Introduction

The development in geophysical techniques leads to routine collection of aeromagnetic and gravity datasets. The large data volume demands automatic techniques like the Naudy (Naudy 1971), Euler (Thompson 1982), and Werner deconvolutions (Werner 1953) for the interpretation of magnetic and gravity anomalies. The Euler deconvolution has emerged as a powerful technique for estimating the depth and the geometry of the buried magnetic sources (Thompson 1982; Reid et al. 1990).

The Euler deconvolution is a popular and effective method particularly suitable for automatic aeromagnetic and gravimetry interpretation because it requires no knowledge about the source magnetization vector. Interesting applications of this method are found, for example, in Reid et al. (1990), Paterson et al. (1991), Beasley and Golden (1993), Hearst and Morris (1993), Roest and Pilkington (1993) and Fairhead et al. (1997). More recent applications for high-resolution aeromagnetic data are presented by Peirce et al. (1998), Roy et al. (2000), Zhang et al. (2000), Mushayandebvu et al. (2001) and Silva et al. (2001). Thompson (1982) developed this technique for profile data while Reid et al. (1990) extended it for gridded data. Barbosa et al. (2000) extend Euler deconvolution to ground surveys with small number of observations.

Semi-automated (classical Euler deconvolution) interpretation methods in applied magnetometry are based on estimation of source parameters directly from the measured (processed) and/or transformed potential fields. These methods are built on introduction of a priori information on the properties of the desired solution, mainly at the mathematical level. In the majority of cases, the a priori information is based on the recognition of a predefined source type response in the interpreted data.

This approach does not fully solve the ambiguity of the inverse problem in potential field's interpretation. By means of this approach, we introduce into the solutions the so-called model error (Dmitriev et al. 1990)—this is connected with the problem of description of the complex real situations by simple models

The magnetic method is one of the best geophysical techniques to delineate subsurface structures. Generally, aeromagnetic maps reflect the spatial variations in the magnetic field of the earth. These variations are related to the distribution of the structures, the magnetic susceptibilities, and/or the remnant magnetization. In general, sedimentary rocks have lower magnetic properties than do igneous and metamorphic rocks, which tend to have a much greater magnetic content. Thus, most aeromagnetic surveys are useful in mapping the structures of the basement and the intruded igneous bodies. Fortunately, recent advances in the airborne magnetic systems have made it possible to cover large areas in a reasonable time and at a high resolution.

In this paper, we present an example of mapping the active subsurface structures using the high-resolution aeromagnetic data that were collected in the western part of the Central Arabian Shield. We have applied two structural index models (contacts and dikes). The solutions were evaluated for onshore and offshore regions based on the criterion of Thompson (1982). The investigated area is located in the west central section of Saudi Arabia between latitudes 21°00′ to 23°00′ N and longitudes 38°48′ to 41°30′ E (Figs. 1 and 2). This region is bounded in the west by the Red Sea, in the north by Harrat Rahat, in the east by Harrat Hadan and in the south by the Al Ta'if and Makkah areas.

Fig. 1
figure 1

Location map of the study area. The main tectonic features of the Arabian plate and surrounding regions are also presented

Full size image
Fig. 2
figure 2

Simplified geological map of the west central part of the Arabian Shield, which shows the traverse across the Ad Damm fault zone and the current interpretation of tectono-stratigraphic terranes (after Johnson 1997)

Full size image

Subsurface active structural mapping in this region led to the discovery of rift-related dike swarms and/or contacts that are cogenetic with isolated exposures of volcanic rocks and that provide new evidence of tectonic characteristics for the initial Red Sea rift magmatism. The style of intrusions in the area is related to the tectonic changes during rifting, and the depth of the magmas produced during rifting records a transition from continental to oceanic domains. Moreover, according to the earthquake clustering and as a result of our 3D Euler Deconvolution, the Ad Damm fault zone was re-evaluated according to its activity. The re-evaluation was performed to determine whether it was largely faulted contemporaneously due to magmatic eruption forces that resulted from the extension of the Red Sea Rift or whether it occurred because of other tectonic sources. Finally, deep magnetic source bodies were also identified (oceanic crustal structures) in the coastal plain and their relation to the ophiolite occurrences in the Arabian Shield.

Geological and structural settings

The Arabian Shield consists of a large outcrop of Precambrian rocks in the western Arabian Peninsula (see Fig. 1). Although granite and gneiss do not dominate the exposed rocks, there are large outcrop areas of both. Cenozoic basalts overly shield the rocks in some areas (Figs. 1 and 2). On the west side, the shield rocks are bordered by Cenozoic deposits of the Red Sea coastal plain, which include volcanics from a number of late Quaternary vents.

Geologically, the oldest rock exposed in the area of study (Fig. 2) includes quartzite, quartz-rich schist, marble, amphibolites, and basalt. According to Moore and Al-Rehaili (1989), five lava intrusion bodies were recognized in the study area. These bodies are Sharqah, Milh, Ju'ranah, Hashafat, and Nu'man complex bodies. The Milh complex consists of metadiorite, gabbro, and tonalite.

The rocks of the complex are metamorphosed and have a pervasive weak foliation. The Ju'ranah complex is a large body of hard-rock types, such as tonalite or granodiorite, that crop out in the northwestern area of the batholith Makka. This complex consists of quartz and diorite (Fig. 2). The Nu'man complex is a large body of granite gneiss that crops out in the traverse area as the Afar batholith (which was named after the largest mountain in the area, Jabal Afar). The batholith is exposed mainly south and southeast of the Ad Damm Fault zone but actually extends to the north into the Ta'if area (Fig. 2). The complex consists of biotite, monzogranite, orthogneiss, and hornblende-biotite monzogranite to granodiorite orthogneiss. At its contacts, the complex is interlayered and structurally conformable with amphibolite (metabasalt), and it contains amphibolite xenoliths, which suggests that the gneiss is a syntectonic intrusion that was emplaced during the deformation of rocks (Johnson 1997).

The tertiary sedimentary rocks are exposed in three faulted grabens (i.e., Ash Sham, Suqah, and Shumaysi). The Suqah formation is occasionally covered by volcanic lava and Quaternary sedimentary rocks. The quaternary sediments are exposed to the greatest extent in the Makka quadrant region, especially in the eastern part (e.g., Wadi Fatima).

Structurally, the study area is dominated by the NE Ad Damm trending fault zone (Fig. 2), which cuts diagonally across the area from northeast to southwest and divides it into two contrasted structural domains. The southern structural domain is characterized by north–south-striking foliation surfaces and south-plunging folds and lineations. The northern structural domain is characterized by northeasterly trends of the type that characterize the Jiddah terrene between the Ad Damm Fault and the Labunah thrust (Johnson 1997). In addition, the Wadi Fatima NE-trending fault is the major fault in the study area.

In our study area, the Jiddah terrane is bounded in the north by the Bir Umq suture zone, which is a well-delineated ophiolite-bearing fault zone (Johnson et al. 2002a; Johnson and Woldehaimanot 2003). The southern boundary has been set as the Ad Damm fault (Fig. 3). The eastern part of the terrane is very poorly exposed, and its relationship to the An Nimas terrane is not thoroughly understood. The Bidah terrane (Al Lith) protolith is composed dominantly of greywacke and other metasediments with a minor but locally important component of basalt-rhyolite bimodal volcanics (Roobol et al. 1983). The Bidah terrane may represent a back-arc basin. This protolith's boundary with the Jiddah domain to the northwest is tentatively placed along the Ad Damm Fault zone (Johnson and Woldehaimanot 2003), but the northern boundary region is complex, poorly mapped, and lacks geochronological studies.

Fig. 3
figure 3

a Terrane map of the Saudi Arabian Shield (from Stoeser and Frost 2006). The Protolith ages are provided in Ma, where those in parentheses indicate a younger arc assemblage that was emplaced into an older protolith. FZ fault zone. The red rectangle is the study area. b Arabian Shield terranes as defined by Stoeser and Camp (1985)

Full size image

Ophiolites, tertiary volcanism, and seismicity

The Arabian Shield consists of at least five Precambrian Terranes (Fig. 3) separated by suture zones (Schmidt et al. 1979). During the late Oligocene and early Miocene, the Arabian Shield was disrupted by the development of the Red Sea and the Gulf of Aden rifts, and from the mid-Miocene to the present, the region experienced volcanism and uplift (Bohannon et al. 1989).

Precambrian Ophiolites in Saudi Arabia occur along the NE- and NNW-trending linear belts that have been identified as the suture zones that separate the composite terranes with different lithostratigraphic units and evolutionary histories (e.g., Abdel-Gawad 1970; Bakor et al. 1976; Nassief et al. 1984; Fig. 3). This distribution of the Precambrian ophiolites and crystalline basement rocks has been affected by the NW-trending Najd fault system, particularly in the central part of the Arabian Shield (Fig. 3; Quick 1991). The uplift and the volcanism are generally assumed to be the result of hot, buoyant material in the upper mantle that may have eroded the base of the lithosphere (Camp and Roobol 1992a). However, details about the nature of the upper mantle, such as its thermal and compositional state, are not known.

Volcanic activity (the Harrats) is observed on the Arabian Shield (Fig. 3), which is likely to be related to the opening of the Red Sea and the mantle asthenospheric upwelling beneath western Arabia (e.g., Camp and Roobol 1992b).

The regional distribution of seismicity in the Red Sea indicates (Fig. 4) a concentrated distribution of events proximal to the main and axial trough in the southern portion. However, the concentration is not uniformly distributed but occurs in clusters on the ridge crests or near the transform faults of the rift axis. Other significant activities appear to occur along other portions of the central rift that do not have transform faults. The activities may be related to an intrusive mechanism, normal fault movements that are associated with the down dropping of blocks, or movements along undetected transform faults. The focal mechanisms for two earthquakes located near the southern Red Sea rift axis indicate a nearly pure strike-slip mechanism on the NE-trending planes, which suggests seismic activity on the rift transform faults (Thenhaus et al. 1986).

Fig. 4
figure 4

Instrumental Seismicity Map of the Arabian Peninsula and Adjoining Regions (the red color circles represent seismic epicenters, and the yellow circles are volcanic centers: 1990–2003 A.D.; Seismic Studies Center, King Saud University)

Full size image

Al-Amri (2004) stated that seismic activity in the Arabian shield appears to be low. There have been a few moderate seismic events composed of 25 (4.0 < Ms < 5.9) and 1 (Ms > 6.0) that have occurred since 1900. However, historical accounts (Ambraseys 1988) indicate that strong ground shaking has been felt in the northwestern portion of the shield during the 1068 event (Ms7.0) and 1588 event (Ms6.7). The Ms (7.0) was accompanied by ground cracking and fissuring that caused widespread destruction. The central portion shook in the year 1269 and was felt at Ta'if. In 1256, the City of Madinah also experienced ground shaking. However, the shaking phenomenon was more related to volcanic activity (Fig. 4). The Arabian Plate boundary extends east-northeast from the Afar region through the Gulf of Aden and into the Arabian Sea and Zagros fold belt.

The boundary is clearly delineated by teleseismic epicenters, although there are fewer epicenters bonding the eastern third of the Arabian Plate south of Oman. Most seismicities occur in the crustal part of the Arabian Plate beneath the Zagros folded belt (Jackson and Fitch 1981). The Zagros is a prolific source of large magnitude earthquakes with numerous magnitudes 7+ events occurring in the last few decades.

Quantitative interpretation of the aeromagnetic map

A magnetometer measures the variations in the earth's magnetic field at the ground, which corresponds to the magnetic susceptibility changes in the subsurface. It should be noted that the earth's magnetic field varies from approximately 30,000 nanotesla (nT) near the equator to approximately 65,000 nT near the poles. The details of the theoretical principles regarding magnetometry can be found in standard geophysical texts (e.g., Sharma 1997).

The magnetic anomaly map is a compilation of different magnetic surveys (that were acquired in 1960s and compared with more recent detailed surveys, which indicated its high quality), which were performed by the Saudi Geological Survey, Saudi Arabia. The total magnetic intensity anomaly grid is derived from a survey with a flight-line spacing of 100 m, a flight-line height of 800 m above the ground and flight-line direction E–W.

The aeromagnetic map (Fig. 5) shows two regions, the onshore and the offshore areas. The former area has a greater density of contour lines, which reflects the changes in intensity of the aeromagnetic data.

Fig. 5
figure 5

Total intensity of the aeromagnetic map of the study area, west central Arabian Shield, Saudi Arabia

Full size image

High-amplitude, short-wavelength magnetic anomalies are observed in the onshore part of the study area. Most of these anomalies are striking in the NNW–SSE and NW–SE regions, which suggest that the causative structures are related to the main structure trend of the Red Sea. In the offshore region of Fig. 5, the wavelength of the anomalies becomes progressively longer, and the density of contour lines is significantly less. Generally, the magnetic field over the central Red Sea offshore takes the form of low-amplitude, long-wavelength variations, on which a number of high-amplitude, short-wavelength anomalies that may be caused by recent intrusion are superimposed. The strong amplitude of these anomalies suggests that the causative sources are diabasic dikes and most likely have a genetic relation to the thermal system of the Red Sea.

Because there is no geological information available for the offshore part of the study area, the present study will provide important information about the source of these anomalies.

The total intensity aeromagnetic map (Fig. 5) registers the small scale variations, and a difference of approximately 200 nT is recorded. Several trends and anomalies are clearly observed in this map.

  1. 1.

    A W–NW trend of magnetic anomalies on the order of −50 to −100 nT marks the Red Sea area in the west. The central parts of the Red Sea are dominated by small variations, which range from −30 to −50 nT.

  2. 2.

    Low-gradient and large-amplitude anomalies with elongated shape marked the southern parts of the study area around the Makkah and AL Ta'if regions, which suggests that non-magnetite or low-magnetic sedimentary rocks could originate these anomalies.

  3. 3.

    Broad elongated positive anomalies, which strike in the E–W, ENE–WSW, NE–SW, and NW–SE directions, are observed in different sites of the RTP map that occupy the Jeddah Terrene, Makah, Harrat Kishb, and Harrat Hadan areas.

Depth estimation by 3D Euler deconvolution

The development in acquisition technique leads to a usual collection of magnetic and gravity datasets (Hartman et al. 1971; Reid et al. 1990). The large data volume demands automatic techniques such as (Naudy 1971), Euler (Thompson 1982) and Werner Deconvolution (Werner 1953) to interpret the magnetic and gravity anomalies. The objective of the 3D Euler Deconvolution process is to produce a map that shows the locations and the corresponding depth estimations of the geological sources of magnetic anomalies in a two-dimensional grid (Thompson 1982; Reid et al. 1990). This technique has become popular because it requires no prior information about the source magnetization, and it assumes no particular geological model. Thompson (1982) developed this technique for profile data, whereas Reid et al. (1990) extended it for gridded data.

The 3D Euler Deconvolution is performed by solving the equation (Thompson 1982; Reid et al. 1990; Silva and Barbosa 2003):

The 3D form of the Euler's equation can be defined (Reid et al. 1990) as:

$$ x\frac{\partial T}{\partial x}+y\frac{\partial T}{\partial y}+z\frac{\partial T}{\partial z}+\eta T={x}_0\frac{\partial T}{\partial x}+{y}_0\frac{\partial T}{\partial y}+{z}_0\frac{\partial T}{\partial z}+\eta b, $$
(1)

where x, y, and z are the coordinates of a measuring point, x 0, y 0, and z 0 are the coordinates of the source locations, b is the base level, and η is a structural index that defines the anomaly attenuation rate at the observation location (e.g., η = 0.5 for a contact, η = 1 for the top of a vertical dike or the edge of the sill, η = 2 for the center of a horizontal or vertical cylinder (pipe), and η = 3 for the center of a magnetic sphere or a dipole (Thompson 1982; Reid et al. 1990)). By considering four or more neighboring observations at a time (an operated window), the source location (x 0, y 0, z 0) and b can be computed by solving a linear system of equations generated from Eq. (1).

Then, by moving the window over the anomaly, multiple solutions for the same source are obtained (Ravat 1996). Finally, a location where these solutions tend to cluster is considered to be the most likely location of the source. It is well known that the Euler Deconvolution method assists in determining the depth and the location of magnetic sources from the observed magnetic data, but the method has several disadvantages. First, only a few simple geometries satisfy Euler's homogeneity equation (Blakely 1995). Second, the technique is best suited for sources for which the anomaly attenuation rate is constant, such as idealized magnetic sources. For arbitrary sources, the structural index changes with the source-to-observation distance, which may produce errors in the depth estimate (Ravat 1996). Third, the structural index must be assumed as a priori information.

However, Thompson (1982) and Reid et al. (1990) showed that the optimum structural index usually yielded the tightest clustering of solutions. Despite these disadvantages, the Euler technique provides satisfactory results in approximating the location of complex bodies. To apply the Euler method, the aeromagnetic data were first digitized at an equal interval of 250 m, and the gradients of the field were calculated in the frequency domain using the FFT method (Blakely 1995).

Discussion

In this study, the Euler method has been applied assuming two models (contact and dike) to locate the possible subsurface structures from the observed aeromagnetic data.

Figures 6 and 7 show the Euler solution results for each case (contact and dike). The solutions were accepted based on the criterion of Thompson (1982).

Fig. 6
figure 6

Solutions of the Euler method using a structural index of the contact model. The circular shallow magmatic-source bodies (green small circles that are nearly >500 m) were bordered with large blue rounded shapes, whereas the constrained source contacts (oceanic) were bordered with large yellow oval-like shapes in the Red Sea offshore regions trended from NW to WNW

Full size image
Fig. 7
figure 7

Solutions of the Euler method using a structural index of the contact model. The circular magmatic-source dikes (green circles, ranging from 0.5 to 1.0 km) were bordered with large light blue rounded shapes, which represent the centers of circular-shaped ring dikes that penetrate the continental part on the west central Arabian Shield along the Ad Damm Fault zone, whereas the constrained oceanic-source contacts were bordered with large yellow oval shapes in the Red Sea offshore regions and are trended from NW to WNW

Full size image

Generally, the magnetic field over the central Red Sea takes the form of low-amplitude, long-wavelength variations on which a number of high-amplitude, short-wavelength anomalies caused by a recent intrusion are superimposed.

The strong amplitude of these anomalies suggests that the causative sources are diabasic dikes and most likely genetically related to the thermal system of the Red Sea. This expectation was reinforced by the field observations, which indicated that the basement rocks on the onshore part of the study are intensively intruded by diabasic dikes trending in the NNW and NW directions (Izzeldin 1982).

On the offshore part of the study area, good clusters of Euler solutions are observed when the contact model is assumed (Fig. 6). Generally, the solutions delineate the boundaries of intrusive small magnetic bodies. The depths associated with the Euler solutions for these bodies range between 0.5 and 6 km. These bodies are most likely representing an intrusion that is associated with the Red Sea rift. Such a small intrusion in the basement would be expected to produce structures higher up in the sedimentary section.

Furthermore, most of the dike solutions (Fig. 7) are trending in the NNW and NW directions and are approximately associated with the main trend of the Red Sea. In a few locations of the dike model (Fig. 7), the solutions are found to be trending in the NE–SW direction. This trend is related to the transformed faults, which is associated with the Red Sea rifting (Said 1992). In the onshore part, the dike solutions appear to be more clustered than those obtained using the contact model. As mentioned by Thompson (1982) and Reid et al. (1990), selection of an appropriated index (model) will result in good clustered solutions. Thus, the dike model appears appropriated for the magnetic anomalies on the onshore part.

Magmatic activity around the Ad Damm fault zone in the Arabian shield

The Ad Damm fault runs from the coastal plain to beyond Ta'if over a length of approximately 130 km. This fault has been reported to have Precambrian right-lateral movement (Pallister 1984) and younger movements. This fault is considered a kinematic extension of one of the most important transform faults of the Red Sea rift on land. It extends in the direction of NE to ENE and continues north for a distance of 85 km from the coast of the Red Sea to the north Harrat Kishb, which is 44 km away from Makkah in the SSE direction. It is close to major cities such as Jeddah and Taif (right-lateral strike-slip fault) and consists of a series of deformational zones, which are displayed to be approximately 1 km wide. The recent seismological studies have shown that there are many recorded earthquake epicenters along this fault and on the scope of volcanoes in Harrat Ad Damm, where they are intersected with the Ad Damm fault zone (Merghelani 1981).

Based on the Cenozoic volcanism along the Red Sea, previous research has argued that the Red Sea is underlain by hot mantle (e.g., Coleman and McGuire 1988; Dixon et al. 1989). Ebinger and Sleep (1998) proposed rifts may present natural channels for hot material from mantle plumes to flow horizontally beneath thin lithosphere. The Red Sea has been proposed as such a channel for hot material from the Afar plume to cause the Cenozoic volcanism in western Arabia (e.g., Hansen et al. 2006; Park et al. 2007, 2008; Ritsema et al. 1999).

Barazangi (1981) stated that the overall lack of seismicity in the interior of the Arabian Peninsula suggests that little internal deformation of the Arabian Plate is occurring at present. There is widespread Quaternary volcanism along the Red Sea coast with at least one documented historical eruption in 1256 A.D.

Although all earthquakes associated with active volcanoes are ultimately related to volcanic processes, volcanic earthquakes are directly associated with magma movement, whereas tectonic earthquakes occur in zones that are separated from the principal areas of the magma movement (Heliker et al. 1986).

Seismicity may also be related to the transform faults in the Red Sea that continue onto land and other causes. To date, few on-land epicenters are accurately located, and there are few available focal mechanisms.

The offshore continuation of such fault is also inferred from the magnetic data of Al-Saud (2008). Her study showed that the seismic activity of the Ad Damm fault zone (Fig. 8) was relatively high in its southern part (southwest of Al Ta'if) and clustered in some parts, which might be attributed to local tectonic activities such as the upwelling of magmas or the intersection of two or more tectonic elements.

Fig. 8
figure 8

Estimated structural map in the west central part of the Arabian Shield showing moderate earthquake epicenters distributed along the tectonic faults (the red-colored bold trend lines are interpreted from the aeromagnetic anomaly map). The yellow and red circles that represent volcanic bodies are derived from the 3D Euler method. We notice that most of these volcanic bodies (yellow circles) beneath the offshore region trended in the NW direction (transform faults direction of Red Sea rift). Nevertheless, the yellow circle dikes beneath the Red Sea offshore region represent the centers of circular magmatic sources, which penetrate to the continental part (on the west central Arabian Shield as red circles) along the Ad Damm Fault zone. Most earthquakes (black circles) are clustering on the estimated fault. The estimated oceanic crustal structures were observed beneath the Red Sea offshore (the bold yellow lines, which are trended in NW directions with ∼10.5 km depth)

Full size image

Nevertheless, using magnetic Euler Deconvolution in this work, we have estimated many circular-shaped ring dikes (Figs. 7 and 8) located along both sides of the Ad Damm fault zone in the NE direction, which occur because of the upwelling of magmatic intrusions. These ring dikes may occupy the tension fractures that are formed under excess magma pressure, whereas some other ring dikes occupy the shear fractures under reduced magma pressure (Robson and Barr 1964). As a result, these ring dikes may have been formed because of the upwelling of magmas (eruptions). Therefore, the main reason that triggers the earthquake activity of the Ad Damm fault zone could be attributed to the actively melting volcanic magmas, which are first generated along the axial zone of the Red Sea rift and moved accordingly to its transform faults. Thus, the activity of this fault zone is facilitated by a slight continental extension due to the opening of the Red Sea rift.

Evaluating the crustal thickness (Moho boundary) in the central Red Sea offshore

The purpose of this section is to delineate some deep structures (oceanic) resulted from our Euler solutions and to describe their relation to the occurrence of Arabian Shield ophiolites and their relation to the Red Sea rift tectonics.

Neoproterozoic ophiolites, ranging in age from c. 870 Ma to c. 627 Ma, occur in several discrete suture and/or fault zones within the Arabian Shield and display a record of rift-drift, sea-floor spreading and collision tectonics during the evolution of the East African Orogen (Yildirim and Zulfiqar 2003). The selected Neoproterozoic ophiolites (e.g., Jabal Tharwah and Bi'r Umq ophiolites) lie on the Bi'r Umq suture between the Hijaz and Jiddah terranes in the west central Arabian Shield, Saudi Arabia (Figs. 2 and 3a).

The ophiolites within the Yanbu and Bi'r Umq suture zones in the west are among the oldest (870–740 Ma) in the shield, and they have chemical compositions typical of modern forearc oceanic crust (Yildirim and Zulfiqar 2003). These ophiolites are spatially associated with coeval and younger volcanic arc assemblages and were incorporated into the Arabian Shield during a series of collisional events that amalgamated these ensimatic arc terranes.

The Tharwah ophiolite complex (Nassief 1981; Nassief et al. 1984; Pallister 1984) consists of mafic–ultramafic rocks, which were preserved as a stack of steeply dipping, northwest- and southeast-vergent thrust sheets (obducted) exposed over an area of 13 km east–west and 6 km north–south (Fig. 3). Together with adjacent pelagic rocks, the ophiolite is part of the Labunah thrust zone (Ramsay 1986) and lies in the zone of deformed rocks that constitutes the southwestern part of the Bi'r Umq suture (Pallister 1984; Johnson et al. 2002b). The ophiolite is exposed in hills that rise 150–200 m above the adjacent Red Sea Coastal Plain.

Heat flow density in the Red Sea is generally high, reflecting the shallow depth to the asthenosphere and its recent age of intrusions. The early surveys, e.g. Girdler and Evans (1977), Haenel (1972), Girdler and Evans (1977), reported high heat flow density values of approximately 600 mW/m2 for the axial trough, whereas the flanks have values of twice the world mean (59 mW/m2). Heat flow in the Arabian Platform (45 mWm−2) lower than in the Arabian Shield (60 mWm−2); (Rolandone et al. 2013).

The fact that part of the mantle at depths below 15 to 17 km is much softened and partially melted leads to this assumption. Melts are concentrated at the base of the crust below the deeps. They transport the heat by convection, are responsible for the intrusions in the shallower parts of the crust and the volcanic activity and deliver the thermal energy which, together with the lateral variation of lithostatic pressure, drives the water circulation in the sediments, where the observed HFD distribution is depressed.

The 15- to 17-km thickness of the lithosphere mentioned above was obtained by microseismicity studies in the Red Sea (Makris et al. 1991a, b) that located the deepest hypocentres at this depth. The isothermal distribution obtained is in good agreement with values published by Haenel (1972). Haenel showed that, along strike of the median trough, the isotherm values change very significantly although the crust is of comparable thickness and composition. The map presented in Fig. 9 confirms just this phenomenon and shows that along the strike, the Red Sea has HFD values which increase and decrease independently of crustal thickness or lithology.

Fig. 9
figure 9

Heat flow density map of the Red Sea (from Makris et al. 1991b). Dots in the map indicate positions of the data used for this compilation. Units are given in milliwatt per square meter. See text for a discussion of this map. P1, P2, and P3 are heat flow profiles

Full size image

In addition, along the offshore part of the central Red Sea region, good clusters of Euler solutions are observed when we assume the contact or dike models (Figs. 6 and 7). Several obvious magnetic contacts of oceanic crust (Fig. 7) are marked by a steep deep-seated magnetic gradient with a NW lineament, which is located at a depth of 8–13 km (with an average depth of 10.5 km). This shallow crustal depth bodies are reflecting the shallow depth to the asthenosphere (shallow Moho depth) and its recent age of intrusions and consequence high heat flow intensity values.

Generally, the solutions delineate the boundaries of intrusive deep magnetic bodies that are trended in the NW direction and tacking the same trend of the Red Sea Rift. Magnetic bodies are located near the ophiolite occurrences (∼20 km to the west of Jabal Tharwah), where they are further extended to the southwestern corner of the study area beneath the Jeddah coastal plain (see Figs. 6, 7, and 8).

Al-Damegh et al. (2005) estimated the average crustal thickness of the late Proterozoic Arabian Shield, which was 39 km. The crust thins to approximately 23 km along the Red Sea coast. They observed shallower Moho for the stations along the Red Sea than in the shield. Moreover, they noticed that the change in the Moho depth becomes more pronounced when they compare stations LTHS and TAIF or FRSS and DJNS in particular (see Fig. 10), where a change of approximately 25 km in Moho depth occurs in a distance of less than 200 km. Toward the east, we observed a slight thickening of the crust near the platform, where the Moho depth reaches approximately 44 km.

Fig. 10
figure 10

Map showing the crustal thickness contour map (Moho depth) of the Arabian Shield region (after Al-Damegh et al. 2005). The seismic stations are shown as open triangles. The contour lines based on the kriging approach are shown in different colors, and the hand-drawn contour lines are shown in solid black. The figure shows also the outline of the Arabian Shield (open small circles), the escarpment (long dashes), and the Red Sea ridge (small dashes)

Full size image

Results of our estimated contacts of oceanic crustal depths using the Euler method along the Red Sea costal region, are consistent and well correlated with the previous Moho depth that was inferred from the seismological data analysis (Al-Damegh et al. 2005, see the correlations in Figs. 6, 7, and 10).

It is tempting to speculate that the Red Sea opening may be one of the consequences of the presence of a mega plume that extends from the core-mantle boundary into the upper mantle beneath east Africa, the Red Sea, and the western portion of the Arabian plate (Romanowicz and Gung 2002 and Nyblade et al. 2000).

It is possible that this mega plume may have rivaled those that drove the breakup of Gondwanaland and Pangea. Based on the surface geological observations and the widespread volcanic activity (Camp and Roobol 1992a), it appears that this plume has affected the lithospheric structure over a large distance (>4,000 km) in east Africa and western Arabia, and it may have resulted in a considerable thinning of the mantle part of the lithosphere. This thinning is partially documented in western Arabia based on the recent tomographic results of the high-attenuation and low-velocity uppermost mantle (Camp and Roobol 1992b; Al-Lazki et al. 2004 and Garson and Krs 1976).

Moreover, the site where the Red Sea sea-floor spreading occurred was a Proterozoic suture and a zone of weakness throughout the lithosphere that connects the proto-Arabian plate and the African plate (Garson and Krs 1976 and Dixon et al. 1987). The above combined observations may explain the relatively abrupt breakup of the Arabian plate and the anomalous nature of the Red Sea margin.

Saleh and Pašteka (2012) presented a case study of structural mapping using geophysical data of northern Red Sea region in Egypt. Their study provides information about buried subsurface structures with the help of the Euler method and bandpass filter maps. Integrated interpretation of the geologic-geophysical, seismicity and topographic data reveals the elements of subsurface structure in the study area. Close correlation between the deep faults, seismicity and geomorphic expression at surface indicates their recent activity and control over sedimentation. Several transform faults striking NE–SW (normal to the spreading axis of the northern Red Sea) were inferred from Euler deconvolution and bandpass of Total Magnetic intensity maps are marked by several interfering dipole anomalies.

Results of their study were able to delineate circular magmatic bodies (volcanic extrusions) depicted from bandpass filter magnetic map (with average depth 10 to more than 18 km) scattered along specific tectonic features of the study area (their locations are well correlated with Meneisy 1990; Plaziat et al. 1998 and Cochran 2005 ).

These bodies are most probably representing intrusions associated with the Red Sea rift, indicating that their source is the heated anomalous upper mantle, which consequently confirms the idea that the asthenospheric upwelling might be responsible for the subsequent rifting of the Red Sea.

Furthermore, in the present study, the magnetic structures are most likely representing the oceanic crustal source formed by the upwelling magmatic eruptions, which are mostly associated with occurrences of ophiolites as a result of thrusting (obducted) onto the edge of continental plates.

Well correlated results of Saleh and Pašteka (2012) with the present work are accomplished, particularly the two regions are related to Red Sea rift and most of oceanic crustal source bodies were formed due to upwelling magmatic eruptions.

Therefore, we can conclude that the most important result from this study is its significant oceanic crustal structures beneath the eastern side, which is approximately 20 km off the coast of Saudi Arabia beneath Jeddah region in the central Red Sea.

In addition, the most remarkable crustal (anomalous upper mantle) structural change (toward the coastal plain of the Red Sea rift) is the transition from the westernmost shield to the Red Sea rift offshore (Fig. 10). The thin oceanic crust of the offshore occurs over a short distance of 20 km from the Neoproterozoic ophiolites (e.g., Jabal Tharwah and Bi'r Umq ophiolites) and lies on the Bi'r Umq suture between the Hijaz and Jiddah terranes in central western Saudi Arabia (Figs. 2 and 8). The seismically determined transition is characterized at the surface of the Red Sea offshore by a sharp boundary between the Precambrian rocks and the Tertiary volcanic and intrusive rocks (Fig. 10).

Conclusion

In this paper, we presented a case study of structural mapping using the high-resolution aeromagnetic data of the west central Arabian Shield area and the central offshore of the Red Sea region of Saudi Arabia. The study provides information about the buried subsurface structures using the Euler deconvolution method. We have found that the dike bodies are appropriated models for the magnetic anomalies of the onshore part, whereas the contact models are a good selection of the offshore magnetic anomalies.

Generally, Euler deconvolution results illustrate that, the study area is affected by sets of fault systems, which are mainly trending in the NNW–SSE, NW–SE, and NE–SW directions.

Euler results of the present work also delineated the boundaries of very shallow and/or exposed small and circular-shaped ring dikes on the offshore part of the study area. These exposed magmatic intrusions are most possibly related to the opening of the Red Sea rift. These volcanic bodies are extended to the onshore part of the Arabian Shield, particularly beneath both sides of the Ad Damm fault zone, which verifies that the fault was initiated synchronously with the opening of the Red Sea rift. Thus, the earthquake activity of this fault may most likely occur because of the magmatic eruption forces instead of the tectonic source.

In addition, Euler deconvolution source values related to depths of certain deep oceanic structures beneath the offshore region were discussed according to previous seismological results. These values are possibly associated with the occurrences of ophiolites that have been thrust onto the edge of the continental plates. Their depths are ranging from 10 to approximately 13 km, which indicates that their source is the heated anomalous upper mantle, which consequently confirms that the asthenospheric upwelling might be responsible for the subsequent rifting of the Red Sea (Saleh et al. 2006).

A detailed 3D modeling is highly recommended to delineate more information about the shape and type (continual or oceanic type) of magnetization of these causative source bodies.