GIS based V s30 , V sz & D bed rock Mapping for the Shallow Sites of Islamabad, Pakistan

-- The location of Pakistan is in a collision zone among the Eurasian and Indian plate boundaries. The geological framework of the Northwest Himalaya makes Northern Pakistan susceptible to frequent moderate to major earthquakes. The devastating earthquake of October 8, 2005 caused 87,000 deaths, 2.8 million displacements and financial loss of 200 Billion USD equivalent to 6% GDP of Pakistan. Islamabad is in seismically active region, the studies conducted for this region addresses the poor soil conditions, non-engineering construction practices and high level of seismic shaking. However, none of these studies provided readily available ground maps (top 30m time average shear wave velocity V s30 , average soil shear wave velocity V sz & depth to bed rock D bed rock ) based on site specific geotechnical database. V s30 , V sz & D bed rock are important parameters for the evaluation of dynamic site characteristics of shallow bed rock sites. In Pakistan, Uniform Building Code (UBC-1997) is currently practiced, and the site classification are defined using site specific V s30 . In this study, geotechnical borehole database of 57 sites of Islamabad was collected to develop GIS based V s30 , V sz & D bed rock maps using Standard Penetration Test value (SPT-N). The borehole database includes, SPT-N, soil description and unit weight of soil for the Islamabad sites. In order to develop shear wave velocity (V s ) profiles, the available V s -SPT empirical correlations were evaluated where both SPT-N & V s measured data is available. The average of selected bounding correlations was applied to SPT-N database to develop representative Vs profiles. Using these V s profiles, V s30 , V sz & D bed rock GIS based maps were developed by applying Kriging interpolation in open source QGIS software. The proposed GIS maps can be used in preliminary earthquake design for seismic resilient earthquake design in Islamabad, Pakistan.


I. INTRODUCTION
In event of earthquake, the surface dynamic site characteristics are predominantly controlled by local geology.The local site effects at surface are evaluated by implying the site specific small strain property known as shear wave velocity (Vs) [1,2].In current geotechnical engineering practice, the top 30m time average shear wave velocity (Vs30) is used to define local site effects.Using Vs30 the site seismic coefficients (long and short period) can be calculated to define the design earthquake spectra at particular site.Based on calculated Vs30 site can be classified into five (05) classes ranging from A to F, Table 1.This top 30m is considered relatively shallow in engineering seismology, however it is typical boring depth around the world for geotechnical investigations.In case of site, where soil profile comprises of number of layers, VS30 can be calculated as, where di= thickness of each soil layer, VSi = shear wave velocity of each soil layer and n is number of layers.
The shear wave velocity profile is developed using various geophysical tests [3,4].However, in many regions the most common and popular test is SPT.Shear wave velocity profile is developed using SPT-Vs correlations [5][6][7][8][9].time average soil shear wave velocity (Vs, soil) and depth to bed rock (Dbed, rock) for the shallow bed rock sites [8,10].Islamabad is in seismically active region, the studies conducted for this region addresses the poor soil conditions, non-engineering construction practices and high level of seismic shaking and Vs,30 mapping [11][12][13].However, none of these studies provided readily available ground maps (average soil shear wave velocity Vsz & depth to bed rock Dbed rock), which are significant parameters in earthquake engineering practice for the evaluation of dynamic site characteristics of shallow bed rock sites.In this study, geotechnical borehole data of 57 sites of Islamabad were collected from different projects.The collected data includes, SPT-N, soil description and unit weight of soil.Vs-SPT correlations were selected and applied to develop Vs profiles.Using these Vs profiles, Vs30, Vsz & Dbed rock GIS based maps are prepared.The proposed GIS maps can be used in preliminary earthquake design for seismic resilient earthquake design in Islamabad, Pakistan.

A. STUDY AREA
The study area, Islamabad is located between Longitude  The collected geotechnical data include 57 sites SPT boreholes data and 04 sites shear wave velocity data as shown in Fig. 2. The SPT N data available with location of borehole, borehole depth, ground water location, soil stratigraphy was some time digitized for each depth interval from figures in the available geotechnical reports.In case the geographical coordinates of any site were unavailable in report, the site location name was used in Google Earth to extract the coordinates.
The digitized SPT-N values along with stratigraphy are converted to the Vs profile using SPT-Vs correlation.Currently, no SPT-Vs correlation is available for the geological conditions of Pakistan.However, numerous SPT-Vs correlations are available for the wide range of soils in literatures.Available correlations in literature were evaluated for clay [3,[15][16][17][18][19][20] and sandy gravel [16,21,22] soils by comparing with available measured Vs profiles.The selected correlations were applied to SPT-N data to develop Vs profiles.

Fig. 2 Distribution of borehole data (SPT & Vs) used in current study
The Vs30 was calculated from the developed Vs profiles.For all the sites the borehole does not extend to 30m depth, therefore, Boore [2] extrapolation model was used, which is mathematically expressed as,   Finally, all the calculated Vs30, Vsz & Dbed rock were interpolated for the location where no sample data was available by applying the kriging method in QGIS [23]. [33]

III. RESULTS AND DISCUSSION
The distribution of borehole data in terms of borehole depth is shown in Fig. 3.It is observed that 34% borehole consists of depth less than 10m, 38% ranges between 10-20m while 29% ranges between 20-24m.It can also be seen that none of the borehole in database exceed the 30m depth.This shallow depth of borehole reveals the shallow presence of stiff strata or bed rock.

Fig. 3 Distribution of borehole depths used in current study
In available borehole database, the SPT was performed at an interval of 1.5m.The value of SPT-N equal to 50 or 100 was considered as refusal and depth to hard strata or rock.The detailed analysis of borehole revealed presence of clay and gravel materials in boreholes.
The comparison of available correlations in literature for clay and sandy gravel soils are shown in Fig. 4. It can be seen that the measured Vs of clays lies within the Vs predicted by Ohta and Goto [16], Imai & Tonouchi Imai [17], Jinan [20] for geological age Holocene (H) & geological deposition Alluvium (A), the average of these bounding correlations is applied to SPT to develop Vs profiles.For the gravelly soils, the measured data points lie within the Andrus [22], Yoshida, et al. [21] and Ohta and Goto [16] correlations therefore, the average of these bounding correlations is applied to SPT to develop Vs profiles of gravelly soils.The proposed Vs30 map is shown in Fig. 5.It can be seen that the Vs30 ranges from 180~400m/s, predominantly Islamabad can be classified as site class D as per UBC & NEHRP [24,25] site classifications which are currently practiced in Pakistan SP-BCP [26].Site class D is considered stiff soil as per description in Table 1 [24][25][26].The proposed average soil Vsz map is shown in Fig. 6.Vsz ranges from 170-300 m/s.Also, the proposed map for the depth to bed rock is shown in Fig. 7.It can be observed that Islamabad has a shallow bed rock sites ranging from 4-24m.Manandhar, et al. [27] has shown that Vs30 does not fit well in capturing the site amplification factors for shallow site, and proposed a new site amplification system based on Vsz & Dbed rock for shallow bed rock sites.Fig. 6 & 7 clearly shows that there is a need to develop new classification system that is a function of soil depth average shear wave velocity, Vsz and depth to bed rock, Dbed rock.
[72°40′ & 73°28′ E] and Latitude [33°29′ & 33°52′ N] in Margalla ranges.These Margalla ranges are also consider as Part of Pakistan's Lesser Himalayas.Also, Islamabad is located in a seismically active zone of Main Boundary Thrust (MBT) and Panjal Thrust (PT) in north and Salt Range Thrust (SRT) to the south.Due to this tectonic configuration, folding, faulting and earthquakes has higher frequency.Area drains into southwards in River Soan by tributaries of Bedar Wali kas, Lei nullah and Jodh nullah.

Fig. 1
Fig. 1 Geologic & Seismic configuration of study area [14] II.METHODOLOGY The general methodology adopted for the development of GIS based maps includes following steps, 1. Collection & Digitization of borehole database.2. Evaluation & application of SPT-Vs correlations.3. Development of typical shear wave velocity profiles.4. Calculation of Vs30, Vsz & Dbed rock for each profile. 5. Development of GIS based Vs30, Vsz & Dbed rock The collected geotechnical data include 57 sites SPT total depth of borehole/velocity profile VSeff = the last value of Vs in profile andThe value of SPT-N equal to 50 or 100 was considered as refusal and depth to hard strata or rock, Dbed rock.The Vsz represents the average soil shear wave velocity of the soil deposit.

Fig. 4
Fig. 4 Selection of SPT-Vs correlation for clay and sandy gravel soils using measured Vs data points

Fig. 5
Fig. 5 Proposed Vs30 Maps for the Islamabad

Fig. 6
Fig. 6 Proposed Vsz Maps for the Islamabad

GIS based V s30 , V sz & D bed rock Mapping for the Shallow Sites of Islamabad, Pakistan Class General Description & Vs Criteria
Muhammad Shahkar a , Muhammad Usman Arshad b , Shamsher Sadiq c, * , Abdul Waheed d a Graduate Student, Department of Civil Engineering, UET Taxila, Pakistan b Assistant Professor, Department of Civil Engineering, UET Taxila, Pakistan c Assistant Professor, Department of Civil Engineering, GIK Institute, Topi, Pakistan TABLE I SITE CLASSIFICATION CRITERIA Vs30 is widely practice for the deep sites, however, most recently a new site classification system is proposed based on