.

Sichuan Earthquake 2008 AD - Shaanxi Earthquake 1556 AD

When the Sichuan earthquake struck at 2:28pm on 12 May 2008, I was at home from school in Xi’an. It took 15 seconds for us to realise what was happening. Wall tiles crashed to the floor in the bathroom, and with shouts of ‘earthquake’ we rapidly descended the emergency stairs to ground level.

As I left the stairway, I turned and saw a tower crane topple onto the roof of a construction some 25 stories high. More than 90 people died in Xi’an that day. Inanimate objects were also affected:

I used Google Earth to determine the distance from the Magnitude 7.9 Sichuan epicentre to our apartment (630km) and the Warriors (688km). If such a distant earthquake could damage the terracotta warriors, what would be the effect of a closer one?

Questions and discussions with students and friends in Xi’an produced the same response: “we never have earthquakes”. Yet an Internet search for ‘Xi’an earthquake’ followed by ‘Shaanxi earthquake’ gave an astonishing result. Pages of sites came up, detailing, ‘… the most devastating earthquake in human history…’ at midnight on 23 January 1556 AD. The ‘Shaanxi Earthquake’ (or ‘Huashan/Huaxian Earthquake’, Magnitude 8.0-8.3, epicentre intensity 11) killed 830,000 people. It devastated a vast area of China, wiping our whole towns. Hundreds of thousands were buried alive in their homes in hand-dug caves in the loess plateau of Shaanxi and neighbouring provinces. After entering the coordinates with the USGS, I found that the epicentre was a mere 40km from the warriors.

A wider search revealed that earthquake records had been kept in China for over 3000 years and were without doubt the most comprehensive in the world. I obtained, ‘A catalog of historical earthquakes in China compiled from recent Chinese publications’, Bulletin of the Seismological Society of America (v 66.6, pp 2003-16), which lists earthquakes from 1177 BC. A simplified fault map based on LANDSAT images with historical (1177 BC – 1900 AD) and recent (1962 – 1974) data plotted separately allowed the authors to conclude that: “the most notable difference… is the apparent seismic quiescence of Western China in the pre-instrumental period. This is due entirely to the lack of data there’… ‘On the other hand, (there is) activity along well-delineated belts such as the Weihe graben in Shensi (Shaanxi) province”.

In one sense, the students and friends I had questioned were right. The data for the province record 35 earthquakes between 1177 BC and 1834 AD, with many coming in groups just a few years apart. From the perspective of people and communities, earthquakes separated by long periods of inactivity are effectively “unknown”. I followed up my single placemark plot with all 35 listed in the catalogue in Shaanxi and several more in neighbouring provinces. In addition to the 8.0 to 8.3 Magnitude event of 1556, nine earthquakes of M4.5 to 6.75 were recorded within 40km of the Terracotta Army site, in years BC 35, AD 600, 791, 879, 1487, 1559, 1563, 1569 (twice). The closest in distance was the 1563 event (M5.5, 7.2km away). When all these are plotted, it looks like someone has been taking pot shots at the necropolis, and often getting very close to a bull’s-eye!

Amateur seismology

In this region of northern China sizeable earthquakes are not frequent, indicating that the underlying Ordos Block is stable and solid. It lies north of the Tibet Plateau and the resulting interaction between the block and the northward moving plateau has produced significant seismic events, including the Haiyuan earthquake (M8.6, 1920). The block rotates counter clockwise, influencing the seismic activity in the Weihe graben basin that is the subject of this article. The geology and seismology of the basin has been the subject of considerable research; it is an active area.

In March 2009 a visit by friend Ed Garrett, a postgraduate student of geography at Durham University, gave my researches renewed impetus. With my wife, Sue, we travelled the 65km east from Xi’an to Huashan to ascend midway up the sacred mountain by cable car.

Approaching the Weihe graben floor, the effects of the 1556 event are clearly visible. Hou J et al. describe the range of techniques available to gain evidence for the occurrence and causes of the earthquake, concluding: “… we need to consider the potential for active faulting, and prepare for another possible large earthquake in the region, since the faults are active now.” (Journal of Structural Geology, vol. 20, no5, pp. 549-557, 1998). Zhang A et al. subdivide the fault zone into two and ascribe the 1556 event to the “…obvious activity of the piedmont fault of Huashan mountain”. They estimate the earthquake recurrence interval at 2000 to 2500 years for the fault and 2000 to 4000 years for the zones. (Quaternary International, vol. 25, pp. 25-31, 1995).

This paper has renewed relevance following a M4.8 event on 5 November 2009 at 07:23 (GMT +8) in Gaoling County, Shaanxi province. The epicentre was 16km from the Qin mausoleum and 5km from the Weihe (River Wei) and therefore in a central position towards the western end of the Weihe graben. Media reports suggested that the Terracotta Warriors were undamaged.

As well as affecting the Emperor’s terracotta guardians, earthquakes seem certain to have caused the reduction in height of the burial pyramid. The combination of high intensity seismic action, high riparian water table and loess closely resembles the New Madrid Fault Zone in the Mississippi valley. Close examination of photographic evidence, the results of ground-penetrating radar investigations and thermal analysis support the view that the mound has been shaken, liquefaction has occurred, and the upper section of loess originally forming a ‘true’ pyramid has collapsed into the burial chamber.

Research has confirmed the veracity of ancient historical records in China. There can be no doubt that the necropolis has been subjected to very strong earthquake action many times since 210 BC. It is my belief that the Emperor lay undisturbed for at least 175 years until the recorded earthquake sequence began in 35 BC with a M 5.0 event, recorded at a distance of 20km.

Sima Qian’s historical records of the events of the Qin Dynasty are very detailed, but do not include any raids by Xiang Yu’s ‘rebel army’ on the Terracotta Warriors. In fact, Sima Qian does not mention the Terracotta Warriors at all. This is not surprising, since modern accounts assert that no one knew about them until their rediscovery in 1974. It is inconceivable that the Warriors were unaffected by at least the 1556 Huashan earthquake, when above ground there was total devastation. The recent M 4.8 event at Gaoling caused no damage; but my estimate of its epicentre intensity is some 10,000 times smaller than the Huashan event. Huashan is 40km from the Warriors location, Gaoling 16km. I have attempted to calculate the effect of the earthquake waves as they passed through the underground pits, using estimates of the accelerations produced, but have decided that this would be taking amateur seismology too far.

I am hoping that this account will act as a catalyst for research.


* John Bell, 68, is a teacher from Durham (UK). He teaches physics at Jinling High School, Nanjing, PRC. He was a teacher and Centre Principal at the Cambridge Advanced Learning Centre in Gaoxin No. 1 High School, Xi’an, PRC, from 2007 to 2009. Following John’s diagnosis with Parkinson’s disease in 1998, he and Sue have continued to live actively, believing that challenging exercise is as good for the brain as it is for the rest of the body.