Assessing archaeological evidence for seismic catastrophies

Archaeologists frequently ascribe “destruction” layers uncovered during excavation to the effects of earthquake-induced phenomena. Undoubtedly, ancient societies that lived in active seismic zones were as plagued by seismic destruction as we are today. However, other geological phenomena can account for many observed features. This paper reviews the geologic and archaeological background for assessing the evidence for prehistoric earthquake destruction.     

Prehistoric human settlement patterns in a tectonically unstable environment: Outer shumagin islands,southwestern alaska

Human population movements into and around the outer Shumagin Islands of southwestern Alaska during the last 5000 years, and temporal gaps in Shumagin habitation correlate inversely with geologically inferred prehistoric earthquakes. Clusters of inferred seismic activity correlate with temporal gaps or with small, sparsely distributed archaeological sites; periods of relative seismic quiescence coincide with settlement florescence. Variations in terrace and archaeological site heights across the study area indicate differential uplift of the outer Shumagin Islands when corrected for the Holocene sea level rise and post-glacial isostatic adjustments.     

The geoarchaeology of Iron Age “moated” sites of the Upper Mae Nam Mun Valley,N.E. Thailand. I: Palaeodrainage,site–landscape relationships and the origins of the “moats”

The study focuses on the “moated” Iron Age sites of N.E. Thailand, first identified as significant prehistoric settlement sites in the 1940s from aerial photography. Two more recent photograph sets are used to map the surficial geology and prehistoric site distribution for a study area west of Phimai, N.E. Thailand, with a focus on site–landscape relationships and, in particular, relationships between site location and form and patterns of palaeodrainage. The derived record of the surficial geology reflects several phases of palaeodrainage, characterized by differing locations and types of former river channels. Of note is the differentiation between a recent period in which river channels, including those presently active, are single-string meandering channels, and an older period of broad belts of meandering multistring channels. The prehistoric site distribution correlates closely with the older drainage, and for many, the encircling channels (the “moats”) are closely associated with former river channels. These relationships provide a critical and novel model for site distribution; several implications arise, supported by emerging field evidence, and introducing issues for archaeological debate: (i) there is no need, as has been done in the past, to invoke prehistoric artificial forms of drainage associated with the sites; (ii) the definition of the encircling channels as “moats” is seriously called into question; and (iii) the inferred geomorphological evolution of the floodplain implies past changes in environmental parameters such as run-off, climate or biophysical environments. Since the sites are all located in or beside ancient meander belts, these parameters should now be introduced into archaeological discussions regarding the establishment, history, evolution, and abandonment of the Iron Age sites. Methodologically, this article illustrates the need to be aware of the complexity of aerial photograph interpretation in archaeological survey, showing that careful analysis of aerial photograph information may have a significant impact upon the modeling of prehistoric interpretations. Further stratigraphical studies will be reported subsequently, and will refine the models presented here. © 1999 John Wiley & Sons, Inc.     

An archaeoseismic approach and method for the study of active strike-slip faults

The use of archaeology to study earthquake hazards provides a human dimension to an issue of modern societal concern. We developed an archaeoseismic approach to the study of prehistoric earthquakes on active strike-slip faults. This approach employs a combination of standard archaeological and paleoseismic techniques. We have successfully applied this approach and its attendant methods to an archaeological site that straddles and has been offset by the San Andreas fault in northern coastal California. Resultant fault parameters, including cumulative rate of slip and timing of the penultimate event, are comparable to results of strictly paleoseismic investigations at other sites on this fault. The archaeoseismic approach furnishes a number of advantages over geologic studies in terms of the availability and number of potential study sites, the abundance of datable materials, and the array of potential piercing features with which to constrain fault history. © 1997 John Wiley & Sons, Inc.     

Rates of fluvial sedimentation: Implications for archaeological variability

Rates of fluvial sedimentation exhibit spatial and temporal variability that is important with respect to the study of prehistoric archaeological sites. Sedimentation rates within fluvial basins vary in response to internal dynamics, geomorphic controls, and external factors including climate and tectonics. Fluvial rates of sedimentation may be estimated using sedimentary, pedogenic, biogenic, and radiometric evidence. Holocene rates of sedimentation vary by three orders of magnitude, as shown by radiometric data from numerous localities in North America. Sedimentation rates define rates of matrix accumulation in archaeological sites. These act as controls on site construction and site modification. Artifact densities, spatial patterning, and preservation are all conditioned by rates of matrix accumulation. These dimensions of prehistoric sites are critical to the evaluation of variability within and between archaeological sites and to the study of past settlement/subsistence systems.     

Laterite nodules: A credible source of iron ore in iron age northeast Thailand?

The advent of iron metallurgy changed the course of prehistoric Thailand. There is, however, little information on the nature of iron production and the extent to which locally abundant and easily quarried lateritic iron may have been the ore source. For northeast Thailand archaeological sites, the presence of both iron slag and laterite iron nodules within and surrounding these sites has widely been assumed to represent localized smelting using locally sourced iron ore. This interpretation is, however, based on untested laterite‐equals‐ore and slag‐equals‐smelting equations. This paper tests these assumptions by examining the chemistry of the lateritic iron nodules from two archaeological sites. At one site, the laterite is not a credible ore source and iron working probably comprised only smithing. Local laterite at the other site may be a usable ore. We conclude that, in northeast Thailand, the use of local laterite as an ore source cannot be assumed and that presence of both laterite and slag is not necessarily evidence for prehistoric smelting. © 2010 Wiley Periodicals, Inc.     

The paleohydrological context of the Iron Age floodplain sites of the Mun River Valley,Northeast Thailand1

Many Iron Age sites on the flood plain of the Mun River in northeast Thailand are encircled by channels commonly known as “moats.” Also, the sites are closely associated with complex paleochannels of the river. A comparison between the seemingly human‐constructed moats and paleochannels provided an opportunity to assess the relationship between prehistoric human settlement and paleohydrological conditions. In this study, the results of physical, sedimentological, and geochemical analyses are used to characterize sediments deposited within the channels around the Iron Age site of Ban Non Wat and within a paleochannel at nearby Ban Non Ngiu. This allowed us to test the results of previous research that has suggested significant changes in the floodplain hydrology and the geoarchaeologically important conclusion that Iron Age human activity was associated with one particular paleohydrological phase. Our analyses broadly confirm the results of previous stratigraphic studies, but add detail regarding sedimentation processes. The evidence indicates that there are significant sedimentological differences within the complex of archaeological channel features, differences that provide critical evidence for the formation and sedimentation processes of the channels. More importantly, comparison between the archaeological features and the natural channel fills highlights the relationships between the archaeological sites and landscape. Drawing also on previously published chronological, geomorphological, and stratigraphical data, it is possible to place the sites into a floodplain hydrological regime that may have been unique to the Iron Age. Specifically, the moats may have been constructed in response to enhanced water availability on the floodplain. The sites, therefore, may reflect a human response to increased availability of water beyond the main river channels. This water supply, however, appears to have been short‐lived (centuries at most), and with its loss, the human adaptation to this enhanced natural resource became unviable. © 2008 Wiley Periodicals, Inc.     

秦岭北麓古滑坡分布特征与地震活动关系研究

地震滑坡是严重的次生地质灾害,也是改变地球表层地貌形态的重要力量,对地震诱发滑坡的规模、数量、类型等研究是地震危险性评价的重要手段,也是认识地震地质灾害的主要方法和途径。位于西安市以南的秦岭山脉北麓中段发育有一条长约50 km的古滑坡群,且基本与山前秦岭北缘断裂带平行展布,普遍认为该古滑坡群可能是由于秦岭北缘断裂的强震活动所诱发,但对于诱发地震的震级大小和影响范围尚没有细致研究。本文通过利用资源3号卫星立体影像制作的高分辨率数值高程模型(DEM)和高分辨率多光谱遥感影像对秦岭北麓古滑坡区域进行了详细的解译工作,并结合对部分古滑坡体进行了野外调查,制作了详细秦岭北麓古滑坡分布图。结果表明:解译出了43处古滑坡,主要集中分布在70 km×10 km的范围内,总滑坡面积是16.57 km2。通过利用地震震级与滑坡面积频度分布的关系分析了诱发秦岭北麓古滑坡群的地震规模,得到了诱发秦岭北麓古滑坡群的地震震级应在7.6~8.1之间。并结合区域地震构造环境以及与现代地震诱发滑坡事件的对比,认为秦岭北麓具有发生7.5级以上地震的潜在能力。该研究对认识现今秦岭北麓古滑坡的成因提供了定量化的数据支持,也对理解秦岭北缘断裂的地震危险性具有重要的现实意义。     

藏南安岗地堑的史前大地震遗迹、年龄及其地质意义

地表调查发现, 沿近南北向亚东-谷露裂谷中段的安岗地堑存在地震大滑坡、多世代断层崖和断层崩积楔等多种类型的史前大地震遗迹.进一步的观测和年代分析表明: 该区的古地震滑坡体至少存在新、老两期, 其中规模最大的"尼续大滑坡体"应该是最新一次大地震所形成.该区T₁到T₆各阶地的形成时代从新到老分别为7.7~2.1 ka、11.0~10.5 ka、17.6~12.1 ka、25.7~22.9 ka、58.4~70.6 ka和130~150 ka, 它们沿主边界正断层的平均垂直断距依次为2.8 m、6.1~7.9 m、10.3~12.5 m、16.6~19.0 m、28.0 m和76.0 m.其中T₁和T₂阶地上的断崖剖面揭示, 最近两次大地震发生在距今约5.8±1.0 ka和2.4±0.2 ka.综合分析认为: 安岗地堑的大地震活动具有较明显的丛集性特征, 并且在距今约23~26 ka以来一直处于大地震活跃期, 期间的断层垂直活动速率为0.8~1.3 mm/a, 大地震的原地复发间隔大致为3.3~3.6 ka, 特征地震的矩震级为7.0~7.2, 推算整个尼木地堑群的大震复发间隔最短可能只有约1.0~1.2 ka.研究结果指示, 藏南裂谷的大地震活动性明显比藏北的近南北向正断层更显著.      

Formation processes affecting submerged archaeological sites:An overview

Although the formation processes operating on submerged archaeological sites are just as varied as those affecting terrestrial ones, nautical archaeologists have not yet devoted much attention to them. Most studies to date are concerned with formation processes at particular sites. This article provides an overview of the major depositional and postdepositional formation processes affecting underwater sites. The most obvious depositional process is shipwreck, which takes several different forms. Submerged sites may also be formed by the drowning of coastal areas due to tectonic or eustatic sea level changes. In these cases, rapid submergence preserves sites better than slow inundation, which allows time for waves and currents to tear the site apart. For both shipwrecks and coastal sites, once submergence occurs, the single most important factor for preservation is rapid burial by sediment. A cover of sediment protects both the artifacts themselves and their spatial patterning from destruction by water and marine organisms. Once deposited, underwater sites are subject to modification by both cultural and natural processes. The best understood postdepositional processes include salvaging, treasure hunting, and destruction by marine borers. Others, such as dredging, construction, and bioturbation, have hardly been investigated at this time. Archaeologists need to devote more attention to the effects of marine animals that live in close association with the seabed, as well as marine plants, whose roots may disturb sites located in shallow water. From this study it is clear that maritime archaeologists must consider formation processes when planning projects, rather than thinking of underwater sites as simply “time capsules.” © 1999 John Wiley & Sons, Inc.     

Large earthquakes and the abandonment of prehistoric coastal settlements in 15th century New Zealand

This paper reports on the effects of large earthquakes and related events, such as tsunamis, on prehistoric coastal settlements in New Zealand. It is based on field observations at several well‐established archaeological sites around the Cook Strait region and on literature reviews. We identify three broad periods of seismic activity in New Zealand since human occupation of the islands: 13th century, 15th century, and the 1750s to 1850s. The most significant, from a prehistoric human perspective, is the 15th century. Using examples from the Cook Strait region, we suggest that the abandonment of coastal settlements, the movement of people from the coast to inland areas, and a shift in settlement location from sheltered coastal bays to exposed headlands, was due to seismic activity, including tsunamis. We expect similar patterns to have occurred in other parts of New Zealand, and other coastal areas of the world with longer occupation histories. © 2003 Wiley Periodicals, Inc.     

The Ohalo II prehistoric camp (19.5 Ky): New evidence for environmental and tectonic changes at the Sea of Galilee

Combined archaeological data, shore surveys, and aerial photos of submerged sediments in the Sea of Galilee provide new insights into environmental and tectonic events, their dating, and their impact on the Ohalo II prehistoric camp (ca. 19,500 yr B.P.) and its surroundings. The Ohalo II waterlogged campsite contains excellently preserved brush hut remains and other in situ features, all embedded in late Pleistocene lacustrine strata. The findings indicate relatively short occupation of the site, not more than months or several years at a time. The high quality in situ preservation of delicate organic materials, as well as the short occupation period, suggests a quick and gentle burial by fine sediments. The evident fast submergence (water level rise of the Sea of Galilee) could have been the result of climatic fluctuations towards the end of the last glaciation and/or small‐scale tectonic subsidence. The site is located on a tectonic block formed in the western fault belt of the Dead Sea Rift. We present new evidence of post‐occupational folding of the late Pleistocene strata and recent tilting and faulting. A westward tectonic tilt may have caused the blockage of the old Jordan River outlet after A. D. 1106. Excellent preservation of the fault traces to the east of the site is attributed to the young age of the displacement on the fault. The last displacement apparently post‐dates the blockage of the old Jordan River. © 2002 Wiley Periodicals, Inc.     

Preliminary environmental historical results to reconstruct prehistoric human-environmental interactions in Eastern Hungary

Palaeoenvironmental research is playing an important role in recent archaeological investigations. We present preliminary results of geoarchaeological analyses conducted at a palaeochannel located between two prehistoric archaeological sites in eastern Hungary. The study area lies within the Körös River Basin in Békés County, a region of intensive human occupation beginning in the Neolithic, ca. 7550 BP, and represents only the second palynological analysis done in conjunction with archaeological investigations and adjacent to an archaeological site in the Körös region. Pollen from an environmental monolith was used to reconstruct the local vegetation composition and the human impact on arboreal and non-arboreal vegetation near the archaeological sites. Sediment analyses helped to reconstruct hydrological activity and human impact on the local palaeochannel. Results indicate that activity from the Neolithic onwards played an important role in local environmental change, including increasing sedimentation and deposition of organic matter in the local waterway, some forest clearance and a shift from primarily arboreal vegetation to more grasses on elevated surfaces. The trophic status of the local channel changed several times during the Holocene. In addition, indications that groundwater levels may have been fluctuating during the period of human occupation, when combined with the other changes in the area, provide a possible partial explanation for changing settlement patterns.     

Geological processes and site structure: Assessing integrity at a Late Paleolithic open‐air site in northern France

The difficulty of reading the archaeological record of caves and rock shelters is becoming increasingly obvious to Paleolithic researchers. Although some open‐air sites are thought to avoid such taphonomic complications, interpreting their archaeological record is less straightforward than assumed. Postdepositional processes may obscure structure in configurations of features and artifacts. Recently developed techniques for the excavation and analysis of Paleolithic cave sites can be applied to open‐air sites before spatial patterning is interpreted for inferences about prehistoric social organization. Analysis of the orientation of elongated artifacts on the occupation surface of the late Upper Paleolithic site of Verberie, France, is employed for evaluation of the integrity of site structure. Results indicate spatial structure largely undisturbed by geological processes. The lack of disturbance in the configuration of archaeological materials allows for a behavioral interpretation of those remains. The proposition that Verberie was a hunting campsite for initial carcass processing is supported. © 2006 Wiley Periodicals, Inc.     

Landforms and archaeological site location in the little Missouri badlands: A new look at some well-established patterns

Patterns in the distribution of archaeological sites in the Little Missouri Badlands of North Dakota have previously been attributed to prehistoric cultural preference for certain landforms. These patterns include the concentration of Plains Archaic sites in upland settings and the predominance of Late Prehistoric sites in the alluvial lowlands. An understanding of Holocene stratigraphy provides new insights into these patterns, and other spatial/temporal aspects of the archaeological record. These insights in turn illustrate the need for comprehensive stratigraphic studies prior to the formulation of prehistoric settlement models in highly eroded environments. © 1993 John Wiley & Sons, Inc.     

Archaeoseismological evidence of past earthquakes in Rome (fifth to ninth century A.D.) used to quantify dating uncertainties and coseismic damage

The transformation of Rome during the Late Antiquity and the Early Middle Ages has been investigated by archaeologists and historians. Social and political changes are the main aspects which led to a progressive modification of the urban framework; abandonment, spoliation and transformation of buildings are quite diffused as documented by the archaeological literature. The consequence of these practices is a higher vulnerability of the buildings which, from the seismological point of view, played a main role in increasing the effects of seismic shaking. A number of earthquakes have struck Rome during the period of investigation (fifth to ninth century A.D.), known from historical sources: 443, 484–508, 618, 801, 847; in some cases (443, 484–508, 801) damage has been documented. In contrast, the archaeological sources characterise collapse layers and evidence of destruction at different sites with changing and not always conclusive chronological constraints. Consequently, collapse and destruction have been alternatively attributed to the above-mentioned earthquakes. Through a geoarchaeological and stratigraphic analysis of potentially coseismic collapse units, we want (1) to describe the archaeoseismic evidence derived from recent excavations and from the available literature (e.g. Piazza Madonna di Loreto, Piazza Venezia, Palazzo Valentini Crypta Balbi, Colosseo, Basilica Hilariana, Basilica di Santa Petronilla, Santa Maria Antiqua,…); (2) to discuss the chronological problems and the uncertainty of attribution of the collapse units to known historical earthquakes; (3) to discuss the earthquake damage exaggeration due to erroneous attribution of seismic origin to the evidence of destruction derived from archaeological data. Finally, we will infer the role that earthquakes may have had on the development of the urban landscape in the fifth to ninth century A.D.     

Tidal marsh stratigraphy, sea-level change and large earthquakes, i: a 5000 year record in washington, U.S.A.

Many of the estuaries of the Pacific Northwest of the U.S.A. and Canada contain stratigraphic sequences typified by alternating peat-mud couplets. Recent studies in this region interpret such couplets as the product of repeated large (magnitude S or 9) earthquakes on the Cascadia subduction zone. The resultant pattern of land-level movements is described by a model, the ‘earthquake deformation cycle’, of coseismic land subsidence followed by land uplift during interseismic strain accumulation. However, peat-mud couplets similar to those recorded in the Pacific Northwest are found on other less tectonically active temperate-latitude coasts, such as northwest Europe and the Atlantic coast of the U.S.A., where they have been interpreted as the product of non-seismic coastal processes. In this paper we apply the methods and scientific framework common to sea-level investigations in northwest Europe to a sequence of peat-mud couplets recorded in the lower Johns River, an estuary in southern Washington, to provide a test of the ‘earthquake deformation cycle’.Stratigraphic investigations of the intertidal sediments along the lower Johns River, using lithological, pollen, diatom and foraminiferal data, show evidence for eight coastal submergence events during the last 5000 years. To evaluate the ‘earthquake deformation cycle’ we assess the lateral extent of peat-mud couplets, the synchroneity of submergence, the presence of tsunami deposits accompanying submergence, and the suddenness and amount of submergence. Each submergence is shown to be accompanied by changes in coastal sedimentation broadly commensurate with those predicted by the ‘earthquake deformation cycle’, demonstrating the continued intermittent seismic activity of the Cascadia subduction zone throughout the mid and late-Holocene.Quantitative analyses of contemporary and fossil biostratigraphic data, using TWINSPAN and Detrended Correspondence Analysis, enable us to estimate the magnitude of submergence accompanying each peat-mud couplet. One event was accompanied by submergence of about 1.5 m or more, four events by intermediate submergence of about 1±0.5 m, and a further three events by submergence of <0.5 m. There is evidence for non-seismic relative sea-level rise prior to two of the eight submergence events, but for at least the last 3500 years the magnitude of relative sea-level rise has been less than the combined influence of sediment accretion following submergence and interseismic land uplift.     

全球海啸地震震源机制解及海啸成因探讨

海啸作为五大海洋自然灾害之一,严重威胁着人类生命财产安全。近些年来,国内外学者对地震海啸进行了大量研究,主要针对海啸的生成、传播、爬高和淹没的数值模拟,以及古海啸沉积物进行研究,但是对于海啸地震震源机制的研究还比较欠缺,尤其是缺乏对震级小于6.5的海啸地震的研究。针对我国的地震海啸研究现状,强调震级小于6.5地震引发海啸的问题不容忽视。本文归纳整理了全球766次地震海啸,利用三角图分类基本法则对海啸地震震源机制解进行分类,并对其中341个发生在1976年后的海啸地震进行震源机制解分析,对其中633次海啸浪高进行统计学方法分析研究。本文认为逆冲型、正断型、走滑型和奇异型机制地震均能引发海啸,逆冲型地震引发的海啸占比最大,震级小于6.5级地震引发的海啸的浪高也有高达10 m的情况,也能产生巨大破坏性。逆冲型、正断型、奇异型地震可直接引起海底地形垂向变化,进而引发海啸,而走滑型地震引发海啸则可能有两种原因,一种是走滑型地震并非纯走滑型而是带有正断或逆冲分量从而引发海啸,另外一种是走滑型地震引发海底滑坡导致海底地形变化进而产生海啸。从海啸地震震源深度分析,能产生海啸的地震震源深度97%以上都是浅源地震,主要集中在30 km深度以内,但是也有中深源地震海啸。本文综合海啸地震的震源特点、我国地理位置以及以往海啸发生的情况,认为未来我国沿海地区威胁性的地震海啸主要集中在马尼拉海沟和台湾海峡区域,在今后海啸预警方面需要格外重视这些区域,通过建立完善海啸预警系统来减少损失。     

Recognition of earthquake-related damage in archaeological sites: Examples from the Dead Sea fault zone

Archaeological structures that exhibit seismogenic damage expand our knowledge of temporal and spatial distribution of earthquakes, afford independent examination of historical accounts, provide information on local earthquake intensities and enable the delineation of macroseismic zones. They also illustrate what might happen in future earthquakes. In order to recover this information, we should be able to distinguish earthquake damage from anthropogenic damage and from other natural processes of wear and tear. The present paper reviews several types of damage that can be attributed with high certainty to earthquakes and discusses associated caveats. In the rare cases, where faults intersect with archaeological sites, offset structures enable precise determination of sense and size of slip, and constrain its time. Among the characteristic off-fault damage types, I consider horizontal shifting of large building blocks, downward sliding of one or several blocks from masonry arches, collapse of heavy, stably-built walls, chipping of corners of building blocks, and aligned falling of walls and columns. Other damage features are less conclusive and require additional evidence, e.g., fractures that cut across several structures, leaning walls and columns, warps and bulges in walls. Circumstantial evidence for catastrophic earthquake-related destruction includes contemporaneous damage in many sites in the same area, absence of weapons or other anthropogenic damage, stratigraphic data on collapse of walls and ceilings onto floors and other living horizons and burial of valuable artifacts, as well as associated geological palaeoseismic phenomena such as liquefaction, land- and rock-slides, and fault ruptures. Additional support may be found in reliable historical accounts. Special care must be taken in order to avoid circular reasoning by maintaining the independence of data acquisition methods.     

Liquefaction evidence for strong earthquakes of Holocene and latest Pleistocene ages in the states of Indiana and Illinois, USA

Sand- and gravel-filled clastic dikes of seismic liquefaction origin occur throughout much of southern Indiana and Illinois. Nearly all of these dikes originated from prehistoric earthquakes centered in the study area. In this area at least seven and probably eight strong prehistoric earthquakes have been documented as occurring during the Holocene, and at least one during the latest Pleistocene. The recognition of different earthquakes has been based mainly on timing of liquefaction in combination with the regional pattern of liquefaction effects, but some have been recognized only by geotechnical testing at sites of liquefaction.

Most paleo-earthquakes presently recognized lie in Indiana, but equally as many may have occurred in Illinois. Studies in Illinois have not yet narrowly bracketed the age of clastic dikes at many sites, which sometimes causes uncertainty in defining the causative earthquake, but even in Illinois the largest paleo-earthquakes probably have been identified.

Prehistoric magnitudes were probably as high as about moment magnitude M 7.5. This greatly exceeds the largest historic earthquake of M 5.5 centered in Indiana or Illinois. The strongest paleo-earthquakes struck in the vicinity of the concentration of strongest historic seismicity. Elsewhere, paleo-earthquakes on the order of M 6–7 have occurred even where there has been little or no historic seismicity.

Both geologic and geotechnical methods of analysis have been essential for verification of seismic origin for the dikes and for back-calculating prehistoric magnitudes. Methods developed largely as part of this study should be of great value in unraveling the paleoseismic record elsewhere.