Introduction

Ubhur district is located at the eastern coast of the Red Sea to the north of Jeddah city. It is considered as the northern extension of Jeddah city, where most of the new investments and constructions are located. It extends from the north of King Abdul-Aziz International Airport in the south to Asfan Road in the north. Figure 1 shows the location of the study area and its surface geology.

Fig. 1
figure 1

Location and geologic map of Ubhur area (after Johnson 1999)

Full size image

The Arabian Plate is a single tectonic plate, which is surrounded by relatively high active tectonic zones (Fig. 2). Active tectonics of the region is dominated by the collision of the Arabian Plate with the Eurasian Plate along the Zagros and Bitlis thrust systems, rifting and seafloor spreading in the Red Sea and Gulf of Aden, and strike-slip faulting occurs along the Gulf of Aqaba Dead Sea fault system. It is evident that the strong earthquakes, as well as most of the microearthquakes, have occurred at previously defined active tectonic trends (Fig. 2). The Gulf of Aqaba is the source of the 22 November 1995 earthquake (Ms = 7.3), which is the largest recent earthquake which shocked the Kingdom (Roobol et al. 1999). Since 1900, some moderate and large intraplate earthquakes hit the Arabian Peninsula; for example, Ambraseys et al. (1994) mentioned the Sa’dah earthquake on 11 January 1941 with Ms = 6.5. Recently, the Saudi National Seismological Network detected several moderate earthquakes within the Arabian Shield; El-hadidy (2015) also mentioned these earthquakes (e.g., Tabuk earthquake on 22 June 2004 (Mw = 5.1), Harrat Lunayyir earthquake on 19 May 2009 (Mw = 5.7), and Sabia earthquake on 21 January 2015 (ML = 5.1)). Figure 3 shows the seismicity of the Arabian Peninsula and its surroundings (Zahran et al. 2016).

Fig. 2
figure 2

Major tectonic elements surrounding Saudi Arabia. The Pan-African Najd fault system, the Aqaba–Levant intracontinental transform boundary, the Mediterranean–Bitlis–Zagros convergence zone, and the East African rift are highlighted. AFFZ = Alula-Fartak fracture zone; BM = Bab al Mandeb; D = Danakil horst; DA = Dahlak Archipelago; ET = Ethiopian trap series; G = Cape Guardafui; GOA = Gulf of Aqaba; GOS = Gulf of Suez; MAP = Makran accretionary prism; MER = main Ethiopian rift; MS = Mesozoic Marib-Shabwa (Sab’atayn) Basin; SI = Socotra Island; SOC = Semail oceanic crust; SR = Sheba Ridge; SS = Shukra al Sheik discontinuity; RS = Ras Sharbithat; YT = Yemen trap series; ZI = Zabargad Island (After Bosworth et al. 2005)

Full size image
Fig. 3
figure 3

Distribution of epicenters of shallow earthquakes (depth < 30 km) and M > 4 in and around the Arabian Peninsula which occurred up to May 2015 (Zahran et al. 2016)

Full size image

Al-Qarni (2015) studied the microzonation of the urban area of Jeddah starting from the Jeddah Islamic Port in the south to the King Abdul-Aziz International Airport in the north. Al-Qarni (2015) acquired microtremor data at 155 sites and 57 (1D) profiles of the multichannel analysis of surface waves (MASW). The results of the microtremor analysis show that f0 ranges between 0.48 and 4.87 Hz. This wide range of f0 reflects the complex and variant structure of the surface of soil in Jeddah. According to Vs30 measurements, the study area was divided into site classes D, C, and B.

The occurrence of such moderate earthquakes close to the urban areas built on weak soils can cause significant damage; therefore, the use of the produced f0 and A0 distribution maps is very important for seismic risk reduction in the Ubhur district.

Geological setting of northern Jeddah

According to the studies of Morris (1975), Bahafzullah et al. (1993), Al-Saifi and Qari (1996), and Aqeel (2007), the geological setting of northern Jeddah (including Ubhur district) can be classified into (1) hills of Precambrian igneous and metamorphic rocks, (2) alluvial fan deposits and sand bodies, (3) isolated lagoons and sabkha deposits, and 4) near-shore marine deposits and coralline terrace. The Saudi Geological Survey (SGS) produced a surface geology map of the study area, as a part of a large-scale geology map of the western Saudi Arabia. Figure 1 shows the geologic setting of the Ubhur area (Johnson 1999).

Bahafzullah et al. (1993) studied the geological setting of northern Jeddah and identified three distinct geomorphological assemblages. The first assemblage starts from the west and includes near-shore marine deposits and is composed of a coralline terrace, which extends a few meters above the mean sea level. The second assemblage consists of isolated lagoonal sediments and salt-encrusted intertidal flats. The third assemblage lies to the east and is occupied by a veneer of alluvial fans, eolian sand bodies, and wadi plains. In addition, other features include isolated remnants and dissected hills of Precambrian crystalline rocks and outcrops of tilted and faulted Tertiary clastic sedimentary rocks. In some areas, the Tertiary deposits and the Precambrian rocks are capped by flat-lying basalt of Harrat Rahat and its erosional remnants.

Geotechnical and Vs30 data

Ali and Hossain (1988) studied the geotechnical and geochemical characteristics of Ubhur subsoils. They investigated the soil up to 15-m depth and classified the soil composition of northern Jeddah into five soil units and three rock units. The classified units are illustrated in Fig. 4, and they are (1) gray to light brown soft to medium sensitive sandy silty clay with gypsum crystals and shells, (2) stiff to very stiff brown sandy silty clay, (3) gray very loose to loose silty sand and sandy silt, (4) light brown to gray very loose carbonate silty sand, (5) gray medium carbonate gravelly sand, (6) yellowish to creamy white coarse-grained shelly coralline limestone, weak to moderately weak, (7) Tertiary to Quaternary basalt, and (8) tertiary deposits of sandstone, siltstone, or claystone.

Fig. 4
figure 4

The identified rock and soil units in Ubhur area (After Ali and Hossain 1988)

Full size image

Ali and Hossain have also divided the study area into seven zones as the following:

Zone I occupies the northern area of the Ubhur district and consists of soft clay (unit 1) overlying stiff clay (unit 2) underlain either by medium dense carbonate gravelly sand (unit 5) or by a hard clay or claystone (unit 8). The soil profile of zone I is described in Fig. 5.

Fig. 5
figure 5

Zone I soil profile (After Ali and Hossain 1988)

Full size image

Zone II extends parallel to the coast in the north Ubhur district and is composed of loose carbonate silty sand (unit 4) which occurs as gravelly sand (unit 5, whose lowest 1 m is very dense), underlain by coralline limestone (unit 6). The soil column of this soil is shown in Fig. 6.

Fig. 6
figure 6

Zone II soil profile (After Ali and Hossain 1988)

Full size image

Zone III covers a small area about 1 km2 at the upstream end of Ubhur creek. Unit 3 is constituted by very loose silty sand with about 11-m thickness whose lower part is slightly clayey overlying a layer of soft clay (unit 1) which is underlain by medium dense carbonate gravelly sand. Figure 7 presents the soil profile of this zone. This zone is considered as part of Wadi Al-Kura which originates from the crystalline rocks eastward and was linked to the creek.

Fig. 7
figure 7

Zone III soil profile (After Ali and Hossain 1988)

Full size image

Zone IV includes a major terrace and two minor ones of medium dense carbonate gravelly sand (unit 5) covering an area about 22 km2 at the northern bank of Ubhur creek. A thin layer of about 10 cm thick consisting mostly of aeolian sand covers this area.

Zone V is located at the south of the Ubhur creek as well as a belt about 0.5 km wide along the coast in northern Ubhur; in the southern part, the limestone is sometimes covered by a thin layer of carbonate gravelly sand, the larger thickness being in the areas away from the coast. Morris (1975) illustrated the occurrence of compact limestone which has an N value of 50 beneath the King Abdulaziz International Airport. This limestone has a thickness range between 2.5 and 10 m.

Zone VI contains a hill of gravelly sand or sandy gravel layers underlain by layers of red or chalky hard clays or mudstone or friable sandstone (unit 8) and capped with basaltic lava (unit 7).

Zone VII covers a small area within zone II as a layer of soft clay about 2.5 m thick similar to the soft clay of zone I, overlying medium dense carbonate gravelly sand. At one location in this zone, large gypsum crystals (probably part of a gypsum layer) were found at depths of 0.5 to 0.6 m, and this layer affected the geotechnical parameters.

Aldahri et al. (2017) measured the Vs30 at 76 sites within the study area using the multiple analysis of surface wave (MASW) technique and the N value obtained from 10 boreholes having a depth range of 10 to 15 m (Fig. 8). They classified the Ubhur district into three soil classes based on the NEHRB classification of B, C, and D as shown in Fig. 8. Site class B occupies a thin strip at the western part of the Ubhur district, while site class C occupies the central part of the study area. It occupies more than 90% of the surface area of Ubhur.

Fig. 8
figure 8

Map showing the soil classes and distribution of boreholes in the study area (Aldahri et al. 2017)

Full size image

Methodology of the H/V spectral ratio

Nakamura (1989) proposed the basis of qualitative arguments that the H/V ratio is a reliable technique of the site response to S-wave, providing reliable estimates, not only of the resonance frequency, but also of the corresponding amplification. These ratios are much more stable than the noise spectra and that, on soft soil sites, they display a clear peak that is well correlated with the fundamental resonance frequency.

The basic concept of Nakamura’s technique is described as follows. The transfer function (site effect) ST of the surface layers is generally defined as

$$ \mathrm{ST}=\mathrm{SHs}/\mathrm{SHB} $$
(1)

where SHs is the horizontal tremor spectrum on the surface, and SHB is the horizontal tremor spectrum incident from the substrate (rock or firm soil) to the surface layers. But the SHs of microtremors is readily affected by the surface waves. The effect of Rayleigh waves should be included in the vertical tremor spectrum SVs on the surface, but not included in the vertical tremor spectrum SVB in the base ground (rock). The effect of Rayleigh waves on the vertical tremor is

$$ \mathrm{ES}=\mathrm{SVs}/\mathrm{SVB} $$
(2)

If there is no Rayleigh waves, ES = 1, ES will take a value larger than 1 with increasing effect of the Rayleigh wave. The value of ST/ES may be considered to offer a more reliable transfer function STT after elimination of the effect of Rayleigh waves (Nakamura 1989; Lermo and Chaves-Garcia 1993).

$$ \mathrm{STT}=\mathrm{ST}/\mathrm{ES}=\mathrm{RS}/\mathrm{RB} $$
(3)

where RS = SHs / SVS and RB = SHB / SVB.

RS and RB were obtained by dividing the horizontal tremor spectrum by the vertical spectrum, corresponding to surface and bedrock tremors, respectively. In the firm substrate, propagation is even in all directions (Nakamura 1989; Huan 1997), RB ≈ 1.

$$ \mathrm{STT}=\mathrm{RS}=\mathrm{SHs}/\mathrm{SVS} $$
(4)

This means that the transfer function (H/V spectral ratio) of the surface layer as defined by this equation may be estimated from the tremor on the surface only.

Microtremor data acquisition

Microtremor data has been collected for 76 sites as shown in Fig. 8. The Vs30 of these were calculated by Aldahri et al. (2017). The data acquisition was conducted using Trillium Compact 120 s and Trillium 120 PA seismometers. These seismometers are manufactured by Nanometrics Company, and they are sufficient to acquire seismic data over a wide frequency range. The used seismometers are suitable for both portable and fixed operation conditions. These seismometers have a very low self-noise. The seismic data were digitized and stored using Taurus portable seismographs. Taurus is equipped with three 24-bit data channels. Time-series data are stored in Steim (1) format, and can be extracted into MiniSEED, Seisan, or ASCII format, and streamed in Nanometrics NP format. Taurus can be deployed to record continuous data for extended periods of time. Taurus can record and store the seismic data at different sampling rates; in the current study, Taurus was adjusted to record and store the seismic data at 100 samples per second. The data acquisition was done through the summer of 2013. As the temperature and density of traffic in Jeddah during the summer are high, data acquisition was performed during the night hours in order to overcome the effect of temperature and traffic on the records. During data acquisition, all the precautions mentioned by Duval et al. (1994), Nakamura (1996), Mucciarelli et al. (1997)), Mucciarelli (1998), Mohamed (2003), and SESAME (2004) were considered. The length of the record started from 90 min to more than 3 h in case of bad recording conditions. The recorded data was tested daily. Any error within the data was discovered, and the acquisition of that site was repeated (Fig. 9).

Fig. 9
figure 9

Selected sites for microtremor data acquisition

Full size image

Data processing

The H/V computations are carried out using Geopsy software which was designed under the SESAME European research project WP12 (2004). Short-duration instabilities of the signal are avoided during H/V processing through anti-trigger window selection to remove the transients. The used time window length is 40 s. These windows were automatically selected considering the short-term average (STA)/long-term average (LTA) between 0.2 and 2.5. The current analysis was carried out for the frequency range between 0.5 and 15 Hz.

The baseline effect for each microtremor record has been corrected using the running average technique. This is completed by sliding an average window of a given length over the data series. For each window, the baseline is calculated as the average value in the data window. Various numbers of windows with 40-s (4000 samples) duration were selected automatically. This time window is proven to be sufficiently long to provide stable results. The time series was tapered with a cosine taper, and an amplitude spectrum is computed for each component. The FFT spectra were smoothed using the methodology of Konno and Ohmachi (1998), paying attention not to destroy the important spectral peaks. The two horizontal components were merged using the squared average methodology. H/V for each time window is calculated, then an average H/V curve is calculated. The reliability of the obtained H/V curves and peaks was examined using the criteria of the SESAME project (2004).

Microtremor records acquired in urban areas usually suffer from machinery noise. This machinery noise usually appears on the H/V curve as local narrow peaks. Discrimination between the peaks of the machinery origin and the peaks from the natural origin is very important. The origin of the obtained H/V peaks was tested using the damping test. Figure 10 shows the sequence of data processing implemented in the current study.

Fig. 10
figure 10

Sequence of the H/V spectral ratio calculations and testing for the origin of the peak which appeared on the curve of site P2. a Microtremor record. b Window selection model. c Spectral curve. d H/V ratio curve. e Damping test of a peak of natural origin at 5.4 Hz

Full size image

Results

The dominant frequency in the study area ranges between 0.77 and 5.7 Hz as shown in Fig. 11. This frequency band matches the frequency of the engineering structures in the Ubhur area. The f0 distribution map shows that the high resonance frequencies concentrate at the east and northeastern parts of the study area. These high dominant frequencies decrease in the west and southwest directions, while the A0 distribution map shows that the area under investigation can amplify the seismic wave by a factor ranging between two to six times. The high amplification factors (A0 = 3.6 to 6.0) are concentrated in two parts of the study area. The first part occupies the area around Ubhur Creek (Sharm of Ubhur), and the second part occupies the northeastern part of the study area (Fig. 12). The middle part of the study area shows low amplification factors (A0 < 2.5), while the eastern and northwestern parts of the study area are characterized by 3.5 > A0 > 2.5 according to the F0 distribution map and A0 distribution map; the study area can be classified into three zones as the following:

  1. 1.

    Zone (1): this zone occupies the area around Ubhur Creek and the southeastern part of the study area. This zone is characterized by f0 ranging between 0.7 and 1.5 Hz. Concerning the distribution of A0 in this zone, we can observe that the area south to Ubhur Greek shows high amplification factors ranging between 3.6 and 6.0 times while the area to the north of the creek shows lower amplification factors ranging between 2.5 and 3.5. Soil class C is dominant at this zone with minor occurrence of some small localities of site class B. Figure 13 shows the microtremor analysis and results of site P17 as an example of a reliable H/V curve representing the characters of this zone.

  2. 2.

    Zone (2): this zone occupies the middle part of the study area. This zone is characterized by f0 ranging between 1.5 and 2.5 Hz. Concerning the distribution of A0 in this zone, we can observe that this zone shows amplification factors less than 2.4. Soil class C is dominant at this zone. Figure 14 shows an example of a reliable H/V curve representing the characters of this zone.

  3. 3.

    Zone (3): this zone occupies the northeastern part of the study area. This zone is characterized by f0 ranging between 2.5 and 5.7 Hz. Concerning the distribution of A0 in this zone, we can observe that this zone shows amplification factors ranging between 2.5 and 3.5. Soil classes C and D are dominating this zone. Figure 15 shows an example of a reliable H/V curve representing the characters of this zone (Fig. 16).

Fig. 11
figure 11

Distribution of the dominant frequency (f0) of Ubhur district

Full size image
Fig. 12
figure 12

Distribution of the soil maximum amplification factor (A0) of Ubhur district

Full size image
Fig. 13
figure 13

Sequence of the H/V spectral ratio calculations and testing for the origin of the peak which appeared on the curve of site P34. a Microtremor record. b Window selection model. c Spectral curve. d H/V ratio curve. e Damping test

Full size image
Fig. 14
figure 14

Sequence of the H/V spectral ratio calculations and testing for the origin of the peak which appeared on the curve for a site in the western part of the study area P76. a Microtremor record. b Window selection model. c Spectral curve. d H/V ratio curve. e Damping test

Full size image
Fig. 15
figure 15

Sequence of the H/V spectral ratio calculations and testing for the origin of the peak appeared on the curve of the site P34. a Microtremor record. b Window selection model. c Spectral curve. d H/V ratio curve. e Damping

Full size image
Fig. 16
figure 16

Soil thickness map of the study area (Aldahri et al. 2017)

Full size image

Discussion and conclusions

Al-Qarni (2015) studied the microzonation of the Jeddah urban area which is located at the south of our study area using both microtremor data and the multianalysis of surface waves. The author found that the resonance frequency ranges between 0.48 and 4.87 Hz and the maximum amplification factor ranges between 1 and 7. The resonance frequency range and maximum amplification of Jeddah coincide with the resonance frequency range and maximum amplification factor of Ubhur. We suggest that the soils of two regions have the same subsurface soil conditions and structures. Also, the f0 maps of the current have the same pattern as the maps produced by Al-Qarni (2015), where the low f0 values are concentrated on the west part (near the Red Sea Coast), and these values are increasing gradually to the east.

The A0 map showed a range of small values extending north-south, indicating the presence of the foundation layer near the Earth’s surface, while these values increased eastward and westward, indicating the presence of large thickness of soft sediments in the eastern part of the Ubhur area. In addition, BH 1 is located within the range between 2.4 and 2 of the amplification factor (these values are considered small) while it lies in the range from 2.1 to 2.5, with f0 considered to be relatively high from the engineering point of view. In addition, the southeastern part in the A0 map has the lowest values of the frequencies. This indicates that the foundation layer is deep while the surface soil is composed of soft deposits. In comparison with the geological map of Ubhur, the southeastern zone is composed of sabkha deposits. It is indicated that the central zone of the Ubhur district is considered suitable for the construction of high-rise facilities compared to other parts of the region, which should be taken into consideration during land use planning.