5. Conclusions

Shear wave velocity VS profiles were determined by performing nine microtremors array surveys in Port of Spain (POS), Trinidad, employing the spatial autocorrelation SPAC method and genetic algorithms (GAs); the results yielded VS between 50 and 2000 m/s at POS. The ellipticity pattern for the first mode of Rayleigh waves explains the resulting predominant peak in the H/V Nakamura ratios for all array sites. We validated the soil profiles retrieved by the SPAC and GAs'schemes comparing the synthetics' horizontal-to-vertical spectral (H/V) ratios

Lots, the depth of sediments yields 80 m. Generally, the VS in the sediments increases with depth in the range mainly of 50 m/s to 600 m/s, and the variants of the stiffness in the soil are mainly found near the surface due to the reclaimed land —compacted or not—during construction works or the presence of swampy soil to the East of POS. We have also corroborated via successive rounds of genetic inver-

Estimation of Shear Wave Velocity Profiles Employing Genetic Algorithms and the Diffuse Field…

period is a superior parameter for classification than the VS30 in this case.

Borehole data in POS reaching the bedrock is very limited, and a parallel research would be focused on conducting boreholes reaching the depth of the bedrock at new strong motion stations or critical facilities, and if possible, to get the VS employing alternative methods (e.g., cross hole or laboratory soil test); borehole data would help to validate the proposed preliminary liquefaction hazard map and can be used to implement remedial measures against such hazard, especially at the coastal port area. The last hazard peer-reviewed maps for Trinidad and the Eastern Caribbean have been proposed by Bozzoni et al. [35] yielding 0.60 g of peak ground acceleration for POS setting 2475 years return period at a rock site class B in NEHRP classification; such shaking level is strong enough to trigger the liquefaction in the saturated alluvium at POS. Since our microtremor survey only permits to study the soil behavior in the linear range, the effect of the non-linearity in the soil is still a big question to solve for the area. Future research lines might be focused on a frequency-dependent

The authors thank IntechOpen for inviting us to write a chapter of this book and editor John Tiefenbacher for his sincere and fruitful comments during the revision

The microtremor recordings were provided by the Ministry of Planning and

We used QGIS Open Software version 2.4 and the GCC Fortran Compiler to generate the maps presented in this work. The VS30 values used in this work were retrieved via WEB server at USGS (earthquake.usgs.gov/hazards/apps/vs30/) last

The authors thank the Ministry of Planning and Sustainable Development of Trinidad and Tobago for funding this research being conceived by the first author as a principal investigator while working at the Seismic Research Centre in Trinidad

Sustainable Development of Trinidad and Tobago and are proprietary.

It is worth mentioning that the DFA even reproduced the whole shape of the H/V ratios, including peaks and troughs and the level of amplification; such characteristics cannot be retrieved employing the Rayleigh wave pattern interpretation. It seems that the H/V spectral ratio technique from Nakamura represents a true piece of information regarding the dynamic properties of the sediments above bedrock in terms of identification of the fundamental period of the soil profile when comparing with the one retrieved with the 1-D SH wave theoretical amplification as well; it is noticed that an amplification level yields between three and five at the fundamental period. However, the H/V spectral ratios do not coincide with the whole shape of the 1-D SH transfer function. If we proved just incorporating the fundamental mode in the phase velocity inversion that our resonant peak in the H/V ratios is genuine, then we did not incorporate higher modes in the inversion; however, future research lines on this topic would allow us to introduce the first overtones in new analysis [34]. The VS30 retrieved from our microtremors array coincides well with the predicted VS30 of Matsuoka et al. [27] when incorporating geomorphological conditions; however, we did not find a good correlation with NEHRP classification on VS30 [29] suggesting that the fundamental

sions that the VS at the bedrock yields 2000 m/s.

DOI: http://dx.doi.org/10.5772/intechopen.85129

quality factor as well.

Acknowledgements

accessed January 9, 2016.

process.

75

#### Figure 14.

Leaning lighthouse tower at the Port Area. See the displacement Δ at the top of the structure due to the soil subsidence at its foundation.

generated by the Diffuse Field Approach (DFA) with the observed ones at the array sites and with empirical formulas to estimate the average shear wave velocity of the upper 30 m (VS30). We conclude that the H/V ratios yield a genuine shear wave fundamental period of vibration of the soil profiles at POS, and that can be used to validate the high-resolution seismic microzonation map proposed by Salazar et al. [1]. The amplification and fundamental period of motion retrieved from the microtremors together with the water table level suggest a high liquefaction potential mainly on the coastal areas. It looks that in terms of seismic amplification and liquefaction hazard, a safe place is the Laventille area at the East characterized geologically by a metalimestone; unfortunately it is classified with the highest crime rate and drug dealers in Trinidad.

The genetic inversion results revealed that the deeper parts of POS are located in the Port Area and South of Woodbrook with a depth of 225 m and the softer materials are located at the South East of POS in Sea Lots with low VS of 50 m/s which correspond to a buried swamp or mangrove; toward the north of the City, the depth of the sediments decreases substantially from 75 m in the Savannah to 30 m in Saint James and Cocorite to the West. Toward the South East part of POS in Sea

#### Estimation of Shear Wave Velocity Profiles Employing Genetic Algorithms and the Diffuse Field… DOI: http://dx.doi.org/10.5772/intechopen.85129

Lots, the depth of sediments yields 80 m. Generally, the VS in the sediments increases with depth in the range mainly of 50 m/s to 600 m/s, and the variants of the stiffness in the soil are mainly found near the surface due to the reclaimed land —compacted or not—during construction works or the presence of swampy soil to the East of POS. We have also corroborated via successive rounds of genetic inversions that the VS at the bedrock yields 2000 m/s.

It is worth mentioning that the DFA even reproduced the whole shape of the H/V ratios, including peaks and troughs and the level of amplification; such characteristics cannot be retrieved employing the Rayleigh wave pattern interpretation. It seems that the H/V spectral ratio technique from Nakamura represents a true piece of information regarding the dynamic properties of the sediments above bedrock in terms of identification of the fundamental period of the soil profile when comparing with the one retrieved with the 1-D SH wave theoretical amplification as well; it is noticed that an amplification level yields between three and five at the fundamental period. However, the H/V spectral ratios do not coincide with the whole shape of the 1-D SH transfer function. If we proved just incorporating the fundamental mode in the phase velocity inversion that our resonant peak in the H/V ratios is genuine, then we did not incorporate higher modes in the inversion; however, future research lines on this topic would allow us to introduce the first overtones in new analysis [34]. The VS30 retrieved from our microtremors array coincides well with the predicted VS30 of Matsuoka et al. [27] when incorporating geomorphological conditions; however, we did not find a good correlation with NEHRP classification on VS30 [29] suggesting that the fundamental period is a superior parameter for classification than the VS30 in this case.

Borehole data in POS reaching the bedrock is very limited, and a parallel research would be focused on conducting boreholes reaching the depth of the bedrock at new strong motion stations or critical facilities, and if possible, to get the VS employing alternative methods (e.g., cross hole or laboratory soil test); borehole data would help to validate the proposed preliminary liquefaction hazard map and can be used to implement remedial measures against such hazard, especially at the coastal port area. The last hazard peer-reviewed maps for Trinidad and the Eastern Caribbean have been proposed by Bozzoni et al. [35] yielding 0.60 g of peak ground acceleration for POS setting 2475 years return period at a rock site class B in NEHRP classification; such shaking level is strong enough to trigger the liquefaction in the saturated alluvium at POS.

Since our microtremor survey only permits to study the soil behavior in the linear range, the effect of the non-linearity in the soil is still a big question to solve for the area. Future research lines might be focused on a frequency-dependent quality factor as well.
