Gulf of Aqaba Turbidites

• from Bektaş et al. (2024:4-9)

Radiographic images of the cores reveal that the background sediments in the Gulf of Aqaba do not exhibit lamination due to intense bioturbation (Figs. 2 and 3). Since the shell density of planktonic foraminifera (1.4 – 1.5 g/cm3; Fok-Pun and Komar, 1983) is generally lower than that of the sediments, they are seen in the radiographic images as light-colored spots. Within the intensely bioturbated background sediments having high biogenic content (predominantly planktonic foraminifera), we identified numerous sedimentary anomalies that are significantly different from these complex background sediments in our cores. These sedimentary anomalies, which produce signals that differ significantly from the background sediments in proxies and radiographic images, are labeled alphabetically from the top to the bottom for each core (Figs. 2, 3, and 4). They appear in the radiographic images as darker intercalations, implying that they have higher density compared to the background sediments.

The examples of the sedimentary events presented in Fig. 2 are the most prominent and well-preserved turbidites in the studied cores. Due to their instantaneous deposition and thickness mostly over 4 cm, bioturbation is limited to only the topmost parts of these turbidites, resulting in well-preserved internal structures. Parallel-to-subparallel laminations (Fig. 2: 2-E, 15-B, 3-A, 4-A, 15-C, 16-B, 17-G, and 15-A), and even cross laminations in some cases (3-E, 4-D, 14-B, and 16-A; Fig. 2), just above the sharp bottom boundaries of these events, can be attributed to multiple coarse sediment pulses that are due to multiple successive mass wasting events along basin slopes, likely caused by an earthquake (Shiki et al., 2000; Nakajima and Kanai, 2000; Goldfinger et al., 2007, 2008; Goldfinger, 2011; Van Daele et al., 2014; 2017; Wils et al., 2021). These distinct laminations are mostly overlapped by more homogeneous and probably finer-grained sediments that are clearly lacking carbonaceous biogenic content, i.e. mostly planktonic foraminifera seen as whitish spots (e.g., 13-I in Fig. 2). Although boundaries at the bottom of turbidites are sharp and distinct, the boundary between the top of the turbidite and the above background sediments is difficult to determine precisely since it is gradational and bioturbated (Goldfinger et al., 2008; Goldfinger, 2011). In addition to their well-preserved internal structures, the lack of biogenic content in these sediments confirms their quasi-instantaneous deposition, which must have been too fast for the biogenic carbonates precipitating from the water column to be included into the sediments. In this study, we classify all of the sedimentary events that appear like the ones presented in Fig. 2 as “Type I: Turbidites”.

Some hazy, but still relatively darker levels, can be seen in the radiographic images (e.g., 9-B, 17-B, 2-C, and 18-B in Fig. 3), although they neither include multiple laminations nor foraminifera-free homogenous parts. We also classified these levels as sedimentary events, since their darker radiographic view implies a sudden influx of coarser hemipelagic sediment arrival at the coring location. We interpret them as either thin and singular turbidites or flood deposits that were dispersed within the background sediments by intense bioturbation. Hence, we classify them as “Type II: Turbidite or Flooding” events. Type III events are characterized by their thickness, darker appearance in radiographic images (e.g., 17-C, 17-F, and 18-D in Fig. 3), and coarser grain-size compared to the background sediments, which is evident from their Sand (%) values (9-F, 11-F, 11-I, 11-M, 17-C, 17-F, and 18-D in Fig. 5). Unlike Type I events, Type III events are bioturbated, lacking any lamination, and containing biogenic remains. We consider three possible explanations regarding their origin. Firstly, it is possible that Type III events are turbidites that underwent complete bioturbation after deposition. However, considering that coarser sediments typically exhibit less vertical penetration of bioturbation (Wheatcroft, 1992), and given that Type III events are noticeably coarser than the background sediments, it is unlikely that these events under went extensive bioturbation after deposition. Another possibility is that Type III events originated from hemipelagic sediments highly rich in biogenic content, hence their turbidites contain abundant biogenic remains as well (e.g., Van Daele et al., 2017; Polonia et al., 2023). However, if this was the case, we would expect to observe some evidence of multiple laminations or fining-upward grain-size trends in Type III events, which is not observed. The most plausible explanation for these events is that, during their deposition, there was sufficient time for organisms to dig and burrow, and for biogenic remains from the water column to be included into the sediments. Consequently, the deposition of Type III events was likely slower than that of Type I and Type II events. These events likely represent a series of successive flooding events that occurred over a period of several years or decades. We classify these events as “Type III: Thick flooding sequence”. For a more detailed classification of sedimentary events, please refer to the Supplementary Material (E-SUPP 1).

Fig. 4 shows radiographic images of all the 18 cores side-by-side along the gulf, together with the depths of ¹⁴C and ²¹⁰Pb measurements and their results (Table in Fig. 4). The stratigraphic order of the dated samples confirms that our samples were collected in regular sedimentation sections and not in any anomalous sedimentary event. Radiocarbon results of 16 samples show that most of the cores include the sedimentary record for at least the last 1000 years.

We applied the “Constant Flux Constant Sedimentation Rate” model (Goldberg, 1963) on the ²¹⁰Pb_ex (excess lead) from the cores 3, 7, and 17, which yielded sedimentation rates of 0.25, 0.19, and 0.38 mm/yr for the top parts of these cores, respectively. It should be noted that event 17-A was excluded from the depth scale for the sediment rate calculation of core 17. Although the number of radionuclide samples collected from core 7 is insufficient to achieve a statistically meaningful sedimentation rate, the ²¹⁰Pb_ex results from cores 3 and 17 can be compared with the radiocarbon ages. For core 3, ¹⁴C at the bottom of the core yields a bulk sediment rate of 0.44 mm/yr, which is inconsistent with the rate obtained by ²¹⁰Pb_ex for the same core (0.25 mm/yr). However, ¹⁴C-based sediment-rate calculation by using composite depths, which are obtained by excluding the sedimentary events, yields a composite sediment rate of 0.26 mm/yr that is consistent with the ²¹⁰Pb_ex sediment rate. Similarly, ¹⁴C-based bulk rates for cores 14 and 15 (0.56 and 0.50 mm/yr, respectively) in Eilat Deep are much higher than the ²¹⁰Pb_ex-based sediment rate obtained for core 17 (0.38 mm/yr) from the same basin. However, once corrected by removing the event layers, the ¹⁴C-based composite sediment rates for cores 14 and 15 (0.34 and 0.33 mm/yr, respectively) are rather consistent with the 0.38 mm/yr derived from the ²¹⁰Pb_ex measurements. Thus, these two comparisons confirm the importance of determining both the bottom and top boundaries of sedimentary events to properly exclude them from the sequences and hence to construct reliable sediment chronology.

In order to cross-check the existence and extent of the sedimentary events in the cores, we compared our visual inspections of the radiographic images to magnetic susceptibility, grain-size (only sand content) and µ-XRF measurements (Fig. 5). For these analyses, at least one core per basin was selected. During floods or mass wasting events, coarser sediments originating from the shallower parts of the basin or from drainages onshore are expected to reach bottom of the basins. Thus, sand content in the event deposits is expected to increase compared to the background sedimentation. Deep marine sediments are normally a mixture of terrigenous clastics (mainly aluminosilicate minerals) and bio/chemical carbonates produced in the water column. However, there is almost no bio/chemical carbonate input during the almost instantaneous deposition of turbidite or flood deposits, which makes them richer in terrigenous clastics compared to the background sediments. Since aluminosilicates have higher magnetic susceptibility values than carbonates (Nowaczyk, 2001), turbidite and flood deposits should show as anomalies along the magnetic susceptibility profiles of the cores. Similarly, Zr/Sr profiles, where Zr and Sr represent aluminosilicates and foraminiferal calcite, respectively (Rothwell et al., 2006; Croudace and Rothwell, 2015), should show anomalies at turbidite and flood levels.

Among the 67 events shown in Fig. 5, significant magnetic susceptibility anomalies are observed for 51 of them. While 11 events (1-C, 3- D, 7-A, 9-C, 10-B, 10-C, 13-I, 15-C, 17-E, 17-G, and 18-B) show almost no magnetic susceptibility anomalies, only three events (3-A, 15-A, and 17-A) have clearly lower magnetic susceptibility values than the background sediments. Sand fraction profiles show distinct anomalies for 57 events out of 67. On the other hand, events 7-B, 15-A, 15-B, and 17-A have lower sand content compared to the background sediments. No sand anomalies are observed for four events (1-A, 1-B, 13-I, and 17-G). Since no µ-XRF scanning was done for core 18, 62 events can be tested for Zr/Sr anomalies. Among these, 49 events show higher and only two events (3-A and 17-A) lower Zr/Sr anomalies. Nine events (3-C, 3-D, 7-A, 9-C, 10-B, 11-A, 11-B, 13-D, and 13-I) show no Zr/Sr anomalies. Accordingly, from the data presented in Fig. 5, we note that magnetic susceptibility, sand content, and Zr/Sr profiles are successful in detecting sedimentary events observed on the radiographic images at rates of 81%, 91%, and 79%, respectively. Although Type I and Type III events are already evident in the radiographic images, some Type II events are unclear in the images (e.g., 7-A in Fig. 3). Magnetic susceptibility, sand content, and Zr/Sr profiles, together with the radiographic images, were therefore particularly useful to confirm/detect the thickness of Type II events, so that all sedimentary events were successfully excluded from the sequences covered by the cores to achieve reliable stratigraphical correlations and sediment chronology.

• from Bektaş et al. (2024:9-10)

Coevality of turbidites at different locations and even in different basins should be tested to achieve successful submarine paleoseismological records (e.g., Goldfinger, 2011), which can be achieved by careful high-resolution stratigraphical correlations. In Fig. 4, some of the sedimentary events can be visually correlated between cores collected from the same basin according to their stratigraphical order, e.g., between the cores 2, 3, and 4 in Tiran basin, cores 9 and 10 in Dakar basin, and cores 14, 15, 16, 17 and 18 in Eilat basin. However, visual observations are not reliable enough for inter-basin correlations between the cores as one cannot assume that the number of turbidites in different basins is the same. Stratigraphical correlations between cores can be achieved by using data reflecting geophysical and geochemical properties of sediments, which may include magnetic susceptibility, bulk density, grain-size distribution, computed tomography (CT) image analysis, μ-XRF data and paleomagnetic secular variation (PSV) records (Patton et al., 2013; Drab et al., 2015; Ikehara et al., 2016; Goldfinger et al., 2017; Usami et al., 2018). In our study, geophysical and geochemical properties of sediments were evaluated by magnetic susceptibility and grain-size measurements (Sand percent), and μ-XRF μ scanning, of which resolutions were 5 mm, 10 mm and 0.5 mm, respectively. μ-XRF data, which has significantly higher resolution than the other proxies, was preferred for core correlation. Reliable chemostratigraphical correlations over large distances, like in our case where the cores are distributed along the ~180-km-long gulf, can only be successfully achieved by using a sedimentary geochemical proxy recording regional environmental conditions effective for the entire gulf. It is known that Sr/Ca ratios of planktonic foraminifera have a strong positive correlation with sea-surface temperature (e.g., Clerouxet al., 2008). Since the sediments of the Gulf of Aqaba are rich in planktonic foraminifera, and by assuming that the surface seawater temperatures would synchronously change over the entire gulf, we used Sr/Ca ratio profiles to correlate the cores in this study.

We intentionally collected core 11, the longest core in this study, from a ridge just to the south of the Aragonese Deep, rather than from the depocenter of the basin (Fig. 1d), so that it would dominantly reflect the background sedimentation recording the climatic conditions rather than being dominated by turbidites. To achieve reliable stratigraphical correlations, sedimentary events should be excluded from the sequences, and correlations should be done on composite profiles representing background sedimentation (Arnaud et al., 2002; Schwab et al., 2009; Avs¸ar et al., 2015; Moernaut et al., 2017). After excluding the sedimentary events, we correlated the composite Sr/Ca profiles of all the cores to the composite Sr/Ca profile of core 11 (Fig. 6). As the sedimentation rates are different between basins, this resulted in squeezing or stretching of the core records. This calibration was done using 5 to 8 characteristic reference levels that can be recognized in all the cores, which we used as tie-points to ensure consistency between the cores through the calibration process. (Gray lines in Fig. 6). Detailed explanation of the procedure for the removal of the sedimentary events and chemo-stratigraphical correlation can be found in Supplementary Material (E-SUPP 2). The original and modified depths of these tie-points are also presented as bi-plots next to each correlation plot in Fig. 6. Almost linear and smooth appearances of these bi-plots confirm that depth modifications did not result in abnormal sedimentation rates. For five cores (1, 6, 8, 12, and 13), the chemostratigraphical correlations result in almost perfect overlaps with core 11. For most of the other cores, although exact overlaps are not achieved, similarities between general trends and fluctuations along the cores are still within the range of uncertainties and thus are deemed acceptable. The discrepancies are probably due to the semi-quantitative nature of ITRAX µ-XRF scanning data or local differences in sedimentation. In addition, cores 11 and 13 were collected by a piston corer, which sometimes causes a disturbance and sediment loss close to the water/sediment interface. However, a good correlation between cores 11 and 12, which was taken by a multicorer ensuring undisturbed recovery of water/sediment interface, confirms that there was no sediment loss at the top of core 11. On the other hand, correlation between cores 11 and 13 shows that approximately 15 cm-thick sediment was lost at the top of core 13 during coring operation. A close-up view of the correlation between cores 13 and 11 is also provided in Fig. 6. Correlation between cores 17 and 18 is based on the stratigraphical order of the events since no µ-XRF data is available for core 18.

The raw ¹⁴C dates listed in Fig. 4 were included in the OxCal P_Sequence code (E-SUPP 3) with respect to their modified depths obtained by the chemostratigraphical correlations, as if all the radiocarbon samples came from core 11. The k value for the P_Sequence function was selected as 3, which is generally used for deep sea environments where hemipelagic sedimentation can be assumed rather constant (e.g., Polonia et al., 2023). Although the dates show a reasonable trend along core 11 (Fig. 7), two dates, the one from the bottom of core 7 (7–38.70) and the youngest date from core 9 (9–33.04), are clearly older than the general trend. Hence, we interpret them as reworked material and did not use them in the P_Sequence code. The resulting age-depth model, which is presented with 68% and 95% confidence intervals in Fig. 7, yields approximate sedimentation rates of 0.35, 0.16 and 0.22 mm/yr for the intervals of 0–32 cm, 32–55 cm and 55–80.5 cm, respectively. All of the sedimentary events projected on core 11 are also plotted with respect to depth in Fig.7, and are included in the P_Sequence code (E-SUPP 3), so that probability density functions (PDFs) for each event detected in Gulf of Aqaba can be calculated.

• from Bektaş et al. (2024:10-13)

PDFs of Type I and Type II events in the cores through the gulf are plotted with respect to calendar dates in Fig. 8, which were obtained according to the age-depth relation presented in Fig. 7. The probability density values for Type II events were multiplied by 0.5 since they may also be floods. Type III events, which are most probably not related to earthquakes, were not included in this plot. In Fig. 8, the plot of summed PDFs is also presented in order to statistically express the spatial extent and coevality of turbidites in the gulf; i.e., multiple coeval turbidites should be seen as distinct anomalies on the summed PDFs profile. Several sedimentary events appear to be coherent for several cores, including cores located in different basins. Hence, they are seen as anomalies in the summed PDFs profile (e.g., around late 10th, 16th and 20th centuries CE), and we interpret these events as the signature of past earthquakes that triggered turbidites (Type I or II) in the Gulf of Aqaba.

The turbidites dated to the late 20th century are unambiguously seen as Type I events at the tops of cores in Aragonese and Eilat Deeps (Cores 11, 12, 14, 15, 16, 17, and 18) and as Type II event in Dakar Deep (Core 10). They constitute a perfect benchmark of our sedimentary system as they are almost certainly the sedimentary traces of the 1995 Nuweiba (M_W 7.2) earthquake, the most recent major earthquake in the Gulf of Aqaba. Given that the Aragonese Fault, and probably partially the Eilat Fault as well, were the source faults for the 1995 Nuweiba earthquake and that the earthquake rupture propagated northward (Baer et al., 2008; Hofstetter, 2003; Klinger et al., 1999; Shamir et al., 2003; Ribot et al., 2021), it is not surprising that turbidites (Type I events) from this earthquake are not found in cores from the southern part of the gulf (i.e., cores 1 to 10). Despite the proximity to the epicenter of the 1995 earthquake, no turbidite associated with this event is visible in core 13 due to sediment loss at the top of that core (Fig. 6).

Among older turbidites, two events stand out and are recognized almost in every core, one in about the early 12th century CE and one in about the late 16th century CE. Starting from the oldest turbidites, one series of turbidites are clearly visible as Type I events in all cores (except core 11), which are long enough to cover at least the last millennium (Fig. 8). Since core 11 is not a turbidite-targeting core, and it was collected from a ridge rather than from a depocenter, it can be expected to see only limited evidence for turbidites in this core. The significant anomaly on the summed PDFs profile, near the beginning of the 12th century CE is most probably associated with the only major earthquake known in the region during that period: the CE 1068 earthquake. The temporal discrepancies observed for dates of turbidites in some cores are most probably related to the difficulties of the inter-core chemostratigraphical correlations due to intense bioturbation in the sediments of the Gulf of Aqaba that can lead to small time shifts when it comes to estimate exact age of specific core sections. The 18 March 1068 Aqaba-Hijaz earthquake was reported to have devastating effects in many locations from the city of Aila (Eilat) to Medina and Cairo (Ambraseys, 2009). Presence of seismo-turbidites all along the gulf in addition to evidence found at the onshore paleoseismic sites north of the Gulf (Amit et al., 1999, 2002; Zilberman et al., 2005; Klinger et al., 2015; Kanari et al., 2020) indicate that the CE 1068 earthquake was a major earthquake in the region that ruptured the Eilat, Aragonese, Arnona and probably Tiran faults together, in addition to the southernmost part of the Wadi Arabah Fault near the gulf. Hence, the total rupture length of this earthquake could have been at least ~200 km.

Similar as for the CE 1068 event, numerous coeval turbidites along the Gulf of Aqaba are represented by the anomaly on the summed PDFs profile about late 16th century CE, which is consistent with the well-documented earthquake of 4 January 1588 (Ambraseys, 2009). Majority of them are of Type I events, except the ones in core 2 in Tiran Deep, and in cores 9 and 10 in Dakar Deep, which are of Type II. The absence of turbidites in core 6 can be explained by the fact that this core was collected from a small and isolated basin (Fig. 1e). Indeed, this basin is isolated from turbidity flows that would come from the main slopes of the gulf, and its relatively smaller slopes may not be sensitive to earthquake shaking as much as the larger main slopes of the gulf. An absence of a turbidite in core 14 is more difficult to explain. One possible explanation could be the sediment clearance on the slopes and the banks of the submarine channels during the preceding earthquake, i.e., in CE 1068, leaving nothing to be wasted during the CE 1588 earthquake. Furthermore, another possibility is that the turbidity flows due to 1588 earthquake might have bypassed the location of core 14 (Goldfinger et al., 2017). Apart from turbidite absence in these three cores, it seems that the CE 1588 earthquake triggered seismo-turbidites along the entire gulf. This earthquake is also known as a devastating event in the historical records, affecting many places from the cities of Eilat and Aqaba at the northern end of the gulf to Cairo in Egypt (Ambraseys, 2009). A second event is also reported in the historical chronicles that happened on 7 April 1588, which was felt in Cairo and in the northern Red Sea. Although it cannot be ruled out that this second event was completely independent of the event in January 1588, it could be an aftershock of the former event (Ambraseys, 2009), triggering additional turbidites for example along the Tiran fault section. Klinger et al. (2015) report no surface rupturing evidence for the CE 1588 earthquake at the Qatar trench site (x in Fig. 1b). On the other hand, Kanari et al. (2020) report both rupturing and paleoliquefaction evidences that could be related to this earthquake in the Eilat Sabkha (t in Fig. 1b). Hence, based on our seismo-turbidite observations, it appears that the CE 1588 earthquake likely also ruptured the entire fault system in the Gulf of Aqaba. Unlike in CE 1068, however, the 1588 rupture does not seem to have propagated inland beyond the northern end of the gulf. In the south, based on turbidites we can trace the rupture to the south of the Tiran Strait and it may have ruptured even further to the south.

Some additional sequences of turbidites are found in our record, although they are not as extensive as the sequences associated respectively to the CE 1068 and CE 1588 earthquakes. According to the historical records, the CE 1212 earthquake caused widespread damage in an extensive area from Al-Shaubak and Al-Karak in the north (ca. 150 km north of Eilat and Aqaba cities) to the St. Catherine Monastery in the south (ca. 50 km east of Dahab) (Ambraseys, 2009). Although Klinger et al. (2015) reported evidence of surface rupturing along the southern Wadi Arabah Fault (Qatar Site, x in Fig.1b) that could be related to the CE 1212 earthquake, the trenching studies closer to the gulf (Zilberman et al., 2005; Kanari et al., 2020; y and t in Fig. 1b, respectively) claim that the CE 1212 earthquake was likely generated by a secondary fault on the eastern edge of the Eilat depression, rather than by the Wadi Arabah Fault. The probable turbidites of the CE 1212 earthquake are only seen in cores 17 and 18 from the northernmost part of the gulf (Fig. 8). Thus, it appears as a minor anomaly in the summed PDFs profile. The turbidite in core 18 is thicker and better preserved with its laminated internal structure. Furthermore, there are no turbidites in cores 14 and 15 around 1200s. Given the absence of consistently coeval turbidites through the gulf that are dated to the beginning of the 13th century (except the one in core 13), and only two turbidites in the northernmost cores (17 and 18), we conclude that the CE 1212 earthquake was likely significantly smaller than the CE 1068 and CE 1588 events. The exact location of this event north of the gulf remains uncertain, but it might have caused a discontinuous rupture on fault segments of the Wadi Arabah Fault system.

Prominent turbidites with well-preserved internal structures are seen around the mid-19th century CE in cores 2, 3 and 4. No turbidites were found around this period in the neighboring cores, except Type II events in cores 1, 7, and 8. The summed PDFs profile has a minor anomaly due to these turbidites that are limited to the most southern part of the gulf. These turbidites are probably related to the CE 1839 earthquake, which is described in the historical records as causing minor damage on the walls of St. Catherine Monastery (Ambraseys, 2009). Furthermore, Purkis et al. (2022) reported an incipient submarine landslide on the southeastern slopes of Tiran Deep that failed within the last 500 years, plausibly triggered by the CE 1839 earthquake. Given that the turbidites in cores 2, 3, and 4 are limited only in Tiran Deep, it is likely that the CE 1839 earthquake only partially ruptured either the Tiran or Arnona fault, or activated one of the secondary faults in the southernmost part of the gulf. Except for the ones related to the CE 1839 earthquake, the absence of distinct coeval turbidites in cores 1 to 10 (southern half of the gulf) since CE 1588 implies that most of the Tiran and Arnona faults has not ruptured since then.

While the cores provide a detailed record for the past millennium, only a few cores provide information for earlier periods. Around the middle 5th century CE, there are implications of coeval turbidites in cores 9, 10, and 11 in Dakar and Aragonese basins. Klinger et al. (2015) report evidence of surface rupturing at the Qatar trenching site on the Wadi Arabah Fault between 9 BCE and CE 492 (Fig. 8), which they attribute to the CE 363 earthquake that affected significantly the southern part of the Dead Sea Fault (Thomas et al., 2007; Ambraseys, 2009). However, this event could not be found north of the Qatar site (Lefevre et al., 2018). Hence, the turbidites around the middle 5th century CE could be due to a large earthquake rupturing several faults in the gulf, in addition to a 50 km long on-shore fault section. In core 11, three more Type I events, which can be other earthquakes in the gulf, are dated to ca. 250, 850 and 1350 BCE. The one around 250 BCE temporally correlates well with the event between 338 and 213 BCE, detected by Klinger et al. (2015) (Fig. 8). Although Klinger et al. (2015) do not report any event around 850 BCE, they report two events occurred during the period 2797–1245 BCE, which also coincide with the date of the turbidite ca. in 1350 BCE.

From the oldest turbidite to the youngest in 1995, the recurrence intervals for major earthquakes in the gulf seem to vary between 400 and 700 years (Fig. 8), with a mean value of 560 years. According to the paleoseismic trenching studies conducted in the north of the Gulf of Aqaba region, the Dead Sea Fault is characterized by earthquake clusters lasting 100–200 years, followed by seismic quiescence periods of 350–400 years (Klinger et al., 2015; Lefevre et al., 2018). On the other hand, Lefevre et al. (2018) have proposed that recurrence intervals for events should be longer due to partial stress partitioning between offshore and on-land fault segments, which is consistent with the 560 years recurrence interval found in the Gulf of Aqaba. In addition, the left-lateral relative plate motion between the Sinai and Arabian plates at this latitude is almost 5 mm/year (Castro-Perdomo et al., 2022; Viltres et al., 2022), which yields about 2 to 3.5 m fault slip deficit in a 400–700 year period, corresponding to an expected surface rupture length of 100–180 km (Wells and Coppersmith, 1994), which would be consistent with multi-segment rupturing events in the gulf. However, recent studies based on InSAR and GNSS observations (Li et al., 2021; Castro-Perdomo et al., 2022) report fault-locking depths that become shallower towards the south and possibility of partial fault creep along the southernmost fault strands of the gulf. Still, our results indicate that the entire Gulf of Aqaba fault system was activated in the 1068 and 1588 earthquakes and probably during the previous major earthquakes. This implies that the southern gulf can be regarded as being close to the end of the earthquake cycle as it has not ruptured in a major earthquake for more than 400 years.

• from Bektaş et al. (2024:13)

Based on ITRAX µ-XRF scanning, radiographic imaging, magnetic susceptibility measurements and grain-size measurements, we detected a total of 86 sedimentary events in 18 sediment cores collected from the Gulf of Aqaba. Of these events, 46 were classified as distinct turbidites, 9 events described as thick flooding sequences due to high precipitation periods lasting probably several decades, and the remaining 31 less distinct events as turbidites or floods. Careful chemostratigraphical inter-core correlations and radiometric dating of these events provide a robust submarine paleoseismic record for the last millennium. The results show that the historical earthquakes in 1068 and CE 1588 were major characteristic earthquakes in the gulf that probably ruptured all the main faults (Tiran, Arnona, Aragonese and Eilat faults) in the gulf. On the other hand, the historical earthquakes of 1839 and CE 1212 were smaller and triggered only local turbidites in the southernmost and northernmost parts of the gulf, respectively. Information on older major events, together with the 1068 and CE 1588 earthquakes, suggests a recurrence interval of 400–700 years (average = 560 years), indicating that the southern gulf is a ripe for a major earthquake.

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)

• from Bektaş et al. (2024)