Open this page in a new tab

Aila

Aila Area J-East

click on image to explore this site on a new tab in Google Earth


Introduction
Introduction

Aila (aka Ailana), the ancient settlement at the head of the Gulf of Aqaba, occupied a strategic position on the Red Sea trade routes connecting Arabia, Egypt, and the Levant. Originally a Nabataean town, it was incorporated into the Roman province of Arabia in 106 CE and flourished as a Roman, then Byzantine, port and administrative center. The town sat along the Via Traiana Nova, and its role in facilitating transshipment from maritime to overland routes made it critical in imperial logistics and commerce.

By the 4th century CE, Aila had developed into a walled city with a rectangular fortification plan. Archaeological excavations have uncovered city walls, gates, towers, a street grid, and residential and commercial buildings, some rebuilt after earthquake damage. Aila became a diocese in the Christian period and was later expanded under the Umayyads, who established a new Islamic city adjacent to the Roman‑Byzantine one. Its importance continued into the Abbasid and Fatimid periods, before declining with the emergence of competing trade hubs.

Today, the archaeological site of Aila lies within modern Aqaba. Excavations from the 1980s–2000s by the University of Chicago (notably under Donald Whitcomb and S. Thomas Parker) revealed deep stratigraphy with evidence of repeated seismic destruction and rebuilding phases. These findings shed light on the resilience and transformation of Red Sea port towns across nearly a millennium.

Aqaba/Elat - Introduction Webpage

Maps, Aerial Views, Plans, Sections, and Photos
Maps, Aerial Views, Plans, Sections, and Photos

Maps

Normal Size

  • Fig. 2.15 - Location Map
    Figure 2.15

    Map of archaeological sites in greater Aqaba, Jordan region (modified after Brückner et al. 2002).

    Allison (2013)
    for Aqaba and environs from Allison (2013)
  • Map showing location
    Map of the modern city of Aqaba with ancient and medieval archeological sites

    JW: Apparently Ailana = Aila

    Whitcomb (1988)
    of Islamic Ayla and Byzantine Aila (aka Ailana) from Whitcomb (1988)
  • Fig. 2 Map showing
    Figure 2

    The succession of settlements at 'Aqaba, Jordan

    Whitcomb (1997)
    location of Islamic Ayla and Byzantine Aila from Whitcomb (1997)
  • Fig 2.8B Map showing
    Fig. 2.8 B)

    Map of Aqaba showing the three archaeological sites near the coast, some major roads, and location of cross faults and possible Aqaba fault strands.

    Niemi (2014)
    location of Area J-east from Niemi (2014)
  • Fig. 1b Quaternary Geologic
    Fig. 3

    Quaternary geologicmap of the region along the Dead Sea Transform north of the Gulf of Aqaba, showing the geometry offaults in the region of Aqaba, Jordan, and Elat, Israel (modified after Garfunkel 1970).

    Click on image to open in a new tab

    Thomas et al (2007)
    Map of the Southern Arava from Thomas et al (2007)

Magnified

  • Fig. 2.15 - Location Map
    Figure 2.15

    Map of archaeological sites in greater Aqaba, Jordan region (modified after Brückner et al. 2002).

    Allison (2013)
    for Aqaba and environs from Allison (2013)
  • Fig. 1b Quaternary Geologic
    Fig. 3

    Quaternary geologicmap of the region along the Dead Sea Transform north of the Gulf of Aqaba, showing the geometry offaults in the region of Aqaba, Jordan, and Elat, Israel (modified after Garfunkel 1970).

    Click on image to open in a new tab

    Thomas et al (2007)
    Map of the Southern Arava from Thomas et al (2007)

Aerial Views

  • Aila Area J-East in Google Earth
    Aila Area J-East

    click on image to explore this site on a new tab in Google Earth

Plans

Normal size

  • Fig. 2.15 Plan of Area J-east
    Fig. 2.15

    Faults exposed in the Roman Aqaba Project excavation Area J-East

    Niemi - Chapter 2 from Parker et al (2014)
    from Thomas et al (2007)
  • Fig. 8 Map/Plan showing
    Fig. 8

    Plan of the areas excavated by the Roman Aqaba Project, after the 1998 season. Note that the location of Nabataean/Roman Aila is in the area not surveyed by Meloy (courtesy of RAP).

    JW: Area B is in upper middle of map

    Dolinka (2003)
    location of Area B from Dolinka (2003:32)

Magnified

  • Fig. 2.15 Plan of Area J-east
    Fig. 2.15

    Faults exposed in the Roman Aqaba Project excavation Area J-East

    Niemi - Chapter 2 from Parker et al (2014)
    from Thomas et al (2007)
  • Fig. 8 Map/Plan showing
    Fig. 8

    Plan of the areas excavated by the Roman Aqaba Project, after the 1998 season. Note that the location of Nabataean/Roman Aila is in the area not surveyed by Meloy (courtesy of RAP).

    JW: Area B is in upper middle of map

    Dolinka (2003)
    location of Area B from Dolinka (2003:32)

Sections

Individual Sections and Photos

  • Fig. 4                
    Fig. 4

    Section drawing Wall 26 and Wall 48, showing earthquake damage and fault offsets.

    Thomas et al (2007)
    Section Walls 26 & 48 from Thomas et al (2007)
  • Fig. 5                
    Fig. 5

    . Stratigraphic sections of the south and north baulks of J-1, showing the faults.

    Thomas et al (2007)
    Section C from Thomas et al (2007)
  • Fig. 5                
    Fig. 5

    . Stratigraphic sections of the south and north baulks of J-1, showing the faults.

    Thomas et al (2007)
    Section A from Thomas et al (2007)
  • Fig. 5                
    Fig. 5

    . Stratigraphic sections of the south and north baulks of J-1, showing the faults.

    Thomas et al (2007)
    Section B from Thomas et al (2007)
  • Fig. 6                
    Fig. 6

    Section drawing of the north corner of Room 6, showing the relationship between the earthquake damage ...

    Thomas et al (2007)
    Section N corner of Room 6 from Thomas et al (2007)
  • Fig. 7                
    Fig. 7

    Section drawing and photo of the blocked archway and slip on fractures in it in the north wall of Room 4

    Thomas et al (2007)
    Section Blocked Archway N Wall Room 4 from Thomas et al (2007)
  • Fig. 8                
    Fig. 8

    Section drawing of the plastered north wall of Room 11. The second phase of plastering probably repairs diagonal tension fractures developed from the earthquake at the end of Phase 2. The corner wall join also shows repair. The west wall is collapsed westward, perhaps in a much later earthquake.

    Thomas et al (2007)
    Section N Wall Room 11 from Thomas et al (2007)
  • Fig. 9                
    Fig. 9

    Section drawing of the south wall of Room 12. This wall lies very close to the trend of the active fault that traverses the site. Several earthquake-induced structural damage features are noted, including a slump failure along the west wall, extensional and shear fractures, collapsed walls, and a wall join repair.

    Thomas et al (2007)
    Section S Wall Room 12 from Thomas et al (2007)
  • Fig. 10                
    Fig. 10

    Section drawing of repairs in the wall near the stairway and arch pillar in Room 4. Most of the west wall is intact, standing to a height of ca. 2 m.

    Thomas et al (2007)
    Section wall near the stairway and arch pillar in Room 4 from Thomas et al (2007)
  • Fig. 11                
    Fig. 11

    Section drawing and photos of the earthquake damage in Rooms 2 and 3.

    Thomas et al (2007)
    Section and photos of quake damage in Rooms 2 and 3 from Thomas et al (2007)
  • Fig. 11                
    Fig. 11

    Section drawing and photos of the earthquake damage in Rooms 2 and 3.

    Thomas et al (2007)
    Section S wall Room 3 from Thomas et al (2007)
  • Fig. 11                
    Fig. 11

    Section drawing and photos of the earthquake damage in Rooms 2 and 3.

    Thomas et al (2007)
    Section S wall Room 3 View W from Thomas et al (2007)
  • Fig. 11                
    Fig. 11

    Section drawing and photos of the earthquake damage in Rooms 2 and 3.

    Thomas et al (2007)
    Photo of collapse of S wall Room 2 view W from Thomas et al (2007)
  • Fig. 11                
    Fig. 11

    Section drawing and photos of the earthquake damage in Rooms 2 and 3.

    Thomas et al (2007)
    Photo of collapse of S wall Room 2 view W from Thomas et al (2007)

Composite Columnar Stratigraphic Section

Fig. 3

Schematic columnar stratigraphic section of the deposits at the J-East site, showing mudbrick tumble from earthquakes on floor levels, sand horizons, occupational levels, and earthquake event horizons.

Thomas et al (2007)

Photos

Tilted and Faulted Mudbrick Walls

Tilted Walls at Aila Jordan (southern Cyril Quake) Wall Collapse at Aila Jordan (southern Cyril Quake) Left

Tilted South Wall of Room 2 at Aila J-East

Right

Normal Faulting of a wall at Aila J-East

photos by Jefferson Williams


Paleoseismic and Archaeoseismic Chronology
Phasing

Stratigraphic Section

Composite Columnar Stratigraphic Section

Normal Size

  • Composite columnar stratigraphic section
    Fig. 3

    Schematic columnar stratigraphic section of the deposits at the J-East site, showing mudbrick tumble from earthquakes on floor levels, sand horizons, occupational levels, and earthquake event horizons.

    Thomas et al (2007)
    for the deposits of the J-east site from Thomas et al (2007)

Full Size

Fig. 3

Schematic columnar stratigraphic section of the deposits at the J-East site, showing mudbrick tumble from earthquakes on floor levels, sand horizons, occupational levels, and earthquake event horizons.

Thomas et al (2007)

Discussion

Thomas et al (2007) excavated and examined area J-east between 1994 and 2003. The J-East area is a multiphase site incorporating Early Islamic to Byzantine domestic occupation and a late third to fourth-century monumental mudbrick structure that has been interpreted as a church (Parker 1998a; 1999a; Mussell 2001; Rose 1998; Weintraub 1999) ( Thomas et al, 2007). This site, in the Roman-Byzantine town of Aila, is located ~500 m north of the modern shoreline of Aqaba and ~500 m NW of the Islamic town of Ayla . Thomas et al (2007) identified 6 or 7 earthquakes from the 2nd century CE onward in J-east and divided up the timing as shown the the table below.

Table
Phasing Table

Area J-East, Byzantine Aila, Aqaba, Jordan — Occupation and Earthquake Chronology (Thomas et al. 2007)
Phase                 Period Dates (CE) Paraphrased and Abridged Discussion
0 Nabatean
Early Roman		 begin 1st – early 2nd
  • The earliest occupation of Area J-East is represented by three phases of substantial Nabataean stone and mudbrick wall construction, built upon beach sands sterile of artifacts. Ceramic data (mainly early Roman amphorae [Peacock and Williams 1986], Terra Sigillata, and Nabataean painted fineware) associated with these structures suggest that occupation began in the first century A.D. and ended at the close of the Early Roman/Nabataean period, or the early second century based on Nabataean painted fineware (Az Zantur Type 3c; Bignasca et al. 1996: vol. 1, plan 11; Schmid 1997). - Thomas et al. (2007)
Earthquake VII Nabatean
Early Roman		 ~106 – ~114
  • These occupation deposits [Phase 0] were subsequently covered by a very thick layer of mudbrick collapse which contained whole or partial bricks visible in the section. The collapse dents the surfaces beneath, indicating a violent fall of the structures. Excavated in the RAP 2002 season, these layers were found to be in excess of 1 m in thickness. Excavators concluded that this collapse horizon could either relate to the Roman annexation of Nabataea in A.D. 106 or a possible earthquake at the beginning of the second century (Russell 1985). Although there is no specific literary evidence for such an earthquake (Schmid 1997: 419-20), a complete section of collapsed wall (Area J, Trench 11, Loci 227, henceforth J.11:227) might suggest earthquake destruction. There is also continuing debate about the degree of Nabataean resistance to the annexation that might have resulted in destruction by human agency in this period (Bowersock 1983: 78-82; Parker 1986: 123-24; Fiema 1987; Freeman 1996). No rupture for this possible earthquake (EQ VII) was documented in the present study because of the limited areas excavated to this depth (about 2 m asl). Furthermore, subsequent building and reuse of the surviving walls have appreciably masked the original geometry. - Thomas et al. (2007)

  • JW: This refers to the Incense Road Quake, which may be alluded to in the Talmud and by Eusebius.
1&2 Late Roman
Early Byzantine		 Late 3rd – Early 4th
  • In the late third to early fourth century A.D., a monumental rectangular structure was constructed over a thin layer of second-century A.D. aeolian sand and the Nabataean collapse layer. In places, this Late Roman building utilized extant Nabataean walls. The construction and subsequent history of the monumental structure were closely dated by several hundred associated coins and sherds of imported ceramic finewares. During the early fourth century, the monumental building was expanded and concluded with the final addition of Rooms 11 and 12 constructed after ca. A.D. 320. The upper sequences of floors contained Early Byzantine pottery of the mid to late fourth century. - Thomas et al. (2007)
Earthquake VI ca. 320 – 363 CE
  • Then the monumental mudbrick structure experienced fault rupture and collapse of some walls, producing a tumble horizon. The southern wall of Room 13 was ruptured by Fault D and the northern wall of Room 21 by Fault C. This tectonic shift caused substantial localized damage. Earthquake VI produced a total of 10 cm of left-lateral strike-slip measured across Fault C on Wall J.1:26, north of Room 21. This damage from the fault was repaired after Earthquake VI. The strike-slip of Fault D in EQ VI could not be measured because Fault D reactivated in subsequent Earthquakes V and IV. The total strike-slip measured along Wall J.1:53 is 30 cm. Since there was no repair to the wall, this suggests that the majority of the slip was caused by EQ VI. Similarly, the dip-slip could not be directly measured, but later releveling of the southwest corner of the monumental building indicates subsidence did occur. Elsewhere on the site, damage appears not to have been quite as severe, but seismically induced wall failures were repaired in the subsequent occupation phase. This seismic event must have occurred at some point in the mid to late fourth century A.D. but before the final extensive collapse of the complex in Earthquake V. - Thomas et al. (2007)
3 Early Byzantine ca. 320 – 363 CE
  • After Earthquake VI, modifications were made to the monumental building, including releveling, blocking doorways, and wall repairs. During this period, mixed fill deposits were used to raise the floor level of sunken areas of the western part of the site, presumably caused by slip in Earthquake VI. A major effort was made to repair damaged walls, and a number of paved floors were also constructed to improve the monumental structure. Repair during this phase is visible along mudbrick Wall J.1:26/J.1:93 and J.1:48, where stones and mudbrick were packed into extensional fractures opened up along Fault C during Earthquake VI. The repair is abutted by subsequent Earthquake V collapse and so must relate to the Phase 3 construction. This wall also highlights evidence of reactivation of all faults and fractures in subsequent earthquakes. Misalignment of the early fourth-century wall allows us to measure both the strike-slip and dip-slip fault motion that has occurred over the past 1700 years. Where walls had subsequently been repaired, the straight wall became curved. Thin wall construction and surface layers produced pottery from the mid to late fourth century A.D. (similar types to Phase 2 described earlier). The latest pottery dates from about A.D. 360 onward (based on several examples of African Red Slip form 67, introduced ca. A.D. 360; Hayes 1972). However, over 100 coins were found on the final floor of this phase. The majority of these coins were found associated with the remains of a broken box in Room 2. The latest coins date to the reign of Constantius II, who reigned from A.D. 337 to 361 (Parker 1999a), and provide a terminus post quem for this building phase. - Thomas et al. (2007)
Earthquake V Early Byzantine 18 May 363 CE ~9 pm
  • The monumental building appears to have been violently shaken in Earthquake V. This is a more severe reactivation of Faults C and D but occurs along a slightly different rupture plane (through the Room 20 north wall) than during EQ VI. The amount of fault slip in this earthquake must exceed 23 cm of dip-slip (measured in sections A and B, fig. 5). Where Fault D shifted Wall J.1:53, a maximum of 30 cm of left-lateral strike-slip was measured. This slip is shared by reactivation in Earthquake IV and the previous Earthquake VI. The collapse layer for Earthquake V exceeds 90 cm in places. The tumble is more evenly distributed throughout the site than was the case for the earlier Earthquake VI, with a bias to the north side of collapsing walls. This thick collapse horizon across the site suggests Earthquake V was stronger in intensity compared with Earthquake VI. The majority of the lateral slip across Fault D is likely to have occurred predominantly in Earthquake V (but also moves in Earthquakes VI and IV). The very refined pottery and coin dates give a secure post A.D. 360 date for the Earthquake V event. The scarcity of post A.D. 360 pottery and the location of the coin hoard at the interface between occupation surface and collapse horizon indicate that this event cannot have occurred long after A.D. 360. We have interpreted this earthquake to be the historically attested earthquake of May 19, A.D. 363 (Russell 1980; Guidoboni 1994: 264-67). - Thomas et al. (2007)
Byzantine Abandonment
  • Following Earthquake V (A.D. 363), there is a notable reduction in occupation and a change in function of this structure. Where collapse layers from Earthquake V were thin, there is some evidence that paving slabs and other architectural components were robbed from the eastern rooms of the monumental building. Subsequently, these hollows were filled with aeolian sand. Later in the Early Byzantine period, portions of the ruined monumental building were adapted and reused for domestic purposes, while other parts of the ruins were used as a dump, collecting much ceramic production waste as suggested by ceramic slag and kiln wasters. This phase of dumping and domestic activity is dated by coins, late fourth- and fifth-century African Red Slip forms, and the Aila amphorae. Domestic use is suggested by the insertion of several clay-lined ovens (tabuns). The monumental structure yielded no evidence of domestic occupation in its earlier history. A sequence of aeolian sands, a thin occupation surface, and further aeolian sands postdate the tabuns and date to the late fourth to fifth century A.D. Late Byzantine and Early Islamic, sixth- to seventh-century aeolian sands deposited across the site denote a period of abandonment prior to the Earthquake IV rupture. - Thomas et al. (2007)
Earthquake IV Umayyad 7th – mid 7th–8th
  • Measured in Section C (fig. 5), Earthquake IV caused 12 cm of dip-slip across Fault D and up to 30 cm of lateral motion on Wall J.1:53. However, since Fault D also slipped in Earthquakes V and VI and appears to have caused more severe structural damage, strike-slip is probably minimal in this event. The pottery constrains the date of Earthquake IV to sometime between the seventh century and the mid seventh to eighth century. In this case, an early to middle seventh-century date would best fit the dating evidence. Earthquake IV probably caused the collapse of the long-abandoned domestic structures. - Thomas et al. (2007)
Umayyad Occupation Umayyad mid 7th – mid 8th
  • A mid-seventh- to mid-eighth-century occupation surface with Umayyad pottery and a related mudbrick wall with stone footings was built on top of the earthquake fault above Room 20. Deep drifts of loose aeolian sands with similar Umayyad pottery were then deposited prior to Earthquake III. - Thomas et al. (2007)
Earthquake III Umayyad/
Abbasid		 mid 7th – mid or possibly late 8th
  • This major event shows rupture along four fault strands (B, C, F, and G), all within the same fault corridor. Faults G and F were clearly visible cutting post-monumental-building tumble in the RAP 2002 excavations of J.29 in Room 13. Fault B caused left-lateral slip on Wall J.1:26 of only 4 cm. However, the dip-slip for all four faults measured in Section 3 was 54 cm, suggesting a major event. The fault rupture was capped by a later occupation dating to the mid to late eighth century. This dates Earthquake III between the mid seventh to mid, or possibly late, eighth century. - Thomas et al. (2007)
8th Century Occupation Abbasid mid–late 8th
  • Construction and reuse of earlier Umayyad structures in the following phase probably dates to the mid to late eighth century, based on related structures in the west of the site that revealed so-called Mahesh Ware pottery. - Thomas et al. (2007)
Earthquake II Abbasid after mid–late 8th
  • These deposits were ruptured and the buildings collapsed. Slip on Fault A produced a left-lateral strike-slip of 5 cm on Wall J.1:26, and Faults A and E caused an accumulated southwest dip-slip of 42 cm (measured in fig. 5C). Wall collapse was minor despite the obvious energy of the earthquake. The pottery within layers capping Earthquake II is earlier than that found in the occupation deposit beneath it. These data suggest that Earthquake II occurred after the mid to late eighth century A.D. - Thomas et al. (2007)
Post 8th Century Abandonment ?
  • A period of aeolian deposition stratigraphically separates Earthquake II from Earthquake I. These sands include Late Byzantine (sixth to early seventh century A.D.) and Umayyad (mid seventh to mid eighth century) pottery. - Thomas et al. (2007)
Earthquake I ? after mid–late 8th
  • The youngest earthquake (Earthquake I) recorded at this site ruptured faults very close to the modern ground surface. - Thomas et al. (2007)

  • Earthquake I ruptured Faults F and H. We measured a total displacement of 35 cm southwest dip-slip in figure 5C, with little or no apparent strike-slip. These faults trend more toward the west (N12°W and N34°W) than the fault rupture in previous earthquakes (ca. 10° more than II to III, and ca. 20° more than the Byzantine Earthquakes V to VI). The fault rupture of Earthquake I was capped by sand and disturbed modern car-park construction deposits, thus preventing finer dating than post–mid to late eighth century. - Thomas et al. (2007)

Phasing Table



Earthquake VII - Nabatean/Early Roman - Early 2nd century CE

Discussion

Discussion

References
Thomas et al. (2007)

Abstract

The ancient ruins of the city of Aqaba are located at the head of the Gulf of Aqaba along the seismically active Dead Sea Transform fault in southern Jordan. Detailed archaeological excavation and geologic mapping were conducted along an active fault that cuts through Late Roman/Byzantine to Early Islamic deposits at the archaeological site of Aila in Aqaba. In this paper, we describe the stratigraphic evidence for ground-rupturing earthquakes that have affected the site and document the associated collapse, damage, and repair of the architecture in antiquity. These data show that there have been seven earthquakes that have disrupted the archaeological deposits since the second century A.D., based on evidence of faulting. Our data clearly show that historical earthquake catalogs are incomplete with regard to some of the less damaging earthquakes that have affected southern Jordan but may have played a significant role in the cultural history of the region.

Introduction

There are many reasons why it is important to study past earthquakes. Natural catastrophes such as earthquakes may be one of many catalysts for social, political, and cultural change that affected a region in antiquity. Furthermore, by understanding the history and repeat pattern of past earthquakes and the area of seismic damage, we can improve our estimates of future earthquake hazards in the region. Data on historical earthquakes are vital for seismic hazard preparedness and for engineering building codes.

The record of historical earthquakes in Jordan extends back over 2000 years. This is one of the longest records of earthquakes anywhere in the world. The accounts and chronology of historical earthquakes have been collected into earthquake catalogs. There are many of these for the eastern Mediterranean and Jordan, including those by Abou Karaki (1987); Ambraseys, Melville, and Adams (1994); Amiran, Arieh, and Turcotte (1994); Amrat (1999); Ben-Menahem (1991); Ghawanmeh (1992); Guidoboni (1994); and Russell (1980; 1985). The earthquake catalogs often list the suggested areas of maximum damage or intensity of an earthquake and use this data to define a magnitude and general epicentral region. It is important to remember that there are very few well-documented historical earthquakes. This is especially true the farther back one looks in time.

Written accounts of earthquakes are incomplete both in their spatial coverage and throughout history. Records from Cairo (Fustat) or Damascus may not fully represent the earthquake history in Jordan. Historical accounts of earthquakes often contain exaggerated reports or omissions. Cities that experienced severe or extensive damage in an earthquake may not be mentioned in the textual data. The historical texts may also contain errors pertaining to the date of an earthquake or the region where it was felt. This is particularly true when the ancient source did not live during the time when the earthquake occurred, but is repeating an account of an earlier writer. Therefore, primary contemporary sources are the most reliable accounts of a given historical earthquake.

An accurate earthquake catalog is very important for assessing both the cultural history of a region and the present-day seismic hazards. As more is learned from the ancient texts, our earthquake catalogs continue to be revised. However, archaeological and geological data can provide us with crucially needed data about the date and intensity of ancient earthquakes. Earthquake data derived from archaeological excavations and geological studies are a completely independent source of information from historical texts that can be used to revise the earthquake catalogs.

The Wadi Arabah Earthquake Project (WAEP) has been conducting archaeological and geological field studies in Jordan since 1996. The Roman Aqaba Project (RAP) excavated a monumental mudbrick structure heavily damaged by successive earthquake faulting in Aqaba (in excavation Area J-East), between 1994 and 2003. A collaborative study of the excavated evidence from this area has identified a sequence of six or seven ground-rupturing earthquakes from the second century A.D. onward. The purpose of this article is to illustrate how careful archaeological excavation across a fault can greatly improve the accuracy of our understanding of the magnitude and frequency of past earthquakes. This is achieved by understanding the rupture history of the southern Dead Sea Transform fault system and documenting the damage from earthquakes in antiquity.

This research requires an approach encompassing historical textual studies; archaeological excavation of damage and fault rupture; and geological, stratigraphic, and geomorphic studies on the evidence of earthquakes. This article focuses on the archaeological evidence for earthquakes, what it tells us, and how it can be used to improve earthquake catalogs.

The Archaeological Site Of Aila

Aqaba lies on the northern shore of the Gulf of Aqaba along the seismically active Dead Sea Transform fault system (figs. 1A, B). The Aqaba fault apparently emerges from the gulf along the east side of the city (Slater and Niemi 2003). Tectonic motion appears to be transferred across the Wadi Arabah valley to the Elat fault on several normal faults that trend parallel to the shoreline (figs. 1B and 1C). These cross faults cut across the alluvial fan sediments of the Wadi Yutim (Niemi and Smith 1999; Mansoor 2002). Slip on the cross faults produces active subsidence at the head of the Gulf of Aqaba.

Archaeological deposits have been excavated at several sites in Aqaba (fig. 1C). The Nabataean and early Roman through Byzantine ruins of Aila (covered in part by Early Islamic–period structures) were excavated in an area approximately 500 m north of the modern shoreline and 500 m northwest of the later Islamic walled city of Ayla (common nomenclature of the Islamic site is Ayla, whereas it is Aila for the Roman site). This area was excavated between 1994 and 2002 by the Roman Aqaba Project, sponsored by North Carolina State University and directed by S. Thomas Parker (Parker 1994; 1996; 1998b; 1999a; 2000; 2002; 2003). The site of Islamic Ayla, founded in the Umayyad period (ca. A.D. 650) and extending possibly through the early 13th century, was excavated between 1986 and 1993 by Donald Whitcomb of the Oriental Institute at the University of Chicago (Whitcomb 1994). Historical sources assert that Ayla was completely destroyed in the earthquake of A.D. 1068 (Poirer and Taher 1980). Approximately 1 km farther to the south is the Mamluk castle (mid-13th century), which was occupied through the Ottoman period and into the 20th century.

Earthquake investigations took place in 2003 in Area J-East of the Aila site as part of the Wadi Arabah Earthquake Project. J-East is located on the east side of Al-Istiqlal Street, 500 m from the shore in Aqaba (figs. 1C, 2). The J-East area is a multiphase site incorporating Early Islamic to Byzantine domestic occupation and a late third- to fourth-century monumental mudbrick structure that has been interpreted as a church (Parker 1998a; 1999a; Mussell 2001; Rose 1998; Weintraub 1999). The monumental structure is also built upon an earlier first- to early second-century A.D. Nabataean complex of substantial stone and mudbrick structures. The objectives of our study were to conduct a detailed architectural survey of earthquake damage to structures at the site and to excavate along the earthquake fault in the area of J-East.

Fault Stratigraphy

Throughout history, motion along the Dead Sea Transform fault system has been the major source of earthquakes in the Levant (fig. 1A). This fault system extends along the Gulf of Aqaba, Wadi Arabah, the Dead Sea, and Jordan Valley. The seismic energy released in an earthquake exerts vertical and horizontal forces on structures that may cause them to collapse or be damaged. The extent of damage to a structure in an earthquake depends on many factors, including the underlying geologic materials, the type of construction, the size of the earthquake, and the distance to the earthquake epicenter. In general, the most severe damage in an earthquake occurs near the epicenter. Also, very large earthquakes cause damage over a greater area than do smaller-magnitude earthquakes. Because earthquakes are generated by sudden slip or motion on a fault, the ground will usually “rupture” or crack right along the fault (in earthquakes generally greater than Richter magnitude 6.5). This type of ground rupture leaves clear geologic evidence on the landscape and in the archaeological record. An archaeological site like Aila is unique as it sits directly across an active fault zone. Information about past earthquakes can therefore be obtained from measurements of offset of the structure across the fault and from careful recording of the stratigraphy, as shown in fig. 3 for the J-East site of Aila. In the following section, we review the evidence for the timing and amount of lateral shifts (strike-slip) and vertical shifts (dip-slip) on several faults (labeled faults A–H) that have occurred during seven earthquakes (labeled EQ I–VII). Several faults have reactivated in later earthquakes.

Nabataean Structures

The earliest occupation of area J-East is represented by three phases of substantial Nabataean stone and mudbrick wall construction, built upon beach sands sterile of artifacts. Ceramic data (mainly early Roman amphorae [Peacock and Williams 1986], Terra Sigillata, and Nabataean painted fineware) associated with these structures suggest that occupation began in the first century A.D. and ended at the close of the Early Roman/Nabataean period, or the early second century based on Nabataean painted fineware (Az Zantur Type 3c; Bignasca et al. 1996: vol. 1, plan 11; Schmid 1997). These occupation deposits were subsequently covered by a very thick layer of mudbrick collapse which contained whole or partial bricks visible in the section. The collapse dents the surfaces beneath, indicating a violent fall of the structures. Excavated in the RAP 2002 season, these layers were found to be in excess of 1 m in thickness. Excavators concluded that this collapse horizon could either relate to the Roman annexation of Nabataea in A.D. 106 or a possible earthquake at the beginning of the second century (Russell 1985). Although there is no specific literary evidence for such an earthquake (Schmid 1997: 419–20), a complete section of collapsed wall (Area J, Trench 11 Loci 227, henceforth J.11:227) might suggest earthquake destruction. There is also continuing debate about the degree of Nabataean resistance to the annexation that might have resulted in destruction by human agency in this period (Bowersock 1983: 78–82; Parker 1986: 123–24; Fiema 1987; Freeman 1996). No rupture for this possible earthquake (EQ VII) was documented in the present study because of the limited areas excavated to this depth (about 2 masl). Furthermore, subsequent building and reuse of the surviving walls has appreciably masked the original geometry.

Monumental Building Phases 1 and 2

In the late third to early fourth century A.D., a monumental rectangular structure was constructed over a thin layer of second-century A.D. aeolian sand and the Nabataean collapse layer. In places, this Late Roman building utilized extant Nabataean walls. The construction and subsequent history of the monumental structure were closely dated by several hundred associated coins and sherds of imported ceramic finewares. During the early fourth century, the monumental building was expanded and concluded with the final addition of Rooms 11 and 12 constructed after ca. A.D. 320. The upper sequences of floors contained Early Byzantine pottery of the mid to late fourth century. Then the monumental mudbrick structure experienced fault rupture and collapse of some walls, producing a tumble horizon. The southern wall of Room 13 was ruptured by Fault D and the northern wall of Room 21 by Fault C. This tectonic shift caused substantial localized damage. Earthquake VI produced a total of 10 cm of left-lateral strike-slip measured across Fault C on wall J.1:26, north of Room 21. This damage from the fault was repaired after Earthquake VI. The strike-slip of Fault D in EQ VI could not be measured because Fault D reactivated in subsequent Earthquakes V and IV. The total strike-slip measured along wall J.1:53 is 30 cm. Since there was no repair to the wall, this suggests that the majority of the slip was caused by EQ VI. Similarly, the dip-slip could not be directly measured, but later releveling of the southwest corner of the monumental building indicates subsidence did occur. Elsewhere on the site, damage appears not to have been quite as severe, but seismically induced wall failures were repaired in the subsequent occupation phase. This seismic event must have occurred at some point in the mid to late fourth century A.D. but before the final extensive collapse of the complex in Earthquake V.

Monumental Building Phase 3

After Earthquake VI, modifications were made to the monumental building, including releveling, blocking doorways, and wall repairs (described in the next section). During this period, mixed fill deposits were used to raise the floor level of sunken areas of the western part of the site presumably caused by slip in Earthquake VI. A major effort was made to repair damaged walls, and a number of paved floors were also constructed to improve the monumental structure. Repair during this phase is visible along mudbrick wall J.1:26/J.1:93 and J.1:48 (fig. 4) where stones and mudbrick were packed into extensional fractures opened up along Fault C during Earthquake VI. The repair is abutted by subsequent Earthquake V collapse and so must relate to the Phase 3 construction. This wall also highlights evidence of reactivation of all faults and fractures in subsequent earthquakes. Misalignment of the early fourth-century wall allows us to measure both the strike-slip and dip-slip fault motion that has occurred over the past 1700 years. Where walls had subsequently been repaired, the straight wall became curved. Thin wall construction and surface layers produced pottery from the mid to late fourth century A.D. (similar types to Phase 2 described earlier). The latest pottery dates from about A.D. 360 onward (several examples of African Red Slip form 67, introduced ca. A.D. 360; Hayes 1972). However, over 100 coins were found on the final floor of this phase. The majority of these coins were found associated with the remains of a broken box in Room 2. The latest coins date to the reign of Constantius II, who reigned from A.D. 337 to 361 (Parker 1999a) and provide a terminus post quem for this building phase. The monumental building appears to have been violently shaken in Earthquake V. This is a more severe reactivation of Faults C and D but occurs along a slightly different rupture plane (through the Room 20 north wall) than during EQ VI. The amount of fault slip in this earthquake must exceed 23 cm of dip-slip (measured in sections A and B, fig. 5). Where Fault D shifted wall J.1:53, a maximum of 30 cm of left-lateral strike-slip was measured. This slip is shared by reactivation in Earthquake IV and the previous Earthquake VI (discussed above). The collapse layer for Earthquake V exceeds 90 cm in places. The tumble is more evenly distributed throughout the site than was the case for the earlier Earthquake VI, with a bias to the north side of collapsing walls. This thick collapse horizon across the site suggests Earthquake V was stronger in intensity compared with Earthquake VI. The majority of the lateral slip across Fault D is likely to have occurred predominantly in Earthquake V (which also moves in Earthquakes VI and IV). The very refined pottery and coin dates give a secure post–A.D. 360 date for the Earthquake V event. The scarcity of post–A.D. 360 pottery and the location of the coin hoard at the interface between occupation surface and collapse horizon indicate that this event cannot have occurred long after A.D. 360. We have interpreted this earthquake to be the historically attested earthquake of May 19, A.D. 363 (Russell 1980; Guidoboni 1994: 264–267).

Byzantine Abandonment

Following Earthquake V (A.D. 363), there is a notable reduction in occupation and a change in function of this structure. Where collapse layers from Earthquake V were thin, there is some evidence that paving slabs and other architectural components were robbed from the eastern rooms of the monumental building. Subsequently, these hollows were filled with aeolian sand. Later in the Early Byzantine period, portions of the ruined monumental building were adapted and reused for domestic purposes, while other parts of the ruins were used as a dump, collecting much ceramic production waste as suggested by ceramic slag and kiln wasters. This phase of dumping and domestic activity is dated by coins, late fourth- and fifth-century African Red Slip forms, and the Aila amphorae. Domestic use is suggested by the insertion of several clay-lined ovens (tabuns). The monumental structure yielded no evidence of domestic occupation in its earlier history. A sequence of aeolian sands, a thin occupation surface, and further aeolian sands postdate the tabuns and date to the late fourth to fifth century A.D. Late Byzantine and Early Islamic, sixth- to seventh-century aeolian sands deposited across the site denote a period of abandonment prior to the Earthquake IV rupture. Measured in Section C (fig. 5), Earthquake IV caused 12 cm of dip-slip across Fault D and up to 30 cm of lateral motion on wall J.1.53. However, since Fault D also slipped in Earthquakes V and VI and appears to have caused more severe structural damage, strike-slip is probably minimal in this event. The pottery constrains the date of Earthquake IV to sometime between the seventh century and mid-seventh to eighth century. In this case, an early to mid-seventh-century date would best fit the dating evidence. Earthquake IV probably caused the collapse of the long-abandoned domestic structures.

Umayyad Occupation

A mid-seventh- to mid-eighth-century occupation surface with Umayyad pottery and related mudbrick wall with stone footings was built on top of the earthquake fault above Room 20. Deep drifts of loose aeolian sands with similar Umayyad pottery were then deposited prior to Earthquake III. This major event shows rupture along four fault strands (B, C, F, and G), all within the same fault corridor. Faults G and F were clearly visible cutting post-monumental building tumble in the RAP 2002 excavations of J.29 in Room 13. Fault B caused left-lateral slip on wall J.1:26 of only 4 cm. However, the dip-slip for all four faults measured in Section 3 was 54 cm, suggesting a major event. The fault rupture was capped by a later occupation dating to the mid to late eighth century. This dates Earthquake III between the mid seventh to mid, or possibly late, eighth century.

Eighth-Century Occupation

Construction and reuse of earlier Umayyad structures in the following phase probably dates to the mid to late eighth century, based on related structures in the west of the site that revealed so-called Mahesh Ware pottery. These deposits were ruptured and the buildings collapsed. Slip on Fault A produced a left-lateral strike-slip of 5 cm on wall J.1:26, and Faults A and E caused an accumulated southwest dip-slip of 42 cm (measured in fig. 5C). Wall collapse was minor despite the obvious energy of the earthquake. The pottery within layers capping Earthquake II is earlier than that found in the occupation deposit beneath it. This data suggests that Earthquake II occurred after the mid to late eighth century A.D.

Post–Eighth-Century Abandonment

The youngest earthquake (Earthquake I) recorded at this site ruptured faults very close to the modern ground surface. A period of aeolian deposition stratigraphically separates Earthquake II from Earthquake I. These sands include Late Byzantine (sixth to early seventh century A.D.) and Umayyad (mid seventh to mid eighth century) pottery. Earthquake I ruptured Faults F and H. We measured a total displacement of 35 cm southwest dip-slip in fig. 5C, with little or no apparent strike-slip. These faults trend more toward the west (N12°W and N34°W) than the fault rupture in previous earthquakes (ca. 10° more than II to III, and ca. 20° more than the Byzantine Earthquakes V to VI). The fault rupture of Earthquake I was capped by sand and disturbed modern car park construction deposits, thus preventing finer dating than post-mid to late eighth century.

In summary, the detailed excavation of the fault within the J-East area of Aila has provided an unprecedented record of seven earthquakes that have ruptured Aqaba since the second century A.D. Based on the type of structural collapse, we suggest that the site was affected by a second-century earthquake (EQ VII). EQ VII is constrained by pottery typology to have occurred in the early second century A.D. EQ VI caused minor structural damage to walls and floors across the site that was repaired in antiquity. These ancient repairs to extant walls are described in the next section. Earthquake VI may have been a foreshock to the much more damaging Earthquake V. Based on sealed coin loci, we have correlated Earthquake V to the earthquake on May 19, A.D. 363 that caused major destruction across sites in Palestine (Russell 1980; Guidoboni 1994). Following Earthquake V, the monumental use of the J-East structure was converted to domestic use in the late fourth to early fifth century. Earthquake IV is constrained to have ruptured through the site after the abandonment of the domestic structures and before major Umayyad occupation. The best date for EQ IV is early to mid seventh century. Earthquake IV appears to be the first of three earthquakes that occurred during the mid-seventh to late eighth centuries (EQ II and III). The most recent earthquake occurred sometime after the eighth century. No stratified materials were found at this site that could be used to further refine the timing of this earthquake (EQ I).

Structural Damage To Exposed Walls

Not all damage to buildings at an archaeological site is caused by earthquakes. Local ground instability can cause structural failures to buildings. This is especially true if the structure's foundation is not intact. If the underlying material is very heterogeneous with different types of fill material, then a portion of a structure may settle or subside. If a structure is located on a hill slope, then it may be damaged or collapse due to downslope movement of the ground. Furthermore, damage and collapse caused by an earthquake happen instantaneously. If the archaeological record shows that a building has deteriorated over a long period of time, then it is probably due to natural decay caused by abandonment and structural disintegration rather than by an earthquake.

Several criteria for the positive identification of earthquakes from archaeological data have been developed by Stiros (1996). The following criteria may represent earthquakes:
  • An ancient structure is offset by seismic surface faulting.
  • Skeletons of people are buried under fallen building debris.
  • There is characteristic structural damage to and failure of many buildings of the same period (see table 1).
  • Destruction and quick reconstruction (repair sometimes with supporting epigraphical data) is found without an overall change in cultural material.
  • There are massive collapse horizons of monumental and smaller structures at a site.
The structural damage in the J-East structure is recognized from several distinct features. Fractures in the walls and corner joins had opened up sometime in antiquity. These had been patched by placing cobbles and pebbles in a matrix of mud into the fissures. The fractures are generally V-shaped in cross section and extend to various depths in the wall. These repairs are easily identified because they have been reactivated in later earthquakes and appear today as open cracks along the margins of the repair.

Intense ground shaking and subsidence of the structure across the active fault has also caused fracturing and localized structural collapse of the building. Diagonal cracks are present on some of the walls, caused by lateral forces of the earthquake accelerations. Northeast-southwest tension due to subsidence along the fault has also created northwest-trending extensional fractures. The southwest side of the structure is built close to or in some cases across active faults. This portion of the building complex has collapsed more than the rest of the area. Some walls are completely missing, and in general there is a lower preservation of wall height. Areas of previous structural repair also sustained more structural damage in subsequent earthquakes.

Several elevation drawings have been constructed of the earthquake-damaged features in the J-East excavation area. These drawings illustrate the repair of earthquake damage, structural collapse, and wall rebuilds that have affected the archaeological constructions in the J-East site over the period of the fourth through the eighth centuries. The earthquake-induced damage repairs found in wall J.1:26 are also a good example of lateral and horizontal offsets caused by active fault motion in an earthquake (fig. 4).

Evidence on the phasing of the structural repair of the walls is clearly seen in the stratigraphic relationship between floors and the partially blocked doorway in Room 6 (fig. 6). The walls in Room 6 have seen several phases of use. They are founded on Early Roman/Nabataean walls. During Phase 2, construction of the monumental structure, wall J.3:77 was added. Phase 2 ended with structural damage and subsequent repair and modifications to the complex. The wall join between walls J.3:77 and J.3:79 was repaired, and the walls were widened with the addition of extra external coursing. The Phase 3 use of Room 4 comes to an abrupt end with the partial collapse of the structure and the deposition of a major mudbrick tumble horizon that correlates site-wide to the A.D. 363 Earthquake V destruction horizon. The doorway between walls J.3:79 and J.3:80 was partially filled at its base with collapsed mudbrick. An occupational surface above the collapse horizon extends from the doorway and into the adjacent room. This surface intersects the structural repair in the corner between walls J.3:77 and J.3:79 about 30 cm above the base of the V-shaped wall join repair. This relationship clearly establishes that the wall repair predates not only the Byzantine domestic occupation but also the A.D. 363 collapse horizon.

The north wall of Room 4 has an arched entranceway (J.20:4, 6) that was blocked in the Early Byzantine domestic phase (J.20:25; fig. 7). The arch was intact when it was excavated but subsequently collapsed after excavation. The base of the blockage sits on aeolian sand deposits that overlie a layer of collapsed mudbrick that correlates to the A.D. 363 earthquake destruction. The west side of the blockage and the arch display diagonal shear fractures. Mudbrick within the blocked archway shows 3 cm of slip oriented along N25°E, 61°NW. A second prominent shear developed between the mudbrick of the arch and the adjoining wall. Diagonal fractures of this type can be characteristic of either subsidence of one pillar of an arch or slippage due to lateral ground acceleration generated by seismic shaking. Damage probably occurred in Earthquake IV or later, after the site was abandoned.

Within Room 11, replastering is seen along the north wall (J.20:11) and east walls. A mud layer between coats of wall plaster appears to be covering diagonal tensional fractures developed in the wall. These fractures have subsequently reopened. Stones and mudbrick mortar fill a separation between the north wall (J.20:11) and a short segment of the west wall north of a doorway (J.20:35) (fig. 8). These repairs probably occurred during the use of the structure between Phases 2 and 3 (EQ VI). However, no independent dating is available for either the replastering or the wall repair. The lower portion of the doorway wall (J.20:35) has partially collapsed outward, and the upper portion is tilted westward. This seismically induced failure type is highly unstable and suggests that the damage may have occurred when the base of the wall was buried by debris and therefore in an earthquake later than the fourth century.

The west wall (J.20:27) of Rooms 11 and 12 experienced severe damage and appears to have collapsed along a line parallel to the inside wall (fig. 9). This suggests that there may have been a preexisting plane of failure within the wall construction. The slip plane is oriented along N22°E. It may be that the exterior wall was repaired by the addition of a mudbrick abutment. A portion of the west wall of the structure lies over the fault in its southwest corner and thus would have experienced southwest-directed extension when the fault slipped.

The south wall of Room 12 (J.1:16) was also severely damaged and is preserved to only 40–120 cm above its stone footing (fig. 9). Large portions of the wall have collapsed, and a large gap exists between the south wall and the east wall. The western section of wall J.1:16 is crossed by faults. Stress fractures induced by subsidence of the southwest corner are evident in a series of southwest-dipping shear planes that cross the mudbrick and a large, cracked undressed rock in the wall footing. Wall J.1:16 abuts to the west and is composed of two large dressed and one decorated ashlar blocks of wall J.1:29 and threshold J.1:76. A fault that crosses between the walls has subsided wall J.1:29 about 25 cm to the west. A large fissure that had opened along the abutment line had been repaired in antiquity with cobbles and mudbrick mortar, probably in Phase 3 (after EQ VI). This fracture was subsequently reactivated by later faulting.

Abutting the east wall of Room 4 is a stairway, arch pillar, and remnants of an arch springer (fig. 10). The stone treads of the stairs (J.2:20) are tilted, with slopes between 5° and 10° northwest. Several subvertical cracks developed between the arch pillar and a small wall that joins the pillar to the west wall of Room 4. Along the north-facing side, a 1.4-m- long extensional fracture was opened along the east side of a structural repair patch. A long extensional joint is also evident on the west wall of Room 4 near the corner join. Mudbrick on the outside of the stone footing of the pillar indicates that structural damage to the pillar occurred and was patched with a new facing. These structural changes probably also occurred in Phase 3 (after EQ VI).

In Room 3, the south wall (J.2:8) is preserved to a height of >2.8 m. A rectangular window (40 × >120 cm) is located 20 cm from the southeast corner (fig. 11). The stones of the wall footing are of different heights on either side of a fracture that extends the height of the wall. The stones in the footing are 45 cm higher on the west side of the fracture. This strongly suggests that the stone portion of the wall was constructed in multiple phases. The mortar lines of the upper mudbrick wall extend across the fracture. The upper 1-m length of the fracture in the wall is V-shaped. The top of the fracture is open to a width of 10 cm and tapers downward to a depth of 40 cm. The base of the fissure is partially filled with cobbles and mud representing a phase of repair. Post-repair shearing has reactivated this fracture and has caused the wall to tilt significantly. The eastern section tilts 15°NE, and the western section tilts 12°NE (fig. 11). An extension of this wall (the south wall of Room 2, J.2:46) has sheared along a plane in the middle of the wall (fig. 11). The wall slid as a unit toward the south and remained upright.

The west wall of Room 3 is only partially preserved to about 1 m in height. This is in stark contrast to the >2.8-m-high south wall. The join between the south wall and the west wall is marked by a repaired fissure that extends to the level of the stone footing (fig. 11). On the east wall of Room 3, a repaired V-shaped fracture was noted about 1 m above the stone footing. This repair seems to accommodate the outward leaning of the south wall (J.2:8).

This study of wall damage and repair from the Byzantine monumental building illustrates the use of these features to determine the extent of damage to archaeological structures by earthquakes. In the monumental building, repairs to walls in Rooms 6, 13, 20, and 21 all occurred in Phase 3; similarly, repairs in Rooms 11 and 12 also probably occurred at the same time. These repairs were contemporaneous with the paving of some rooms (1 and 7) that was only finished just prior to the main Earthquake V that caused the partial destruction and later change in function of the structure.

Discussion And Conclusions

This interdisciplinary approach to recording earthquakes in antiquity has improved the quality of data available for answering questions of interest to both archaeologists and geologists. Tightly controlled dating available from detailed material culture studies, when combined with detailed stratigraphic excavation and careful measuring of fault dip-slip and strike-slip, allows geologists to estimate the magnitude based on empirical data of slip from measured earthquakes (Wells and Coppersmith 1994) and accurately date earthquakes in antiquity. The frequency and magnitude of these earthquakes aid archaeologists’ interpretation of the short- and long-term effect of these seismic events upon social, cultural, and political processes in antiquity. Also, the identification of deposits that can be interpreted as being caused by a single event, at a specific point in time, is of particular use to archaeologists. These deposits, like shipwrecks, act like time capsules. However, unlike shipwrecks, earthquakes can affect a large area, creating stratigraphic relationships across sites. These, in turn, help date structures more accurately and refine material cultural typologies. Then we come to the obvious relation to historically attested earthquakes. Here we must be careful, as a number, or a sequence, of earthquakes may occur in a region over a short period of time. The automatic association of collapse horizons to documented earthquakes produces a circular argument that does not improve the study of ancient seismic events or produce any of the benefits described above. When properly identified, however, these events can further refine the dating, but only if the fault ruptures are independently dated by archaeological means. We must learn to document carefully the earthquake data from archaeological sites so that this information can be used to check and, if necessary, revise the established earthquake catalogs.

Earthquake VI - 1st half of 4th century CE

Discussion

Discussion

Earthquake V - Early Byzantine - 363 CE

Discussion

Discussion

References
Notes by JW

Thomas et al (2007) identified earthquake destruction (Earthquake V) in a collapse layer which they dated to the southern Cyril Quake. A terminus post quem of 360 CE for Earthquake V was established with coins and pottery.

Thin wall construction and surface layers produced pottery from the mid to late fourth century A.D. (similar types to Phase 2 described earlier). The latest pottery dates from about A.D. 360 onward (based on several examples of African Red Slip form 67, introduced ca. A.D. 360; Hayes 1972). However, over 100 coins were found on the final floor of this phase. The majority of these coins were found associated with the remains of a broken box in Room 2. The latest coins date to the reign of Constantius II who reigned from A.D. 337 to 361 (Parker 1999a) and provide a terminus post quem for this building phase.
They added
The very refined pottery and coin dates give a secure post A.D. 360 date for the Earthquake V event. The scarcity of post A.D. 360 pottery and the location of the coin hoard at the interface between occupation surface and collapse horizon indicate that this event cannot have occurred long after A.D. 360. We have interpreted this earthquake to be the historically attested earthquake of May 19, A.D. 363 (Russell 1980; Guidoboni 1994: 264-67).
Powers (2010) adds the following background information:
At the end of the troubled third century, the Legio X Fretensis was transferred from Jerusalem to bolster Diocletian’s new Limes Arabicus, to the effect that the population increased substantially and the city emerged as a regional centre.61 A church was built in c. 300 – one of the oldest in the world – testifying to the early progress of Christianity in Palestine; it was apparently destroyed by the earthquake of 363 and subsequently covered by the new city wall. This stone and mud-brick wall was complete by the late fourth or early fifth century, suggesting something of the seriousness which the continued threat of Saracen raiding was taken.62
Footnotes

61 Parker, 1996: 234, 253; 2000: 392. Eusebius, Onomasticon, 6.17-21 (1904).

62 Parker, 2003: 332.

Earthquake IV - 7th - 8th centuries CE, probably early to mid 7th century CE

Discussion

Discussion

Earthquake III - Umayyad/Abassid - mid 7th - late 8th century CE

Discussion

Discussion

Earthquake II - Abbasid - after mid to late 8th century CE

Discussion

Discussion

Earthquake I - after 8th century CE

Discussion

Discussion

Paleoseismic and Archaeoseismic Effects
Earthquake VII - Nabatean/Early Roman - Early 2nd century CE

Effect Location Image (s) Comments
  • Wall collapse (mudbricks)       
  • dented floor
  • debris over 1 m thick
 J-east
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
  • These occupation deposits [Phase 0] were subsequently covered by a very thick layer of mudbrick collapse which contained whole or partial bricks visible in the section. The collapse dents the surfaces beneath, indicating a violent fall of the structures. Excavated in the RAP 2002 season, these layers were found to be in excess of 1 m in thickness. ... No rupture for this possible earthquake (EQ VII) was documented in the present study because of the limited areas excavated to this depth (about 2 mast). Furthermore, subsequent building and reuse of the surviving walls have appreciably masked the original geometry. - Thomas et al (2007)

  • a complete section of collapsed wall (Area J, Trench 11, Loci 227, henceforth J.11:227) might suggest earthquake destruction - Thomas et al (2007)
  • Wall collapse (inward)
  • Photo in Image(s) shows preservation of some original courses
 some structures in Area B
Fig. 8

Plan of the areas excavated by the Roman Aqaba Project, after the 1998 season. Note that the location of Nabataean/Roman Aila is in the area not surveyed by Meloy (courtesy of RAP).

JW: Area B is in upper middle of map

Dolinka (2003)
Fig. 14

Tumbled-over mudbricks from the domestic complex in B.1/3 bear witness to the earthquake that ushered in the Abandonment II phase at Aila during the early-2nd century AD (courtesy of RAP)

Dolinka (2003)
  • In Area B, Dolinka (2003:32) found that some structures exhibited inwardly collapsed walls and/or tumbled-over mudbricks (Fig. 14) which was attributed to earthquake destruction (footnoting Blakely & Christen 1998: 4, for the collapsed wall B.1:93=B.1.23 illustrated in Fig. 14. See also Harvey 1998: 4, for wall tumble B.2:109)

Earthquake VI - 1st half of 4th century CE

Effect Location Image (s) Comments
  • Some walls collapsed
  • Some walls failed - inferred from repairs
  • tumble horizon
  • Fault D ruptured S wall of Room 13
  • Fault C ruptured N wall of Room 21
  • 10 cm of left-lateral strike-slip measured across Fault C on Wall J.1:26, north of Room 21
  • total strike-slip measured along Wall J.1:53 is 30 cm. (majority of slip caused by EQ VI)
  • subsidence or dip slip of SW corner of monumental building
 various parts of J-east
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
The monumental mudbrick structure experienced fault rupture and collapse of some walls, producing a tumble horizon. The southern wall of Room 13 was ruptured by Fault D and the northern wall of Room 21 by Fault C. This tectonic shift caused substantial localized damage. Earthquake VI produced a total of 10 cm of left-lateral strike-slip measured across Fault C on Wall J.1:26, north of Room 21. This damage from the fault was repaired after Earthquake VI. The strike-slip of Fault D in EQ VI could not be measured because Fault D reactivated in subsequent Earthquakes V and IV. The total strike-slip measured along Wall J.1:53 is 30 cm. Since there was no repair to the wall, this suggests that the majority of the slip was caused by EQ VI. Similarly, the dip-slip could not be directly measured, but later releveling of the southwest corner of the monumental building indicates subsidence did occur. Elsewhere on the site, damage appears not to have been quite as severe, but seismically induced wall failures were repaired in the subsequent occupation phase. - Thomas et al (2007)

Earthquake V - Early Byzantine - 363 CE

Effect Location Image (s) Comments
  • severe reactivation of Faults C and D along slightly different rupture plane (through Room 20 N wall)
  • dip slip must exceed 23 cm.
  • max of 30 cm. left lateral strike-slip where Fault D shifted Wall J.1:53
  • majority of the lateral slip across Fault D is likely to have occurred predominantly in Earthquake V (but also moves in Earthquakes VI and IV)
  • collapse layer exceeds 90 cm in places
  • tumble evenly distributed with bias to N side of collapsing walls
  • collapsed walls
 Faults C and D, N Wall of Room 20, Wall J.1:53
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
Fig. 4

Section drawing of Wall 26 and Wall 48, showing earthquake damage and fault offsets.



Thomas et al (2007)
Fig. 5

Stratigraphic sections of the south and north baulks of J-1, showing the faults.

Thomas et al (2007)
The monumental building appears to have been violently shaken in Earthquake V. This is a more severe reactivation of Faults C and D but occurs along a slightly different rupture plane (through the Room 20 north wall - see Fig. 4) than during EQ VI. The amount of fault slip in this earthquake must exceed 23 cm of dip-slip (measured in sections A and B, fig. 5). Where Fault D shifted Wall J.1:53, a maximum of 30 cm of left-lateral strike-slip was measured. This slip is shared by reactivation in Earthquake IV and the previous Earthquake VI (discussed above). The collapse layer for Earthquake V exceeds 90 cm in places. The tumble is more evenly distributed throughout the site than was the case for the earlier Earthquake VI, with a bias to the north side of collapsing walls. This thick collapse horizon across the site suggests Earthquake V was stronger in intensity compared with Earthquake VI. The majority of the lateral slip across Fault D is likely to have occurred predominantly in Earthquake V (but also moves in Earthquakes VI and IV). - Thomas et al (2007)

Earthquake IV - 7th - 8th centuries CE, probably early to mid 7th century CE

Effect Location Image (s) Comments
  • 12 cm of dip slip across Fault D
    and up to 30 cm of lateral motion (but probably much less) on Wall J.1.53
  • collapse of the long-abandoned domestic structures
 Fault D and Wall J.1.53
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
Fig. 5C

Stratigraphic sections of the south and north baulks of J-1, showing the faults.

Thomas et al (2007)
Measured in Section C (fig. 5), Earthquake IV caused 12 cm of dip-slip across Fault D and up to 30 cm of lateral motion on Wall J.1.53. However, since Fault D also slipped in Earthquakes V and VI and appears to have caused more severe structural damage, strike-slip is probably minimal in this event. ... Earthquake IV probably caused the collapse of the long-abandoned domestic structures. - Thomas et al (2007)

Earthquake III - Umayyad/Abassid - mid 7th - late 8th century CE

Effect Location Image (s) Comments
  • rupture along 4 fault strands (B, C, F, and G), all within the same fault corridor
  • Faults G and F cut post monumental building tumble in J.29 in Room 13
  • Fault B caused left-lateral slip on Wall J.1:26 of only 4 cm.
  • dip-slip for all four faults measured in Section 3 was 54 cm
 Faults B, C, F, and G
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
Fig. 5C

Stratigraphic sections of the south and north baulks of J-1, showing the faults.

Thomas et al (2007)
Fig. 4

Section drawing of Wall 26 and Wall 48, showing earthquake damage and fault offsets.

JW: may refer to Fault C

Thomas et al (2007)
  • This major event shows rupture along four fault strands (B, C, F, and G), all within the same fault corridor. Faults G and F were clearly visible cutting post monumental building tumble in the [Roman Aqaba Project] RAP 2002 excavations of J.29 in Room 13. Fault B caused left-lateral slip on Wall J.1:26 of only 4 cm. However, the dip-slip for all four faults measured in Section 3 was 54 cm, suggesting a major event. - Thomas et al (2007)

Earthquake II - Abbasid - after mid to late 8th century CE

Effect Location Image (s) Comments
  • deposits were ruptured and the buildings collapsed
  • minor wall collapse
  • Slip on Fault A produced left-lateral strike-slip of 5 cm on Wall J.1:26, and Faults A and E caused an accumulated southwest dip-slip of 42 cm
 Fault A and E and Wall J.1:26
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
Fig. 5C

Stratigraphic sections of the south and north baulks of J-1, showing the faults.

Thomas et al (2007)
These deposits were ruptured and the buildings collapsed. Slip on Fault A produced a left-lateral strike-slip of 5 cm on Wall J.1:26, and Faults A and E caused an accumulated southwest dip-slip of 42 cm (measured in fig. 5C). Wall collapse was minor despite the obvious energy of the earthquake. - Thomas et al (2007)

Earthquake I - after 8th century CE

Effect Location Image (s) Comments
  • ruptured faults F and H very close to the modern ground surface
  • 35 cm southwest
    dip-slip with little or no apparent strike-slip
  • ruptured faults rotated 10°-20° westward compared to previous events
 Faults F and H
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
Fig. 5C

Stratigraphic sections of the south and north baulks of J-1, showing the faults.

Thomas et al (2007)
The youngest earthquake (Earthquake I) recorded at this site ruptured faults very close to the modern ground surface. ... Earthquake I ruptured Faults F and H. We measured a total displacement of 35 cm southwest dip-slip in figure 5C, with little or no apparent strike-slip. These faults trend more toward the west (N12°W and N34°W) than the fault rupture in previous earthquakes (ca. 10° more than II to III, and ca. 20° more than the Byzantine Earthquakes V to VI). - Thomas et al (2007)

Paleoseismic and Archaeoseismic Deformation Maps
Earthquake VI - 1st half of 4th century CE

Deformation Map

click on image to open in a new tab

Modified by JW from Fig. 2 of Thomas et al (2007)

Earthquake V - Early Byzantine - 363 CE

Deformation Map

click on image to open in a new tab

Modified by JW from Fig. 2 of Thomas et al (2007)

Earthquake IV - 7th - 8th centuries CE, probably early to mid 7th century CE

Deformation Map

click on image to open in a new tab

Modified by JW from Fig. 2 of Thomas et al (2007)

Earthquake III - Umayyad/Abassid - mid 7th - late 8th century CE

Deformation Map

click on image to open in a new tab

Modified by JW from Fig. 2 of Thomas et al (2007)

Earthquake II - Abbasid - after mid to late 8th century CE

Deformation Map

click on image to open in a new tab

Modified by JW from Fig. 2 of Thomas et al (2007)

Earthquake I - after 8th century CE

Deformation Map

click on image to open in a new tab

Modified by JW from Fig. 2 of Thomas et al (2007)

Paleoseismic and Archaeoseismic Magnitude and Intensity Estimates
Earthquake VII - Nabatean/Early Roman - Early 2nd century CE

Intensity Estimate from Earthquake Archaeological Effects (EAE) Chart

Effect Location Image (s) Comments Intensity
  • Wall collapse (mudbricks)       
  • dented floor
  • debris over 1 m thick
 J-east
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
  • These occupation deposits [Phase 0] were subsequently covered by a very thick layer of mudbrick collapse which contained whole or partial bricks visible in the section. The collapse dents the surfaces beneath, indicating a violent fall of the structures. Excavated in the RAP 2002 season, these layers were found to be in excess of 1 m in thickness. ... No rupture for this possible earthquake (EQ VII) was documented in the present study because of the limited areas excavated to this depth (about 2 mast). Furthermore, subsequent building and reuse of the surviving walls have appreciably masked the original geometry. - Thomas et al (2007)

  • a complete section of collapsed wall (Area J, Trench 11, Loci 227, henceforth J.11:227) might suggest earthquake destruction - Thomas et al (2007)
 VIII +
  • Wall collapse (inward)
  • Photo in Image(s) shows preservation of some original courses
 some structures in Area B
Fig. 8

Plan of the areas excavated by the Roman Aqaba Project, after the 1998 season. Note that the location of Nabataean/Roman Aila is in the area not surveyed by Meloy (courtesy of RAP).

JW: Area B is in upper middle of map

Dolinka (2003)
Fig. 14

Tumbled-over mudbricks from the domestic complex in B.1/3 bear witness to the earthquake that ushered in the Abandonment II phase at Aila during the early-2nd century AD (courtesy of RAP)

Dolinka (2003)
  • In Area B, Dolinka (2003:32) found that some structures exhibited inwardly collapsed walls and/or tumbled-over mudbricks (Fig. 14) which was attributed to earthquake destruction (footnoting Blakely & Christen 1998: 4, for the collapsed wall B.1:93=B.1.23 illustrated in Fig. 14. See also Harvey 1998: 4, for wall tumble B.2:109)
 VIII +
The archeoseismic evidence requires a minimum Intensity of VIII (8) when using the Earthquake Archeological Effects chart of Rodríguez-Pascua et al (2013: 221-224).

Earthquake VI - 1st half of 4th century CE

Magnitude Estimate from fault slip

  • Calculations made using calculators on this webpage
  • Mcalpin (2009:312) suggests a minimum moment magnitude MW of 6.5 when on site rupture is observed
Rupture Type Min (cm) Max (cm) Average (cm) MW min MW max MW avg
left-lateral strike-slip 10 30 20 6.0 6.6 6.3
dip-slip ? ? ? ? ? ?
Thomas et al (2007) noted that there was subsidence in the SW corner of the monumental building but were unable to get a quantitative measure of dip-slip.

Intensity Estimate from Earthquake Archaeological Effects (EAE) Chart

Effect Location Image (s) Comments Intensity
  • Some walls collapsed
  • Some walls failed - inferred from repairs
  • tumble horizon
  • Fault D ruptured S wall of Room 13
  • Fault C ruptured N wall of Room 21
  • 10 cm of left-lateral strike-slip measured across Fault C on Wall J.1:26, north of Room 21
  • total strike-slip measured along Wall J.1:53 is 30 cm. (majority of slip caused by EQ VI)
  • subsidence or dip slip of SW corner of monumental building
 various parts of J-east
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
The monumental mudbrick structure experienced fault rupture and collapse of some walls, producing a tumble horizon. The southern wall of Room 13 was ruptured by Fault D and the northern wall of Room 21 by Fault C. This tectonic shift caused substantial localized damage. Earthquake VI produced a total of 10 cm of left-lateral strike-slip measured across Fault C on Wall J.1:26, north of Room 21. This damage from the fault was repaired after Earthquake VI. The strike-slip of Fault D in EQ VI could not be measured because Fault D reactivated in subsequent Earthquakes V and IV. The total strike-slip measured along Wall J.1:53 is 30 cm. Since there was no repair to the wall, this suggests that the majority of the slip was caused by EQ VI. Similarly, the dip-slip could not be directly measured, but later releveling of the southwest corner of the monumental building indicates subsidence did occur. Elsewhere on the site, damage appears not to have been quite as severe, but seismically induced wall failures were repaired in the subsequent occupation phase. - Thomas et al (2007) VIII +
The archeoseismic evidence requires a minimum Intensity of VIII (8) when using the Earthquake Archeological Effects chart of Rodríguez-Pascua et al (2013: 221-224).

Earthquake V - Early Byzantine - 363 CE

Magnitude Estimate from fault slip

  • Calculations made using calculators on this webpage
  • Mcalpin (2009:312) suggests a minimum moment magnitude MW of 6.5 when on site rupture is observed
Rupture Type Min (cm) Max (cm) Average (cm) MW min MW max MW avg
strike-slip 20 30 25 6.3 6.6 6.45
dip-slip 23 54 38.5 6.2 6.6 6.4

Intensity Estimate from Earthquake Archaeological Effects (EAE) Chart

Effect Location Image (s) Comments Intensity
  • severe reactivation of Faults C and D along slightly different rupture plane (through Room 20 N wall)
  • dip slip must exceed 23 cm.
  • max of 30 cm. left lateral strike-slip where Fault D shifted Wall J.1:53
  • majority of the lateral slip across Fault D is likely to have occurred predominantly in Earthquake V (but also moves in Earthquakes VI and IV)
  • collapse layer exceeds 90 cm in places
  • tumble evenly distributed with bias to N side of collapsing walls
  • collapsed walls
 Faults C and D, N Wall of Room 20, Wall J.1:53
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
Fig. 4

Section drawing of Wall 26 and Wall 48, showing earthquake damage and fault offsets.



Thomas et al (2007)
Fig. 5

Stratigraphic sections of the south and north baulks of J-1, showing the faults.

Thomas et al (2007)
The monumental building appears to have been violently shaken in Earthquake V. This is a more severe reactivation of Faults C and D but occurs along a slightly different rupture plane (through the Room 20 north wall - see Fig. 4) than during EQ VI. The amount of fault slip in this earthquake must exceed 23 cm of dip-slip (measured in sections A and B, fig. 5). Where Fault D shifted Wall J.1:53, a maximum of 30 cm of left-lateral strike-slip was measured. This slip is shared by reactivation in Earthquake IV and the previous Earthquake VI (discussed above). The collapse layer for Earthquake V exceeds 90 cm in places. The tumble is more evenly distributed throughout the site than was the case for the earlier Earthquake VI, with a bias to the north side of collapsing walls. This thick collapse horizon across the site suggests Earthquake V was stronger in intensity compared with Earthquake VI. The majority of the lateral slip across Fault D is likely to have occurred predominantly in Earthquake V (but also moves in Earthquakes VI and IV). - Thomas et al (2007) VIII +
The archeoseismic evidence requires a minimum Intensity of VIII (8) when using the Earthquake Archeological Effects chart of Rodríguez-Pascua et al (2013: 221-224).

Earthquake IV - 7th - 8th centuries CE, probably early to mid 7th century CE

Magnitude Estimate from fault slip

  • Calculations made using calculators on this webpage
  • Mcalpin (2009:312) suggests a minimum moment magnitude MW of 6.5 when on site rupture is observed
Rupture Type Min (cm) Max (cm) Average (cm) MW min MW max MW avg
strike-slip 5 30 17.5 5.8 6.6 6.2
dip-slip 12 12 12 6.0 6.2 6.1
Thomas et al (2007) noted that although there was up to 30 cm of lateral motion on Wall J.1.53 by Fault D, since Fault D also slipped in Earthquakes V and VI and appears to have caused more severe structural damage, strike-slip is probably minimal in this event.

Intensity Estimate from Earthquake Archaeological Effects (EAE) Chart

Effect Location Image (s) Comments Intensity
  • 12 cm of dip slip across Fault D
    and up to 30 cm of lateral motion (but probably much less) on Wall J.1.53
  • collapse of the long-abandoned domestic structures
 Fault D and Wall J.1.53
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
Fig. 5C

Stratigraphic sections of the south and north baulks of J-1, showing the faults.

Thomas et al (2007)
Measured in Section C (fig. 5), Earthquake IV caused 12 cm of dip-slip across Fault D and up to 30 cm of lateral motion on Wall J.1.53. However, since Fault D also slipped in Earthquakes V and VI and appears to have caused more severe structural damage, strike-slip is probably minimal in this event. ... Earthquake IV probably caused the collapse of the long-abandoned domestic structures. - Thomas et al (2007) VIII +
The archeoseismic evidence requires a minimum Intensity of VIII (8) when using the Earthquake Archeological Effects chart of Rodríguez-Pascua et al (2013: 221-224). however since the site was abandoned at the time, the walls may have been weakened. Since Thomas et al (2007) estimated that earthquakes V (S. Cyril Quake) and VI (Aila Quake) were more energetic at the site and an Intensity of VIII (8) was estimated for these earthquakes, it seems prudent to downgrade the intensity estimate one count to VII (7).

Earthquake III - Umayyad/Abassid - mid 7th - late 8th century CE

Magnitude Estimate from fault slip

  • Calculations made using calculators on this webpage
  • Mcalpin (2009:312) suggests a minimum moment magnitude MW of 6.5 when on site rupture is observed
Rupture Type Min (cm) Max (cm) Average (cm) MW min MW max MW avg
left-lateral strike-slip 4 4 4 5.7 5.8 5.75
dip-slip 54 54 54 6.4 6.6 6.5

Intensity Estimate from Earthquake Archaeological Effects (EAE) Chart

Effect Location Image (s) Comments Intensity
  • rupture along 4 fault strands (B, C, F, and G), all within the same fault corridor
  • Faults G and F cut post monumental building tumble in J.29 in Room 13
  • Fault B caused left-lateral slip on Wall J.1:26 of only 4 cm.
  • dip-slip for all four faults measured in Section 3 was 54 cm
 Faults B, C, F, and G
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
Fig. 5C

Stratigraphic sections of the south and north baulks of J-1, showing the faults.

Thomas et al (2007)
Fig. 4

Section drawing of Wall 26 and Wall 48, showing earthquake damage and fault offsets.

JW: may refer to Fault C

Thomas et al (2007)
  • This major event shows rupture along four fault strands (B, C, F, and G), all within the same fault corridor. Faults G and F were clearly visible cutting post monumental building tumble in the [Roman Aqaba Project] RAP 2002 excavations of J.29 in Room 13. Fault B caused left-lateral slip on Wall J.1:26 of only 4 cm. However, the dip-slip for all four faults measured in Section 3 was 54 cm, suggesting a major event. - Thomas et al (2007)
 VIII +
The archeoseismic evidence requires a minimum Intensity of VIII (8) when using the Earthquake Archeological Effects chart of Rodríguez-Pascua et al (2013: 221-224).

Earthquake II - Abbasid - after mid to late 8th century CE

Magnitude Estimate from fault slip

  • Calculations made using calculators on this webpage
  • Mcalpin (2009:312) suggests a minimum moment magnitude MW of 6.5 when on site rupture is observed
Rupture Type Min (cm) Max (cm) Average (cm) MW min MW max MW avg
left-lateral strike-slip 5 5 5 5.8 5.9 5.85
southwest dip-slip 42 42 42 6.3 6.5 6.4

Intensity Estimate from Earthquake Archaeological Effects (EAE) Chart

Effect Location Image (s) Comments Intensity
  • deposits were ruptured and the buildings collapsed
  • minor wall collapse
  • Slip on Fault A produced left-lateral strike-slip of 5 cm on Wall J.1:26, and Faults A and E caused an accumulated southwest dip-slip of 42 cm
 Fault A and E and Wall J.1:26
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
Fig. 5C

Stratigraphic sections of the south and north baulks of J-1, showing the faults.

Thomas et al (2007)
These deposits were ruptured and the buildings collapsed. Slip on Fault A produced a left-lateral strike-slip of 5 cm on Wall J.1:26, and Faults A and E caused an accumulated southwest dip-slip of 42 cm (measured in fig. 5C). Wall collapse was minor despite the obvious energy of the earthquake. - Thomas et al (2007) VIII +
The archeoseismic evidence requires a minimum Intensity of VIII (8) when using the Earthquake Archeological Effects chart of Rodríguez-Pascua et al (2013: 221-224).

Earthquake I - after 8th century CE

Magnitude Estimate from fault slip

  • Calculations made using calculators on this webpage
  • Mcalpin (2009:312) suggests a minimum moment magnitude MW of 6.5 when on site rupture is observed
Rupture Type Min (cm) Max (cm) Average (cm) MW min MW max MW avg
southwest dip-slip 35 35 35 6.3 6.5 6.4

Intensity Estimate from Earthquake Archaeological Effects (EAE) Chart

Effect Location Image (s) Comments Intensity
  • Fault Scarps - ruptured faults F and H very close to the modern ground surface
  • Seismic uplift/subsidence - 35 cm southwest
    dip-slip with little or no apparent strike-slip
  • ruptured faults rotated 10°-20° westward compared to previous events
 Faults F and H
Fig. 2.15

Faults exposed in the Roman Aqaba Project excavation Area J-East

Niemi - Chapter 2 from Parker et al (2014)
Fig. 5C

Stratigraphic sections of the south and north baulks of J-1, showing the faults.

Thomas et al (2007)
The youngest earthquake (Earthquake I) recorded at this site ruptured faults very close to the modern ground surface. ... Earthquake I ruptured Faults F and H. We measured a total displacement of 35 cm southwest dip-slip in figure 5C, with little or no apparent strike-slip. These faults trend more toward the west (N12°W and N34°W) than the fault rupture in previous earthquakes (ca. 10° more than II to III, and ca. 20° more than the Byzantine Earthquakes V to VI). - Thomas et al (2007) VI+ - VII+
The archeoseismic evidence requires a minimum Intensity of VII (7) when using the Earthquake Archeological Effects chart of Rodríguez-Pascua et al (2013: 221-224). However as so many structures at the now long abandoned site had already collapsed, there is limited archaeoseismic evidence and this is likely an under estimate. A minimum Intensity of VIII (8) is more likely.

Calculators
Normal Fault Displacement

Source - Wells and Coppersmith (1994)

Variable Input Units Notes
cm.
cm.
m/s Enter a value of 655 for no site effect
Equation comes from Darvasi and Agnon (2019)
Variable Output - not considering a Site Effect Units Notes
unitless Moment Magnitude for Avg. Displacement
unitless Moment Magnitude for Max. Displacement
Variable Output - Site Effect Removal Units Notes
unitless Reduce Intensity Estimate by this amount
to get a pre-amplification value of Intensity
  

Strike-Slip Fault Displacement

Source - Wells and Coppersmith (1994)

Variable Input Units Notes
cm. Strike-Slip displacement
cm. Strike-Slip displacement
Variable Output - not considering a Site Effect Units Notes
unitless Moment Magnitude for Avg. Displacement
unitless Moment Magnitude for Max. Displacement
  

Site Effect Explanation

The value given for Intensity with site effect removed is how much you should subtract from your Intensity estimate to obtain a pre-amplification value for Intensity. For example if the output is 0.5 and you estimated an Intensity of 8, your pre-amplification Intensity is now 7.5. An Intensity estimate with the site effect removed is helpful in producing an Intensity Map that will do a better job of "triangulating" the epicentral area. If you enter a VS30 greater than 655 m/s you will get a positive number, indicating that the site amplifies seismic energy. If you enter a VS30 less than 655 m/s you will get a negative number, indicating that the site attenuates seismic energy rather than amplifying it. Intensity Reduction (Ireduction) is calculated based on Equation 6 from Darvasi and Agnon (2019).

VS30 Explanation

VS30 is the average seismic shear-wave velocity from the surface to a depth of 30 meters at earthquake frequencies (below ~5 Hz.). Darvasi and Agnon (2019) estimated VS30 for a number of sites in Israel. If you get VS30 from a well log, you will need to correct for intrinsic dispersion. There is a seperate geometric dispersion correction usually applied when processing the waveforms however geometric dispersion corrections are typically applied to a borehole Flexural mode generated from a Dipole source and for Dipole sources propagating in the first 30 meters of soft sediments, modal composition is typically dominated by the Stoneley wave. Shear from Stoneley estimates are approximate at best. This is a subject not well understood and widely ignored by the Geotechnical community and/or Civil Engineers but understood by a few specialists in borehole acoustics. Other considerations will apply if you get VS30 value from a cross well survey or a shallow seismic survey where the primary consideration is converting shear slowness from survey frequency to Earthquake frequency. There are also ways to estimate shear slowness from SPT & CPT tests.

Notes and Further Reading
References

Articles and Books

Dolinka, B. J. (2003). Nabataean Aila (Aqaba, Jordan) from a Ceramic Perspective: Local and intra-regional trade in Aqaba Ware during the first and second centuries AD. Evidence from the Roman Aqaba Project.

Niemi, T. M. (2014). Chapter 2. The Regional Environment, in S. T. Parker et al. III (eds.), The Roman Aqaba Project Final Report volume 1: The Regional Environment and the Regional Survey, Archaeological Reports 19. Boston, MA, American School of Oriental Research, 33-80.

Parker, S. T. (1986). Romans and Saracens: A History of the Arabian Frontier. Winona Lake.

Parker, S. T. (1987a). Peasants, Pastoralists and Pax Romana: A Different View. Bulletin of the American Schools of Oriental Research 265: 35-51.

Parker, S. T. (1996). The Roman ʿAqaba Project: the 1994 Campaign. Annual of the Department of Antiquities of Jordan 40: 231-257.

Parker, S. T. (1997). Preliminary Report on the 1994 Season of the Roman Aqaba Project. Bulletin of the American Schools of Oriental Research 305: 19-44.

Parker, S. T. (1998). The Roman ʿAqaba Project: the 1996 Campaign. Annual of the Department of Antiquities of Jordan 42: 375-394.

Parker, S. T. (1999). Brief Notice on a possible Fourth-Century Church at Aqaba, Jordan. Journal of Roman Archaeology 12: 372-376.

Parker, S. T. (2000). The Roman ʿAqaba Project: the 1997 and 1998 Campaigns. Annual of the Department of Antiquities of Jordan 44: 373-394.

Parker, S. T. (2002). The Roman ʿAqaba Project: the 2000 Campaign. Annual of the Department of Antiquities of Jordan 46: 409-428.

Parker, S. T. (2003). The Roman ʿAqaba Project: the 2002 Campaign. Annual of the Department of Antiquities of Jordan 47: 321-333.

Parker, S. T. (2006). Roman Aila and the Wadi Arabah: An Economic Relationship. In P. Bienkowski & K. Galor (eds.), Crossing the Rift: Resources, Routes, Settlement Patterns and Interaction in the Wadi Arabah. Oxford, 223-230.

Parker, S. T. (2009). The Roman Port of Aila: Economic Connections with the Red Sea Littoral. In L. Blue, J. Cooper, R. Thomas & J. Whitewright (eds.), Connected Hinterlands: Proceedings of the Red Sea Project IV, held at the University of Southampton, September 2008. Oxford, 79-84.

Power, T. (2010). The Red Sea during the "Long" Late Antiquity, AD 500-1000. Faculty of Oriental Studies. Oxford, University of Oxford. PhD dissertation.

Thomas, R., et al. (2007). Structural damage from earthquakes in the second-ninth centuries at the archaeological site of Aila in Aqaba, Jordan: PERA. Bulletin of the American Schools of Oriental Research 346: 59-77.

Excavation Reports

Parker, S. T. (1987b). History of the Roman Frontier East of the Dead Sea. In T. S. Parker (ed.), The Roman Frontier in Central Jordan: Interim Report on the Limes Arabicus Project, 1980-1985. 2 vols. Oxford, 793-823.

Websites

Aila Port (Ayla) (Aqaba) at Nabataea.net