Active faulting in northern Israel

In this project I am collaborating with Amihai Sneh. At this stage of the research, we are presenting the tectonic framework of the northern sector of the Dead Sea Rift.

INTRODUCTION

The Hula depression in northern Israel and the Marj Ayoun depression in Lebanon and the blocks between them comprise an important sector of the Dead Sea Rift (DSR). Recent mapping of the Metulla quadrangle at a scale of 1:50,000 (map enclosed) provides new insights regarding the tectonics and stratigraphy of this highly complicated sector of the DSR. Based on geophysical and/or stratigraphic data as well as on kinematic considerations, investigators have suggested a left-lateral offset along the Rift. Dubertret (1932) suggested "a left-handed shear of 160 km" and later workers such as Quennell (1958, 1959), Freund (1961, 1965), Freund et al., (1968), Freund et al., (1970), Bartov, (1974), Druckman (1974, 1985), Steinitz et al., (1978), Bartov et al., (1980), Garfunkel (1981, 1989), Garfunkel et al., (1981), Joffe and Garfunkel (1987), Weissbrod (1981), and Sneh (1996) presented evidence for a 105 km left-lateral offset of the Arabian plate with respect to the African plate. In all these cases, the evidence includes isopachs and facies belts as well as fault systems.

The original geological survey of the Metulla area was carried out by Picard (1952) and Rosenberg (1960) in the Naftali Mountains, by Saltzman (1968) and Mor (1987) in the Golan, and by Dubertret (1951, 1960) in the Marj Ayun and Mount Hermon areas. Various parts of the quadrangle were later resurveyed by Glikson (1966a,b), Kafri (1991), Shimron (1998), Heimann (1985), Heimann and Ron (1987), Heimann et al. (1990), Hirsch (1996a) and Baer et al. (1997). Additional sources for the present mapping were Renouard (1955), Schulman (1966), Bein (1967), Ilani (1992), Zilberman et al. (2000) and Minster and Ilani (2001). The present mapping has allowed for another opportunity to compare the stratigraphy on both sides of the DSR. The stratigraphic column of rocks cropping out in the area includes the Jurassic and up to the Quaternary, and comprises units attributed to eight groups (Figure 1): the Arad and Kurnub groups are exposed in the Hermon anticline (at the Arabian plate) whereas in the west, in the Naftali Mountains (at the African plate) only the Kurnub Group crops out. The Judea, Mount Scopus and Avedat groups are found both in the northern Golan and in the Naftali Mountains. All these units, from Jurassic to Eocene in age are 3000 m in thickness and are of marine origin. The Miocene rocks are continental and more than 400 m thick. The overlying Plio-Pleistocene Bashan Group which builds the volcanic plateau of the Golan Heights and the northern margins of the Hula Valley, varies in thickness reaching several hundred meters, and comprises basaltic flows and pyroclastics. Contemporaneous sediments of the Hula Group (Horowitz, 1973) accumulated within the subsided DSR and alternate with the Plio-Pleistocene volcanic flows.

Here we present the main tectonic elements of this sector of the DSR in order to present a reliable and up to date tectonic framework.

TECTONIC FRAMEWORK

The Hula sector of the DSR is a deep depression within which a thick Pliocene to Holocne fill accumulated. It is considered to represent a pull-apart basin (Garfunkel, 1981) bounded by longitudinal and transversal marginal faults. However, the latter were deduced mainly from geophysical surveys (e.g., Rybakov et al., 2003, this volume) and there are only limited indications of them on the surface. Likewise, the marginal faults of the Marj Ayoun depression in Lebanon are not well defined. The area between these depressions comprises highly deformed blocks that form the Metulla saddle.

Faults and folds on the eastern side of the Hula depression, the Golan, and Mount Hermon

The Hula Valley is bounded to the east by the Hula eastern border fault which trends N-S. All along its length it is covered by sediments of the Hula Group and its trace can only be determined by geophysical methods (Rybakov et al, 2003). The precise vertical throw is difficult to measure because of the absence of clear guide horizons within the basalts of the Golan Heights and within the Hula depression. Ar/Ar ages of basalts from the Notera 3 borehole drilled in the Hula Valley indicate that the Dalwe flows which occur at the western margins of the Golan at approximately 200 m msl, were encountered in the borehole between 500 and 900 m (-400 and –800 msl), i.e., a throw of 600 to 1000 m (Heimann and Steinitz, 1988a, 1989). However, it should be noted that many basalt units exposed at the rift margins show steep tectonic dips towards the eastern N-S trending faults (Sneh, 1996).

The surface continuation of the Hula eastern border fault is the Azaz fault which runs along the western foot of Tel Anafa, Tel Qalil, Giv’at HaEm and Giv’at Azaz and continues in a NE direction under the name of Si’on fault; it is known farther to the north in Lebanon as the Rachaya fault. Several other faults run along the slopes of the Golan. The Kefar Szold fault trends nearly parallel to the border fault; east of Kefar Szold it runs along the foot of the Zuq Zaqef cliffs. The Shamir fault first parallels the border fault then changes direction toward the NE ending north of the Shamir ‘windows’. This fault is delineated by a prominent morphological step on the basalt slope facing the Hula Valley.

The Azaz, Kefar Szold and Shamir faults displaced Plio-Pleistocene basalt flows, and we were unable to detect faults that affected only “old” basalts (e.g., Dalwe flows) as suggested by Michelson (1972) and Mor (1973). Consequently, we could neither divide the faults into age groups nor could we place them geographically in a time-arranged pattern as proposed by Heimann (1990), i.e., faults becoming younger in age as they approach the DSR.

The Hermon anticline is part of the Syrian Arc, a series of folded strata that extends from the Sinai Peninsula through Israel to the Palmyra in Syria. The generally accepted time of its formation is from the Cretaceous to the Early Tertiary (Picard, 1943). The Shamir 'windows' and a few other small 'windows' (Mimran et al., 1985) present evidence of local post-Campanian - pre-Maastrichtian and Paleocene tectonic activity, and, two unconformities (respectively) were identified. However, Freund (1980) proposed that like Mount Lebanon, the formation of the Hermon anticline is connected to a strike-slip along the DSR.

Faults which generally parallel the axis of the Hermon anticline, divide it into several blocks. The two main blocks lie west and east of the Si’on fault. Other structures are the Barahta Valley graben associated with the Si’on fault - Rachaya fault (Heimann et al., 1989) and the Newe Ativ graben, whose age was determined as Neogene on the basis of tectonic analysis (Baer et al. 1997). A series of Lower Cretaceous volcanic rocks related to a collapsed caldera was found in the graben area (Shimron, 1998) but has no connection to its time of formation.

Hula western border faults and the block of Naftali Mountains

The Hula Valley is bounded in the west, as in the east, by a large border fault which trends N-S from the mouth of Nahal Qadesh (in the Metulla quadrangle) to the southern part of Qiryat Shemona. The eastern slope of the Naftali Mountains is much steeper than that of the Golan Heights and therefore the course of the main fault in the west is closely associated with the foot of the slope. Nevertheless, the Hula Group sediments cover parts of the fault where its precise location can only be inferred.

Two N-S parallel faults, En Te’o fault and Yesha fault, occur 500 and 1000 m to the west of the main border fault, merging into a single fault on the eastern slope of Mount Nezer, and disappearing about 1 km south of the Qiryat Shemona cable-car. They initially form two distinct morphological steps (~70 and 300 m above the Hula Valley). Patches of alluvium pave the upper step. The stratigraphic throw of the Yesha fault is just a few tens of meters; and that of the En Te’o fault is more than 100 m. There are no grounds to consider either of them older than the Hula western border fault. Another strike-parallel fault was postulated along the high-lying step at Margaliyyot. (Apart from the characteristic morphology, this fault may explain an otherwise unexplained thickness decrease of the relevant stratigraphic units, e.g., Karkara Mbr., in this area.)

There seems to be no doubt regarding the association of En Te’o and Yesha faults with the subsided Hula Valley. However, assessing their exact time of formation or the last time they were active, requires tools of greater precision, although two geomorphological observations might be helpful in this regard. They both lead to the conclusion that the last activity along these faults took place in Late Pleistocene and even Holocene times. The first observation concerns the formation of landslide scars above the faults and the accumulation of slope movement material below the faults. Both features are fresh and could not have survived long periods of weathering and erosion. In the Qiryat Shemona area, complex of slope movement material and three overlying lobes of rock and debris flow deposits were identified, indicating a young, direct or indirect, tectonic trigger. The second observation concerns the development of the drainage rills on the slope. When crossing the N-S parallel-to-rift faults, they lack significant relief and when the lower step is reached, they disappear completely. This state of landform development suggests that processes targeted at forming equilibrium with the erosional base level have hardly begun.

South of the Margaliyyot fault, the ridge-forming morphology of the Naftali Mountains follows the geological structure. The Cretaceous beds exposed on the slope (e.g., Manara Cliff), generally dip gently westwards beneath the thick Tertiary formations that dominates the Yir’on-Nabatiya syncline. Several E-W faults cross the ridge and their relation to the rift is not clear; some are short and appear to die out away from the rift.

The Misgav Am and Metulla blocks

Picard (1952), Glikson (1966a), Kafri (1991), and Ron et al. (1997), described intensive deformation of various Cretaceous and Neogene strata in the Misgav Am and Metulla blocks, respectively. Picard (1952) distinguished between the main folding phase at the beginning of the Miocene and another younger deformation phase which resulted from compression between the blocks on either side of the Tel Hay fault. Ron et al. (1997) studied mesostructures in the Misgav Am and Metulla blocks and suggested an E-W shortening during the Neogene.

West of Mount Noter, the Roum fault branches towards the northwest and then traces a northward course toward the villages of Arab Salim and Roum, sub-parallel to the NE-trending Yammunneh fault. This setting forms the Gebel Niha elongated block (in Lebanon), which is squeezed between the two faults. The Gebel Niha block is highly deformed and tilted generally to the west. Towards the Roum fault, strata in places dip more than 50º westward (Dubertret, 1951).

The Misgav Am block extends to the west of the Qiryat Shemona fault and Roum fault and is bounded to its west by the Margaliyyot fault and its extension in Lebanon. The Margaliyyot fault branches off the Qiryat Shemona fault at the southern entrance to Qiryat Shemona. The block reached a maximum width of about 3 km opposite the bending point west of Mount Noter and decreases northwards to less than 1 km at the Litani knee. Farther north, its width remains small all the way to Roum. The block is intersected by a set of faults generally trending SW-NE. They divide the block into sub-blocks, each displaying highly deformed beds, which in places are near-vertical. Gebel Nabi el Aouedi, west of Misgav Am, is crossed by a fault with a N-S trend, and the strata along it are overturned. The long and narrow sub-block along the Litani River and northwards exhibits vertical strata all along (Dubertret, 1951).

The Metulla block lies between the Qiryat Shemona fault and the Tel Hay fault. It is only 2 km wide at the Mount Noter bending area and like the Misgav Am block, it is divided into several NE-SW sub-blocks which are also highly deformed. In the south, on both sides of the Muftalah fault, the Kefar Giladi beds are vertical and in places display intensive contortion. The Metulla fault to the north is a thrust fault with the Timrat Formation beds overriding the Kefar Giladi lacustrine beds on the northwestern side of the fault. The Har Zefiyya fault occurs between the Bar Kokhba limestone on its east side (with dips of 70º and more toward the E-SE) and the “overlying” Karkara beds on the west (with similar dips and directions) and which are probably overturned in the subsurface.

Based on the following stratigraphic considerations, it may be concluded that the Qiryat Shemona-Yammunneh fault separates the African plate from the Arabian plate and is the master fault (i.e., transform) through which most of the left-lateral offset along the DSR took place.