Dawood et al. (2024) used the ³⁶Cl exposure dating method to recover the last 30 ka scarp exhumation history of three fault segments from the Bet Kerem fault system. Results indicate that the three faults were active simultaneously in at least three distinguished activity periods, during which a minimum of 1.2 m of surface rupturing occurred in each period. This synchronized fault activity in turn suggests that the three individual earthquakes ruptured the faults during the same events.

• from Dawood et al. (2024:2)

The ³⁶Cl cosmogenic exposure dating method was developed and applied on carbonate normal fault scarps over the last two decades (Akçar et al., 2012; Benedetti et al., 2002; Goodall et al., 2021; Iezzi et al., 2021; Mitchell et al., 2001; Mozafari et al., 2022; Schlagenhauf et al., 2010; Zreda & Noller, 1998). This method is based on the fact that the cosmogenic isotope ³⁶Cl is primarily produced and accumulated in the carbonate fault rocks due to the interaction between cosmic radiation and Ca-rich minerals (Gosse & Phillips, 2001). This production of ³⁶Cl atoms in the fault rocks allows to determine the exhumation history of the scarp during Late Pleistocene to Holocene periods. Hence, both, the ages and magnitudes of the surface slip are obtained for the most recent earthquakes that exhumed the exposed fault scarp (Mitchell et al., 2001; Schlagenhauf et al., 2010).

• from Dawood et al. (2024:14)

The seismic activity recorded along the Bet Kerem fault system exhibits a close correlation with periods of intense seismic activity along the Dead Sea Transform. The earliest recorded activity along the Bet Kerem fault system coincides with a series of earthquakes along the Dead Sea Transform and its branching Carmel Fault (CF), which occurred approximately 4–5 ka (Braun et al., 2009; Gluck, 2001; Katz et al., 2009; Matmon et al., 2006; Rinat et al., 2014). Further evidence of this synchronized seismic activity is provided by a shaking event recorded along the CF, dated approximately to 10.4 ± 0.7 ka (Figure8; Braun etal., 2009), which corresponds to the third activity period at the Bet Kerem fault system. Additionally, records of shaking events along the eastern shore of the Sea of Galilee correlate with periods of intense seismic activity along the Dead Sea Transform (Figure 8). These records include five shaking events, the youngest of which is 4–5 thousand years old and aligns with the first activity period. The remaining four events, of which the most recent is 9,200 ± 1,900 years old, could be associated with either the second or third activity period event. The remaining three events have been dated to between 30 and 40 ka (Katz et al., 2009). This period of high seismic activity coincides with the fifth activity periods and the pre-exposure duration along the Deir Al-Assad segment (Figure 8). These observations of synchronized seismic activity between the Bet Kerem fault system and the Dead SeaTransform suggest that earthquakes along the Dead Sea Transform and its branches may trigger seismic activity along the Bet Kerem fault system and vice versa (Scholz, 2010, 2019).

• from Dawood et al. (2024:13-14)

Evidence from the archeological record around the Bet Kerem fault system suggests two distinct seismic events, likely caused by earthquakes along the fault system. The first event, roughly 16 km northwest of the fault system, damaged a castle near Kibbutz Kabri (Figure 7). This shaking has been interpreted as a seismic event with an age of 3.8–3.7 ka (Lazar et al., 2020). The age of this shaking event aligns with the first activity period of the Bet Kerem fault system (Table 1), indicating that it may have been caused by earthquakes along the Bet Kerem fault system.

The second event is evident in the Peqi’in Cave, located 5–4 km north of the fault (Figure 7). Here, evidence suggests a collapse involving cave debris and speleothems, possibly triggered by a strong earthquake that occurred sometime after 6.5 ka (Bar-Matthews et al., 2003; Harney et al., 2018). While this age doesn’t match any known activity period (Table 1), the collapse could be linked to the first activity period or an earthquake along the fault that didn’t produce a surface rupture detectable by the applied ³⁶Cl dating method. It is worth noting that the two recorded events, the collapse of Peqi’in Cave and the damage to Kabri castle, could have been caused by earthquakes along the Dead Sea Transform rather than the Bet Kerem fault system. While we cannot definitively determine which fault system was the source of the earthquakes, the temporal association between the seismic events to the timing of the Bet Kerem fault system earthquakes, as well as their spatial proximity (compared to the Dead Sea Transform), suggest that the Bet Kerem fault system is the more likely source of these events.

• from Dawood et al. (2024:11)

Before we discuss the results, it is important to note, that due to the limitation of the ³⁶Cl approach, seismic events that generate less than 25 cm of surface rupture with a recurrence time of less than a few hundred years, cannot be separated as distinguished events. Therefore, the number of events indicated for each segment is a minimum value. This implies that some so-called “events” may in fact include several earthquakes that occurred within a few hundreds of years or one single event that generated a large amount of slip (Schlagenhauf et al., 2010). Hence, we will refer to the scarp exhumation events as activity period.

• from Dawood et al. (2024:11-13)

Our results suggest the amount of surface slip that occurred at each of the dated faults during each activity period. However, our results do not directly provide information about the number of earthquakes during an activity period, earthquake magnitude, surface rupture length, or maximum surface slip generated by each earthquake. Currently, there is no direct paleoseismological technique that can accurately determine all these parameters. However, the probable magnitude, surface rupture length, and maximum surface slip can be approximated using empirical relationships that link earthquake magnitude to surface rupture length and maximum surface slip. Therefore, if we determine one of these parameters, we can use these empirical relationships to determine the two other missing parameters.



The studied fault scarps are all less than 4.5 km long (Figures 1 and 2). Based on empirical relationships, earthquakes that rupture short faults are expected to produce surface displacements of 10–30 cm and have a magnitude of approximately 5.5 (Figure 6) (Pavlides & Caputo, 2004; Wells & Coppersmith, 1994; Wesnousky, 2008). However, the smallest observed surface rupture on each of the studied fault segments is more than 1.2 m (Figure 5 and Table 1). This suggests that the large amounts of surface slip observed on these faults are likely not caused by a single earthquake rupturing the entire fault segment. Therefore, these large amounts of surface slip along relatively short fault segments can be generated in two ways: (a) The measured fault-slip is generated by a dense sequence of moderate (magnitude around 5.5) along a single segment, (b) The investigated segment is part of a linked fault array that ruptured the Earth's surface simultaneously during a single earthquake that is, multi-segment earthquake. If the first mechanism is true, at least four earthquakes would have had to occur on each segment during an activity period. However, earthquakes that rupture only single, short fault segments (5–10 km long) rarely cause surface ruptures at all. Even when they do, the maximum surface slip is only a few centimeters. This is despite existing models suggesting that these fault segments are capable of generating a maximum surface slip of 10–30 cm (Abdelmeguid et al., 2023; Manighetti et al., 2005; Natawidjaja et al., 2021; Pavlides & Caputo, 2004; Wells & Coppersmith, 1994; Wesnousky, 2008). Therefore, many more than four earthquakes must have occurred along each segment during each activity period to generate the observed surface displacement. The rare occurrence of relatively short (4–5 km) surface rupture by earthquakes and the large number of earthquakes required to generate the observed displacement make it very unlikely that this is the case for the Bet Kerem fault system.

Multi-segment earthquakes are the most common type of earthquake to generate surface rupture in normal fault systems. These earthquakes rupture multiple segments at once, resulting in long surface rupture lengths and large amounts of surface slip (Abdelmeguid et al., 2023; Bello et al., 2021; Bernard & Zollo, 1989; Chiaraluce et al., 2017; Iezzi et al., 2019; Manighetti et al., 2005; Natawidjaja et al., 2021; Wells & Coppersmith, 1994; Wesnousky, 2008). Our data show that the dated fault segments in the Bet Kerem fault system experienced synchronized activity periods, and that the large observed surface displacement is consistent with the occurrence of multi-segment earthquakes (Figure 5 and table 1). The along-fault maximum scarp height projection (MSP) (Figure 2) also supports multi-segment rupture. Fault segments ruptured by multi-segment earthquakes commonly show a scalene triangle shape of the MSP (Manighetti et al., 2005, 2007, 2009; Wesnousky, 2008), which is also observed in the MSP of the dated segments (Figure 2). Therefore, the dated fault segments in the Bet Kerem fault system were likely ruptured by multi-segment earthquakes. In general, considering all the above mentioned observations, one can suggest that for similar cases of normal fault systems, where paleoseismological data shows large surface slip (>1 m) on relatively short faults (LT 5 km) they were likely ruptured by multi-segment earthquakes.

Based on the previously suggested empirical relationships, we hypothesis that multi-segment earthquakes in the Bet Kerem fault system typically have a magnitude of 6.5, surface rupture length of approximately 17 km, and surface slip of 0.7–1.5 m (Pavlides & Caputo, 2004; Wells & Coppersmith, 1994; Wesnousky, 2008). We estimate the 17 kmlongsurfacerupture based on the combined length of the three dated fault segments and the Nahf segment, which are the four fault segments that exhibit seismically exhumed fault scarps in the Bet Kerem fault system (Figures 1, 2, and 6) (Bhat et al., 2007; Scholz, 2019). Using empirical relationship models, we estimate both the surface slip and the magnitude of the earthquake based on this length (Pavlides & Caputo, 2004; Wells & Coppersmith, 1994; Wesnousky, 2008). It is worth noting that if the length of the surface rupture is actually smaller, both the surface slip and the magnitude of the earthquake will also be smaller and vice versa.

To estimate the minimum number of earthquakes within each activity period, it is essential to determine the maximum surface slip that occurred at each period. Analysis of surface ruptures caused by earthquakes reveals that the distribution of surface rupture along a fault follow the shape of the fault scarp. For instance, if the fault scarp profile resembles a scalene triangle, the distribution of surface rupture magnitude along the fault will also exhibit a scalene triangle shape, with the maximum surface slip occurring at the point where the fault scarp is highest (Iezzi et al., 2019; Puliti et al., 2020). Our two sampling sites, as well as the Mitchell et al. (2001) sampling site are located close to the maximum height of the fault scarp (Figure 2). Consequently, the sampling sites show the maximum surface slip that generated at each activity period along each of the segment. Based on that, we assume that the largest surface slip observed during each of the activity period is a good indicator of the maximum surface slip that occurred during that period.

During the first activity period (4–5 ka), the largest surface slip recorded was 2.8 m (Table 1). This suggests that at least two to four earthquakes occurred during this time, based on known correlations between surface slip and rupture length, which indicate that a 17-km-long surface rupture would produce a surface slip of 0.7–1.5 m (Figure 6). Similarly, maximum surface slip observations from other activity periods indicate that at least 4–10 earthquakes occurred during the second activity period, two to four earthquakes occurred during the third activity period, and at least two earthquakes occurred during the fourth and fifth periods.

The time between earthquakes within an active period was too short (hundreds of years at most) to be accurately determined using ³⁶Cl exposure age dating method. Therefore, each activity period likely represents a distinct earthquake cluster. The recurrence interval of these clusters is estimated to range from 3.5 to 5 thousand years during the Holocene and the very Late Pleistocene and between the fifth and fourth activity periods along the Deir Al-Assad segment (Figure 5, Table 1). Notably, there was a period of approximately 13 thousand years without earthquakes between the fourth and third activity periods in the Deir Al-Assad segment. This period of inactivity coincides with the pre-exposure duration observed on the Sajur and Nahef east segments (Figure 5 and Table 1).

Our results, therefore, point to three superimposed recurrence interval wavelengths on the Bet Kerem fault system:

1. Maximum a few 100's of years interval, which separates between discrete earthquakes within a cluster (activity period).
2. A 3.5–5 ka interval between activity periods during the Holocene and Late Pleistocene (activity cycle).
3. An approximately 13 ka interval of total quiescence, that separates between activity cycles.

This phenomenon of the long quiescence period, followed by an activity cycle during which the faults release the strain that has been accumulated during the long quiescence period has been observed in other paleoseismological studies (e.g., Friedrich et al., 2003; Marco et al., 1996; Rockwell et al., 2000; Schlagenhauf et al., 2011), and referred to as “earthquake supercycles” (Salditch et al., 2020). Thus, earthquakes within the Bet Kerem fault system seem to follow the supercycle pattern.

Source - Wells and Coppersmith (1994)

Variable	Input	Units	Notes
Average Displacement		cm.
Maximum Displacement		cm.
VS30		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
MW		unitless	Moment Magnitude for Avg. Displacement
MW		unitless	Moment Magnitude for Max. Displacement
Variable	Output - Site Effect Removal	Units	Notes
Ireduction		unitless	Reduce Intensity Estimate by this amount to get a pre-amplification value of Intensity

Calculate   

Variable	Input	Units	Notes
Max. Vertical Displacement		m	Max. Vertical Displacement
Variable	Output - not considering a Site Effect	Units	Notes
MS		unitless	Avg. Surface Magnitude from Pavlides and Caputo (2004) Eqn. 2 (developed for the Aegean)
MSmin		unitless	Min. Surface Magnitude from Pavlides and Caputo (2004) Eqn. 2 (developed for the Aegean)
MSmax		unitless	Max. Surface Magnitude from Pavlides and Caputo (2004) Eqn. 2 (developed for the Aegean)

Calculate   

• Source - Wells and Coppersmith (1994)

Variable	Input	Units	Notes
Rupture Length		km.	Rupture Length
Variable	Output - not considering a Site Effect	Units	Notes
MW		unitless	Moment Magnitude

Calculate   

Variable	Input	Units	Notes
Rupture Length		km.	Rupture Length
Variable	Output - not considering a Site Effect	Units	Notes
MW		unitless	Moment Magnitude from Ambraseys (1988) (developed for the Middle East)
MW		unitless	Moment Magnitude from Bonilla and Lienkaemper (1984)
MS		unitless	Surface Magnitude from Ambraseys and Jackson (1998) Eqn. 2
MS		unitless	Surface Magnitude from Ambraseys and Jackson (1998) Eqn. 3
MW		unitless	Moment Magnitude from Ambraseys and Jackson (1998) Eqn. 11
MS		unitless	Avg. Surface Magnitude from Pavlides and Caputo (2004) Eqn. 1 (developed for the Aegean)
MSmin		unitless	Min. Surface Magnitude from Pavlides and Caputo (2004) Eqn. 7 (developed for the Aegean)
MSmax		unitless	Max. Surface Magnitude from Pavlides and Caputo (2004) Eqn. 5 (developed for the Aegean)

Calculate   

Variable	Input	Units	Notes
Fault Length		km.
Fault Width		km.
Variable	Output	Units	Notes
Mw		unitless	Moment Magnitude computed using Wesnousky (2008)
Mw		unitless	Moment Magnitude computed using Hanks and Bakun (2008)
Fault Area		km.2

Calculate   

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 V_S30 greater than 655 m/s you will get a positive number, indicating that the site amplifies seismic energy. If you enter a V_S30 less than 655 m/s you will get a negative number, indicating that the site attenuates seismic energy rather than amplifying it. Intensity Reduction (I_reduction) is calculated based on Equation 6 from Darvasi and Agnon (2019).

V_S30 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 V_S30 for a number of sites in Israel. If you get V_S30 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 V_S30 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.