The data collected were analysed to obtain the estimated shear wave velocity profile. The H/V ellipticity of Rayleigh Wave method is used for single point measurement analysis, and the SPAC method is adopted for array measurement. The shear wave velocity (Vs) profile was altered until the ellipticity curve matched well with the H/V spectral ratio and ellipticity of Rayleigh Wave method. In the SPAC method, the proposed shear wave velocity (Vs) profile was altered until the dispersion curve matched well with the measured dispersion of the surface wave.
3.1.1 Single Point Measurement Analysis
The H/V ellipticity of Rayleigh Wave method is adopted. The estimated Vs profiles are proposed for each site to generate an ellipticity curve. Similar to the SPAC method, many attempts were exerted by altering each layer’s thickness and shear wave velocity until the ellipticity curve matched well with the H/V spectral ratio. Amongst all the results, the comparison of the ellipticity curve and H/V spectral ratio at 21 sites were selected for discussion. The selected sites, which cover the whole Penang Island, are shown in Fig. 7. The estimated shear wave velocity profile of these selected sited are presented in Fig. 8(a). The ellipticity curve developed from the proposed ground structures at all three sites correlate well with the H/V spectral ratio of the measured data, as shown in Fig. 8(b). Hence, all the proposed shear wave velocity profiles are accepted.
The sites with higher frequency, such as site D047, E007 and G015, have higher Vs,30 value, indicating that the sites are stiffer. This result correlates well with the topography of Penang Island, where the sites with higher frequency are located near a hilly area. Meanwhile, site F045 shows a low frequency (< 2 Hz) due to the thick soft sediment underlain. Meanwhile, the rest of the results show average frequency, ranging from 2 Hz to 4 Hz. According to the analysis, the areas at Georgetown, Batu Maung and Balik Pulau tend to show a thicker soft sediment with a lower value in frequency and Vs,30. This condition is in agreement with the geological map shown in Fig. 2, indicating that these areas mainly consist of unconsolidated marine clay, sand and gravel.
The estimated Vs profiles of other sites are obtained by repeating the same process. The results are then used to plot the microzonation map and establish the correlation between predominant frequency (f) and sediment thickness (h).
3.1.2 Array Measurement Analysis
Microtremor array measurements were conducted at four different locations, namely, TT, SQ, LIB and JCH, with varying array sizes, as summarised in Table 3. In the SPAC method analysis, the correlation between microtremor sensors with different distances will generate two dispersion curves with an identical trend but different values. SPAC (outer to outer) shows a correlation between the microtremor of the center sensor and the outer sensors of the array measurement, as indicated in Figs. 9(a) to 9(d). Meanwhile, the correlation between the outer microtremor sensors of the array was indicated as SPAC (outer to outer). After the analysis, the estimated Vs profile of the sites are summarised in Table 4.
Table 4
Estimated shear wave velocity profile for array measurement sites
Location Code
|
Layer
|
Thickness (m)
|
Vs (m/s)
|
Frequency
(Hz)
|
Vs,30 (m/s)
|
TT
|
Soft
|
10
|
170
|
|
|
Medium
|
24
|
340
|
|
|
Hard
|
18
|
650
|
2.04
|
255.0
|
Engineering Bedrock
|
10
|
800
|
|
|
SQ
|
Soft
|
12
|
140
|
|
|
Medium
|
33
|
310
|
|
|
Hard
|
20
|
550
|
1.56
|
208.7
|
Engineering Bedrock
|
10
|
1000
|
|
|
LIB
|
Soft
|
3
|
160
|
|
|
Medium
|
14
|
290
|
|
|
Hard
|
20
|
550
|
3.95
|
330.9
|
Engineering Bedrock
|
20
|
800
|
|
|
JCH
|
Soft
|
1.5
|
170
|
|
|
Medium
|
15
|
250
|
|
|
Hard
|
30
|
550
|
3.40
|
321.31
|
Engineering Bedrock
|
10
|
1000
|
|
|
3.1.3 Comparison of Array Measurement Result with Available Borelog Data
To validate the accuracy of the proposed shear wave velocity profile, the Vs profile was cross-checked with the Standard Penetration Test N-value (NSPT) profile from the available site investigation (SI) report near the measurement points. The equivalent Vs and NSPT profiles from the available SI report near the array measurement points were compared, as shown in Fig. 10. The left figure is the NSPT profile taken from the SI report, whilst the right figure is the estimated Vs profile based on the soil properties in Table 5. The Vs profiles consist of four layers, which are soft, medium, hard and engineering bedrock with the Vs range of 0–180 m/s, 180 to 360 m/s, 360 to 800 m/s and more than 800 m/s, respectively.
Table 5
Vs range and its corresponding NSPT value
Layer
|
Description
|
Parameters
|
Vs (m/s)
|
NSPT (blows/ 300mm)
|
1
|
Soft
|
< 180
|
< 15
|
2
|
Medium
|
180–360
|
15–50
|
3
|
Hard
|
360–800
|
> 50
|
4
|
Engineering bedrock
|
> 800
|
-
|
The figures depict that the estimated ground structures of all sites exhibit a similar trend with NSPT profile, especially the ground structure of the top 15 m. The top soil layer is softer and gradually becomes stiffer with the increase in depth. This phenomenon correlates well with the assumption of this research, assuming that the ground structure consists of soft soil at the surface, followed by medium, hard and eventually engineering bedrock. According to the SI report, the engineering bedrock cannot be identified in the borelog because the maximum NSPT value is 50, which is considered soil class C or hard layer. By contrast, subsurface exploration using microtremor measurement can estimate the ground structure deep down to engineering bedrock, equivalent to Vs more than 800 m/s.
Considering that the medium and small array sizes were conducted at site TT, a better precision close to the actual soil profile can be obtained. Accordingly, the estimated Vs profiles of sites SQ and LIB show similarity with those obtained from borelog with some minor deviations. The NSPT and estimated Vs profiles in site JCH show some deviations after 15 m. However, the ground structure at top 15 m is similar with the NSPT profile. A larger array size should be adopted to estimate a deeper ground structure because it could provide a better estimation of the Vs profile at the deeper layer.
3.1.4 Cross-Section of Penang Island
Information about the subsoil geological structure is derived from the subsurface structure estimation results. The simplified sketches of geological sections across northern, eastern and western parts of Penang Island are drawn for the ease of showing the subsurface structure of Penang Island. The locations of the cross-sections are summarised in Fig. 11(a). Connecting as many measuring sites and available borelog as possible is the main consideration for selecting the cross-section location. Considering more measuring sites and borelog data could provide better precision of the estimated subsurface structure. In the cross-sections shown in Figs. 11(b) to 11(f), the soil layers are indicated as soft, medium and hard with different colours.
3.2 Seismic Microzonation Maps
3.2.1Predominant Frequency Map and Vs,30 Map
After obtaining the shear wave velocity profile of all measurement sites, the results are plotted in the microzonation map, representing the predominant frequency (f) and shear wave velocity of the top 30 m (Vs,30) of the research area. Figures 12(a) and 12(b) present the microzonation maps of predominant frequency and Vs,30, respectively. The inaccessible areas, especially the hilly ones, are indicated as dark green colour in the maps.
Based on the predominant frequency map, the western region and northeast of Penang Island have lower frequencies, ranging from 1.45 Hz to 2.5 Hz. This condition corresponds to the thick layers of unconsolidated marine clay deposits in these areas. Conversely, the northern and southern parts of Penang Island show higher frequency, indicating that they are underlain by more shallow bedrock or lesser thickness of soft sediment. Higher frequencies are generally found in the regions adjacent to hilly areas. This situation shows a good correlation with the geological map of Penang Island. These mountainous regions are associated with thin layers of soil deposits over the bedrock.
Penang Island has a ground structure with Vs,30 varying from 168 m/s to 468 m/s. The contour map depicts that the western and eastern parts of Penang Island have lower Vs,30, which corresponds well with the predominant frequency map shown in Fig. 12(a). The data show a similar trend to the predominant frequency map, where the northern and southern regions tend to show stiffer ground conditions, where the Vs,30 is higher. According to the analysis, the lowest frequency is found in the middle of the northeast region (Georgetown). The region has been found to have a softer ground condition, which corresponds with the low predominant frequency.