Harrow and Hillingdon Geological Society

Previous Meetings

Tectonic archaeology in Japan:
volcanoes and earthquakes in the archaeological record

Gina L. Barnes

Introduction

The speaker, who is an archaeologist with a recent degree in geoscience, explained the tectonic situation of Japan with its history of earthquakes and volcanoes arising from the surrounding subduction zones.

Japanese archaeology differs from the standard European, with the main periods being:
• Palaeolithic 42,000 to 14,500BCE;
• Jomon hunter-gatherers 14,500 to 500BCE;
• Yayoi agriculturalists 900BCE to 250CE;
• Kofun chieftains and kings 250 to 645 CE;
• Ritsuryo (Nara) archaic state 645 to 1185

A lot of Japanese archaeology is post-hole archaeology since houses were built of organic material. The first monumental architecture was in the Kofun period in tombs for the rulers. Chinese-style architecture came into Japan with Buddhism as at Yakushiji (752) and Todaiji (753) temples. Medieval castles such as Himeji had the first real use of stone in buildings, particularly for the foundations. The wooden building with clay tile roof at Yakushiji had a central wooden pillar which was not connected to the floor, allowing the building to move around the pillar during earthquakes and is an early example of earthquake-resistant building.

Earthquake archaeology

Japan is exposed to subduction earthquakes since it has 2 oceanic plates (Pacific and Philippine) subducting under 2 continental plates (North American and Asian) with 2 triple junctions. Really bad subduction earthquakes occur about every 100 years or so but there are also lots of compressive earthquakes.

One of the most active areas is around Osaka Bay where the area is bounded by the Arima-Takatsuki and Median tectonic lines, east – west right-lateral strike-slip faults and the north – south Konda fault. The Kondayama tomb, which was examined by Akira Sangawa, a geomorphologist, is cut by the Konda fault.

There are 4 types of earthquake evidence, which differ in Japan from those in European archaeoseismology.

Japan
Fault displacement
Landslips
Cracks and fissures
Liquefaction structures
(Sangawa, 1995)

Mediterranean
Displacement
Building damage
Building deformation
Destruction layers
(Galadini and others, 2006)

Landslips have been found on mounded tombs, one example having 5 different landslips thought to be from 5 different earthquakes. Landslides are more feared in Japan than other natural disasters since they can strike everywhere in such a mountainous country. Fissures in tombs are also interpreted as earthquake evidence.

Fault cutting tombs
Liquefaction sand boils can be significant indicators. They may be truncated by erosion and subsequent deposition of cultural layers or cut such layers, enabling the dating of the earthquake. Liquefaction dykes can cut through archaeological sites and, unlike ditches, always connect to a sand layer below. They also always have fining upwards, while ditches tend to have fining downwards. Draw-in and ring structures, which involve the material around a liquefaction sand boil, may take cultural material with it. One liquefaction event accompanying a very large earthquake brought cobbles to the surface and so impressed the people that it resulted in the Yayoi cobble ritual. Sectioning liquefaction sites to see the inter-relationships between liquefaction structures and cultural material can be very useful.

Fault & liquefaction structures
Identification of stratigraphical deformation is achieved not by looking at individual sites but by examining the pattern through a region. The example was cited of the Kawachi lowland plain, a former delta east of Osaka, which allows reconstruction of the earthquake history of the area with 7 major earthquakes in recorded history and 2 in pre-history. The next Nankai/Tokai earthquake, which has a 70-year return period, is overdue.

Earthquake archaeological methodology has some positives:
• Structures identified in archaeological sites;
• Avoids misidentification as cultural;
• Monitor prehistoric behaviour towards natural disaster;
and some negatives:
• Assumption of seismites, particularly liquefaction structures;
• Sites reported individually without regional linkage;
• Little discussion of methodology.

Volcanic ash archaeology

While Japan is often thought of as a volcanic island, most of the mountains are folded mountains with some Quaternary volcanoes, mostly caused by the Pacific plate subducting under Japan, eg in Gumma Prefecture, with others caused by the Philippine plate in southern Kyushu. Volcanoes in Japan have various eruptive types but explosive eruptions of ash and pyroclastic flows are the norm. There are 109 active volcanoes in Japan, dating back to over 130,000 years ago.

In southern Kyushu, there are 4 volcanoes in a row, Sakurajima, a caldera-ridge volcano on the Aira crater, Satsuma, Ata, and Kikai. Aira had a caldera explosion at about 24,000BP. The Satsuma ash fall in 11,000BP resulted in about 1,000 years without occupation. Kikai exploded in 7,300BP. There are records in AD874 of house destruction by ash fall and, indeed, volcanic bombs and ashfall account for most fatalities caused by volcanoes in Japan, in addition to those caused by flooding resulting from ash falls damming streams.

In Gumma Prefecture is the Mount Haruna dacite cone and Mount Asama. Excavation has revealed paddy fields covered with tephra at different times, in the 4th century, early 6th century and in 1108. Ridge and furrow gardens at Kuroimine have been preserved beneath ash and there is preservation of horse hoofprints and harrow marks.

After eruptions, there are long periods without occupation because the high rainfall leaches minerals from the ash. Water collects in gels and does not allow humus to decay naturally resulting in very infertile soils.

The usual risk calculation is HAZARD X ELEMENT X VULNERABILITY, but this is usually ignored. One archaeological result was the finding of a woman and child who were escaping to the temple for shelter but were caught in a pyroclastic flow. There are 5 levels of volcano alert in Japan but they are not well publicised. The responsibility for both earthquake and volcano warnings and action is the Japan Meteorological Agency.

Back to Top