Fig. 1. Geological map of the region under study (simplified after Bellizzia et al., 1976), location of the El Pilar fault after Audemard et al. (2000). The star indicates the location of the epicenter after FUNVISIS et al. (1997), Mm=Mesozoic metamorphic rocks, Cl=Cretaceous limestones, Ts=Tertiary sediment, Qp=Quaternary sediments (Pleistocene), Qr=Quaternary sediments (recent).

Fig. 2. The regional distribution of macroseismicity related to the 1997 Cariaco earthquake with the estimation of the MMI intensities (above) and zoom on the area related to the surface rupture (thick line; below). Observed intensities at individual data points (see Appendix A) are symbolized by rhombs: MMI=4; squares: MMI=5; circles: MMI=6; triangles: MMI=7 and stars: MMI=8. The dark star indicates the location of the epicenter.

Fig. 3. Decay of macroseismic intensities in E–W direction (solid line) and N–S direction (dashed line) from epicenter. The towns of Cariaco and Cumaná show average intensities of VIII and VI, respectively.

Fig. 4. Fourier spectra with the three components (above) and H/V values (below) for two sites in (a) Cariaco (c07) and (b) in the Cretaceous bedrock south of Cariaco (cc2). The site within the town (c07; for location see Fig. 5) shows a peak at 0.8 s and the bedrock site (cc2) at 0.3 s, considered as predominant periods.

Fig. 5. Map of Cariaco with the predominant periods from microtremor measurements. The circles indicate the soil sites where measurements were performed (c07 from Fig. 4 is located northwest of line 7). The bedrock site (cc2 from Fig. 4) is located about 500 m south of the surface rupture, which separates soft soil to the north from more consolidated sediments to the south. The location of the seismic refraction lines (black lines with arrows indicating the direction of observation) as well as the composed profiles A and B (dashed lines) are indicated. 1=Raimundo Martínez Centeno high school; 2=Valentín Valiente school; 3=Bank of Orinoco; 4=Brekerman Street; 5=Las Flores Street; 6=Bermúdez Street.

Fig. 6. Time–distance plots (a) and seismic sections (b–c) for line 2 at the northwestern border of Cariaco (for location see Fig. 5). As an example for data coverage and the control of the reciprocal times at each shot point, the picks of the first arrivals from and S-waves (above) and P-waves (below) are displayed as time–distance plots (a). Shot-point locations 480 m (b) and 1080 m (c) with the observed seismograms with picked arrivals (above), the observed (circles) and calculated (lines) travel times (center) and the calculated raypaths with the S-velocity model (below). Raypaths with velocity model and calculated P-wave travel times of shot point 480 m are displayed as well (b, lower part). First breaks of S-waves are partially overlain by signals of air blast (c).

Fig. 7. Seismic section along line 3 at the southern limit of Cariaco of shot-point location 480 m with the observed seismograms with picked arrivals (above), the observed (circles) and calculated (lines) travel times (center) and the calculated raypaths with the S-velocity model (below) and the P-velocity model (lower part). No disturbance by signals of air blast is observed along this profile.

Fig. 8. Seismic sections along line 1 at the eastern limit of Cariaco of shot-point locations 0 m (a) and 510/580 m (b) with the observed seismograms with picked arrivals (above), the observed (circles) and calculated (lines) travel times (center) and the calculated raypaths with the S-velocity model (below). Raypaths with velocity model and calculated P-wave travel times of shot point 0 m are displayed as well (a, lower part). A deeper high velocity layer, which is inclined towards the north, is observed in the S-waves (700 m/s) as well as in the P-waves (2500 m/s). The geometry of the top of the deeper layer is well confirmed by the two shot points at 510 and 580 m (b).

Fig. 9. Composed profiles A (top) and B (bottom) located at the southeastern border and in the northwestern part of Cariaco, respectively. The numbers represent the S-wave velocities for each layer. Solid lines: measured, dashed lines: interpolation. The type profiles indicate the locations for which response spectra (Fig. 10) were calculated. For location, see Fig. 5; vertical exaggeration for profile A, 3:1; for profile B, 6:1.

Fig. 10. Response spectra (damping factor 5%) for three type-profiles of Cariaco: type-profiles 1 and 2 correspond to the representative structure of profiles A and B (Fig. 9), while the type-profile 3 is located at the northeastern end of profile A, where the thickness of the Quaternary sediments decreases (for details see Table 2). The thin and thick lines correspond to the response on the bottom and on the top of the sediments, respectively. The accelerogram used is a horizontal component (recorded on stiff sediment at 10.6 km distance to fault rupture) of the 1979 (Ms 6.9) Imperial Valley earthquake at an azimuth of 140°, which had a rupture mechanism similar to the 1997 Cariaco earthquake.

Fig. 11. Induced effects by the Cariaco earthquake around the Cariaco Gulf (A), and around its eastward continuation, the Cariaco sedimentary basin (B). T=Tocuchare, EH=Ensenada Honda, C=road 9 (north of Cariaco), LC=Calzadilla beach, ACA=Aquacam C.A., AM=Atún Margarita fish cannery, H=Veteran Hospital, Cumaná. Qal=Quaternary alluvial sediments, Ms=Mesozoic sedimentary rocks, Mm=Mesozoic metamorphic units.

Fig. 12. Induced effects by the Cariaco earthquake along the shoreline in the city of Cumaná: (A) in the harbor district to the west and (B) the El Peñon area to the east; arrows indicate the direction of sliding. Numbers coincide with the description of the locations in the text.

Plate 1. (A) Aligned sand blows vented through an open-fracture related to sea-front relaxation on the Punta Baja beach, east of Cumaná and NW of El Peñon (for location, refer to Fig. 11C); (B) detail of one of those aligned sand blows; (C) bird's-eye view of northward lateral spreading of almost flat-lying ground of a coconut plantation at the Piragua pool, on the southeastern edge of the Buena Vista swamp, next to the surface rupture of the Cariaco 1997 earthquake (arrow indicates the direction of sliding); and (D) detail view to the east of previous locality exhibiting clockwisely rotated slabs over a shallow liquefied layer. Also notice water shallowness in the southernmost crack.

Fig. 13. Map of Cariaco with the location of geotechnical drillings and their respective soil type percentages (for details of sand layers see Table 3). Surface evidence for liquation has only been observed between Brekerman, Las Flores and Bermúdez streets in the south of the town. The location of the geotechnical profiles 1, running from drilling P2 in the south to drilling P3 in the northwest and profile 4, running from drilling P1 in the center to drilling P6 in the southeast (Fig. 14) is indicated.

Plate 2. (A) Bird's-eye view of sand blowing at the Marigüitar delta, next to the Maigualida beach on the southern coast of the Cariaco Gulf (for location, refer to Fig. 11); (B) detail of one of those isolated sand blows measuring between 4 to 5 m across; (C) collapse of rock embankment at the tip of the fishery port located on the western coast of Cumaná and near the Manzanares river mouth; and (D) sliding of hydraulic fill edge and rails at a dry dock induced by submarine slumping at the Manzanares river mouth (for location refer to Fig. 12).

Plate 3. (A) Small lateral spreading in few-meter-high Plio-Pleistocene cliffs on the northern coast of the Cariaco Gulf, nearby Chiguana (for location, refer to Fig. 11); (B) lateral spread affecting small resting area sitting on artificial fill, at the Calzadilla beach, on the southern coast of the Cariaco Gulf; (C) bird's-eye view of seaward lateral spreading of the southernmost earth embankment of the Aquacam C.A. shrimp farm, east of Chiguana; and (D) sliding of travertine pools originally capping the El Pilar fault scarp developed on Mesozoic mudstones at the Aguas Calientes farm. To the south, some unbroken pools can be spotted, whereas travertine blocks are piling up to the north (left). Arrows indicate the direction of sliding.

Fig. 14. Profiles summarizing the geotechnical information in Cariaco crossing drillings P3, P1 and P2 from NW to SE (profile 1, top) and P1, P4 and P6 from NW to SE (profile 4, bottom). The number of SPT counts are indicated for each drilling; CL=clay, SM=silty sand, SP=poorly sorted sand, SG=sand with gravel.

Table 1. P-wave and S-wave velocities for the seismic lines in Cariaco (for location see Fig. 5)

Generally, two strata are displayed for P-wave velocities, the unsaturated sands and clays (400–1040 m/s) in the upper 4–14 m and the water-saturated sediments (1700–1900 m/s) below. The groundwater level (top of the second layer) is quite deep compared to the observations right after the earthquake (see Section 4.2), which is attributed to the dry weather conditions during seismic measurements. From the evaluation of P-velocities, deeper layers are only detected on lines 3 and 1 with velocities of 2250 and 2500 m/s, respectively, the latter one interpreted as base of the Quaternary sediments. The distribution of S-wave velocities is displayed in Fig. 9.

Table 2. Input data of the Proshake 1.1 program (based on Schnabel et al., 1972) for the dynamic response evaluation of type-profiles 1, 2 and 3 (Fig. 10)

Input data consider the geotechnical characteristics (Section 4.2) as well as S-wave velocities (Fig. 9). The shear module Gmax was calculated based on the curves of Vucetic and Dobry (1991). Modulus and damping curves for clay after Sun et al. (1988) and for sand after Seed and Idriss (1970).

Table 3. Borehole profiles in Cariaco indicating the most important geotechnical parameters of the sand layers with respect to their liquefaction potential (for location see Fig. 13)

Surficial evidence for liquefaction was observed only at drilling P2 in the Brekerman Street. The level of seismic acceleration considered for the estimation of the liquefaction potential is 0.3 g. NSPT is normalized to 60% of blow counts from SPT tests.