High-Resolution P- and S-wave Seismic Reflection Investigation of a Shallow Stratigraphic Sequence

 

Erich D. Guy

U.S. Army Corps of Engineers, Huntington, West Virginia, USA
Email: Erich.D.Guy@usace.army.mil

ABSTRACT

The effectiveness of P- and S-wave reflection surveys for mapping a shallow stratigraphic sequence (flat-lying unsaturated and saturated overburden materials above consolidated units) was evaluated through the acquisition and analysis of high-resolution multicomponent data. The combined P- and S-wave common-mode reflection information allowed the near-surface sequence to be imaged more effectively than using solely the P- or S-wave information. S-wave reflections from the bedrock and overburden interface were consistently measured in both the XX component (inline-inline, or SV-SV) and the YY component (crossline-crossline, or SH-SH) field data. However, surface wave noise resulted in the optimum reflection window of XX component data being relatively narrow, and stacked YY component data had a higher signal-to-noise ratio and better imaged the top-of-bedrock. P-wave reflections from the unsaturated and saturated overburden interface were recorded in ZZ component (vertical-vertical, or P-P) field data, but S-wave reflections from this interface were not observed. P-wave events from deeper contrasts in impedance could not be resolved in field data due to surface wave and air wave noise, a high P-wave reflection coefficient at the top of the saturated overburden, low P-wave reflection coefficients at deeper interfaces, and interference effects and poor resolution. Calculations based on P- and S-wave velocities (VP and VS) and dominant wavelengths suggest that the vertical resolution of S-waves in the study area dry overburden was more than 1.7 times the resolution of P-waves, while the resolution of S-waves in the saturated overburden was more than 4.7 times that of P-waves. The potential for determining detailed variations in Poisson’s ratio (n) using VP/VS ratios was found to be limited due to the small number of reflection events and the fact that P- and S-wave reflections did not correlate to similar interfaces; however, representative lithology values of n were able to be estimated using measurements and reasonable assumptions. Although P-wave reflection data have traditionally been acquired during shallow reflection surveys, results of this study demonstrate that due to differences in P- and S-wave propagation, media compressional and shear impedance contrasts, and variations in receiver sensitivity (as a function of orientation), it is necessary to consider the probable usefulness of different data components/wave-type reflections prior to conducting a shallow reflection survey. Near-surface imaging and characterization may be best accomplished through the acquisition and analysis of one particular, or multiple data components/wave-type reflections.

Keywords: Seismic Reflection, Multicomponent, Imaging, Characterization, P-waves, S-waves

INTRODUCTION

Seismic reflection (predominately P-wave methods) has been applied during the past several decades towards the solution of shallow earth (0 - 100 m deep) engineering and environmental problems (Hunter et al., 1984; Miller and Steeples, 1990; Bachrach and Nur, 1998; van der Veen and Green, 1998). Successful applications of shallow S-wave reflection have been demonstrated (Carr et al., 1998; Harris et al., 2000; Guy et al. 2003a); however, the number of shallow reflection reports concerning S-waves is quite small relative to those concerning P-waves, and most S-wave studies have considered only a single S-wave data component. Potential benefits of testing, recording and analyzing P- and S-wave reflections concurrently for improving shallow earth characterization potential have not been fully recognized or demonstrated. Few reports concerning the concurrent use of P- and S-wave reflections exist, but have shown promising results. Hasbrouck (1990) (without published data) and Goforth and Hayward (1992) discussed sites where saturated overburden and a target bedrock horizon had nearly the same P-wave velocity; P-waves imaged the water table at both sites but were unable to image bedrock. Clark et al. (1994) modeled a shallow geologic sequence and showed that only a combined usage of P- and S-waves would allow the detection of all modeled boundaries.

Potential benefits for shallow earth studies that may be obtained through concurrent P- and S-wave data testing and analysis remain to be recognized and explained, and demonstrated using field data acquired in different geologic settings. Because P- and S-waves travel at different speeds and respond to changes in elastic moduli and density differently, achievable resolution may be different for P- and S-waves, P- and S-wave reflection coefficients may be different, and P- and S-wave reflections may be recorded under different (more or less favorable) noise conditions. It is also possible in theory to estimate insitu elastic properties of near-surface materials using common-mode P- and S-wave reflection information. This study was conducted to investigate these issues, with the objective of providing results and discussion that will improve near-surface imaging and characterization understanding and ability. High-resolution nine-component (9C) seismic reflection data were acquired over a shallow stratigraphic sequence type that is commonly encountered during engineering studies. The acquired common-mode P- and S-wave reflection components were analyzed (as were other data components), and advantages and disadvantages associated with each common-mode component were determined, and are detailed in this paper in terms of near-surface imaging and characterization potential.

MULTICOMPONENT REFLECTION DATA

High-resolution 9C seismic reflection data were acquired along a line (named line EBTravel) located to the immediate south of Interstate 70 in Guernsey County, Ohio (Figure 1, Table 1). The reflection line was located over a geologic sequence consisting of Paleozoic sedimentary rocks with unconsolidated overburden materials containing a water table. The upper 5 to 15 ft beneath the ground surface consists of silt and clay fill materials, with silts, clays, and interbedded sand lenses down to bedrock. The total unconsolidated materials thickness above bedrock in the study area ranges from 40 to 45 ft. Bedrock is the Lower Mahoning Sandstone and Shale; it is arenaceous shale that lies above the Upper Freeport (bituminous) Coal. The coal seam ranges from 5 to 6 ft thick, is underlain by claystone, and was previously mined to the west of the study area. Hydrologic well data indicate that static water levels in overburden materials range from 26-30 feet of the ground surface. Further discussion regarding the area’s stratigraphy is presented in Conroy and Guy (2005).

 


Figure 1. Photographs taken during multicomponent seismic reflection data acquisition: multi-configuration-seismic source (left), and 3 component receivers (right). Refer to Table 1 for data acquisition geometry and recording parameters information.

 

Table 1. Field acquisition and recording parameters for multicomponent seismic reflection data.
Description Parameters
Spread typeInline CDP split-spread
Energy sourceIVI “Minivib II” buggy, capable of generating preferential shear particle motion inline and transverse to the line, and preferential compressional particle motion
Source configurationsInline shear, transverse shear, and compressional (vertical) configurations: 3, 3-component (3-C) records (i.e. 9C data) were acquired for each source station
Source intervalInline and transverse shear configurations: 1 ft, compressional (vertical configuration): 2 ft
Source offsetSource baseplate offset 6 ft from lines on soil (average offset)
Sweep typeLinear (start taper = 0.1 sec, end taper = 0.1 sec)
Sweep frequencies50-500 Hz
Sweep, record lengths4 sec, 1 sec (5 sec listen time minus 4 sec sweep)
Recording instrumentGeometrics 48-channel StrataView and StataVisor modules connected in series, 24 bit A/D resolution
Recording channels240 total, 80, 3-component geophones deployed for each shot (3 channels used per phone, channel 1 of 240 used to record pilot)
Data formatRecorded in SEG-2 format, converted to SEG-Y format
Field filter and gainNo field filters applied, pre-amplifier gain applied as a function of absolute offset (channels 0-30 = 0 dB, channels 30-32 = 24 dB, channels 32 - 240 = 36 dB, channel 240 = 48 dB)
Sample interval0.25 ms
GeophonesGeospace model GS-20DX (10 Hz), one 3-component geophone with orthogonal elements planted at each station
Geophone interval2 ft
Geophone locationsRoad stations 48900-48600, when the source reached the last geophone the first 16 phones were leapfrogged to the line end

Nomenclature used to describe acquired multicomponent data, in terms of source and receiver orientations and preferential polarizations, is illustrated in Figure 2. Three seismic source orientations were used that generated preferential shear particle motion in the direction of the line (sometimes called inline, or SV), transverse to the line (sometimes called crossline, or SH), and preferential compressional particle motion (sometimes called vertical, or P). The inline and crossline shear source components are referred to as source components X and Y, respectively, while the vertical source component is referred to as source component Z. A single 3-component (3C) geophone was placed at each receiver station; receivers contained horizontal (orthogonal) elements oriented inline and transverse to the line (receiver components X and Y, respectively), and a vertical element (receiver component Z). Data are referred to throughout this paper in a way that indicates the specific source and receiver type of the component. For example, a common-mode data component acquired using an inline source and an inline receiver is referred to as the XX component (sometimes called inline-inline, or SV-SV). A converted-mode data component acquired using a vertical source and inline receiver is referred to as the ZX component (sometimes called vertical-inline, or P-SV).

XX DATA VERSUS YY DATA: FACTORS AFFECTING IMAGING POTENTIAL

Shot gathers (i.e. field records showing all of the traces acquired for a single source location) for the XX and the YY components acquired at four locations along line EBTravel are shown in Figure 3. The source locations for the gathers correspond to road stations: 48581, 48638, 48727, and 48789. The gathers are shown as unprocessed data (top), as data with bandpass filter and AGC applied (middle), and as interpreted data (bottom). S-wave reflections are indicated on the interpreted data records at zero-offset arrival times of about 0.11 seconds (110 ms). The character of reflection events relative to reflection events observed to the east and west, and the depth estimates using velocities derived from reflections when correlated with the available drill log data (Table 2), indicate that the observed reflections (indicated on Figure 3) are caused by the overburden and bedrock boundary. The average dominant frequencies of the reflections in the field records that were acquired using X- and Y-oriented sources had similar values. Processed multicomponent gathers (source and receiver component combinations: XX, XY, XZ, YX, YY, YZ) with the same source locations as the gathers in Figure 3, are shown in Figure 4. S-wave reflections from the top-of-bedrock are also indicated on the gathers in Figure 4. Gathers in Figures 3 and 4 have individual trace amplitude scaling applied to enhance the later time events.

Noise, Optimal Reflection Windows, and Reflection Coefficients

In order to produce an accurate image of the subsurface using seismic reflection data, reflection energy (signal) must first be confidently identified. Random noise and different types of coherent noise (non-reflection energy) recorded by geophones can destructively interfere with signal of interest, and this can make accurate seismic imaging difficult or even impossible. A main goal of seismic reflection data processing is to enhance identifiable signal by suppressing noise, and to generate images of the subsurface containing minimal noise-related artifacts.


Figure 2. Multicomponent reflection data nomenclature in terms of source and receiver orientations and preferential polarizations. For sources, the X and Y symbols indicate sources configured to preferentially generate shear particle motion inline and transverse (crossline) to the seismic line, respectively, while Z indicates a vertical source configured to preferentially generate compressional particle motion. For receivers, X and Y indicate horizontal geophone elements oriented inline and transverse to the line, respectively, while Z indicates a vertical geophone element. The source and receiver pairs: XX, YY, and ZZ are referred to as the common-mode components of the matrix (e.g. XX means a source and receiver both oriented in the X direction).

 


Figure 3. Line EBTravel shot gathers: (a) XX component, and (b) YY component. Gathers are shown: (top) unprocessed, (middle) with a bandpass filter (50-80-160-200 Hz) and AGC (100 ms window) applied, and (bottom) interpreted. S-wave reflections from the top-of-bedrock are indicated, and apparent NMO velocities and approximate depths are given. The x-axis scales of absolute offset (AOFFSET) from the source locations (SOU_X; source locations are given in feet from the western county line) are in feet.

 

Table 2. Information from drill logs near line EBTravel (Figure 1).
Boring Station Distance (in feet) from seismic line Depth (in feet) to water Depth (in feet) to bedrock
GC-20848419 41’ North 25-27, 31-43 42
GC-21848421 6’ South 22-28, 36-42 41
GC-20948459 39’ North Not recorded 43
P-221A48500 6’ South 31-42 42
B-38*48525 5’ South 27 (24 hour level) Not available
B-12248570 24’ North 27 42
B-12348600 10’ North Not recorded 41

 


Figure 4. Processed (bandpass filter and AGC gain applied) line EBTravel multicomponent shot gathers: source components X (a), and Y (b). The S-wave reflections interpreted in Figure 3 are superimposed on the common-mode component (XX and YY) gathers. The x-axis scales of absolute offset (AOFFSET) from the source locations (SOU_X; source locations are given in feet from the western county line) are in feet.

There are several types of coherent noise that can complicate or prevent shallow seismic reflection imaging efforts. A blast of air wave noise is typically generated by seismic sources, and coherent noise can also be generated by other environmental sources, such as roadway traffic. Surface waves propagate along or near the ground surface, and often serve as noise that is detrimental to reflection energy. Two types of surface waves are commonly a concern during near-surface seismic reflection surveying: Rayleigh waves and Love waves (Sheriff and Geldart, 1982). Rayleigh waves propagate with a retrograde elliptical particle motion confined to the vertical plane in the direction of the seismic source, and involve a combination of both compressional and shear waves (P- and SV-waves). Love waves have a particle motion that is parallel to the surface and perpendicular to the direction of propagation (SH), with no vertical component of motion. Rayleigh waves will propagate along a surface regardless of the near-surface velocity structure. Love waves however, require a near-surface layer of relatively low velocity, as they result from the interference of multiples of reflected SH-wave energy beyond the critical angle that are confined to a near surface layer.

Traffic noise on the study area field records was higher for data recorded using X- and Z-oriented receivers, than for data recorded using Y-oriented receivers. Recorded traffic noise was predominantly low frequency (e.g. 5-25 Hz), and was therefore significantly suppressed on all components through the application of a low-cut frequency filter. Air wave noise was present in all data that were acquired using X- and Z-oriented receivers, and could not be substantially suppressed without degrading reflection signal quality. Air wave noise propagated along the receiver spread before the return of S-wave reflections from the bedrock. Processing procedures to attenuate air wave noise were not necessary to prevent it from severely degrading the S-wave reflection signal in the XX component data.

Detrimental, high amplitude, dispersive Love wave noise was not observed in most of the YY component data. This is because the S-wave velocity structure across most of the study area was such that near-surface road construction-related fill materials had a higher velocity than underlying overburden materials at most locations. For YY component data at far source-to-receiver offsets where the top-of-bedrock S-wave event was degraded by non-reflected energy, the noise was suppressed through f-k filtering and post-NMO stretch muting. High amplitude Rayleigh wave noise was observed in the data recorded using X- and Z-oriented receivers (Figure 4). The velocity of Rayleigh waves (VR) is dependent upon Poisson’s ratio (n), and is close to that of the S-wave velocity (VS) with the relationship is: VR = aVS, where a is a constant that is dependent upon the value of n; an equation relating a and n can be found in Davis and Selvadurai (1996). Measurements made using acquired field data indicate a representative range for values of Vr (linear group velocity) in the study area of 575 ft/s to 650 ft/s. Rayleigh wave noise tended to mask the top-of-bedrock S-wave reflection that was recorded in XX component data. The Rayleigh wave noise could not be sufficiently suppressed through frequency filtering, resulting in the optimum reflection window being narrower than the window for the YY component data (Figure 3).

S-wave energy reflected from the top of the bedrock (from an X-oriented source) is strongly polarized in the horizontal component (X-oriented receiver component) in the direction of the line (Figure 4a). This polarization effect is shown by high amplitude reflections in the X-oriented receiver gathers (top of Figure 4a), relative to those (when even observed) in Y- and Z-oriented receiver gathers (middle and bottom of Figure 4a respectively). The S-wave energy that was reflected from the top of the bedrock (from a Y-oriented source) is strongly polarized in the horizontal component in the crossline direction as shown in Figure 4b. Reflections have high amplitudes in Y-oriented receiver gathers (middle of Figure 4b), relative to those (when even observed) in X- and Z-oriented receiver gathers (top and bottom of Figure 4b respectively). The amplitude and polarization characteristics of the recorded noise modes, and the amplitude and polarization characteristics of reflected energy, determines whether or not a reflection can be observed in a given data component. In practice, acquisition geometry is never perfect, media are not perfectly isotropic and laterally homogeneous, and reflections may be affected by dipping interfaces. Further, seismic sources and receivers are not perfectly pure in a polarization sense (i.e. more than one type of wave is typically generated by sources, and receivers typically are not completely sensitive to only a single wave type). Miller and Pursey (1955) calculated that a vertically oscillating disk on a half-space medium (having a Poisson’s ratio of 0.25) radiates 67 percent of the total energy as Rayleigh waves, 26 percent as SV-waves, and 7 percent as P-waves. It is unlikely when using a Z-oriented source for example, that the X-oriented component would not contain any reflected energy.

YY component field records generally contained a higher signal-to-noise ratio, and a larger optimum reflection window than corresponding XX component records (Figure 3). Therefore, accurate estimates of the average overburden S-wave velocity could generally be better obtained from the YY component data. The signal-to-noise ratio and the size of the optimum reflection window for the XX and YY component data also affected the quality of images that could be constructed using these components.

Square root energy coefficients were calculated using the PSHSV computer program (equations) presented in Guy et al. (2003b) for the overburden and bedrock interface. For common-mode reflected P- and S-waves, the terms “square root energy coefficient” and “reflection coefficient” represent the same quantities. Representative P- and S-wave velocities were calculated from the seismic data for the saturated overburden (directly above bedrock) of 5150 ft/s and 700 ft/s respectively. An S-wave velocity of 2750 ft/s for the bedrock, and an approximate P-wave velocity of 5500 ft/s (assuming a VP/VS ratio of 2.0), were also used for calculations. The P-wave velocity of the bedrock could not be measured directly from the field data. A density contrast of 0.63 g/cm3 was estimated for the region across the interface between saturated overburden and bedrock for calculations. This approximation was based on an average density value for wet overburden materials of 1.92 g/cm3 (Telford et al., 1990), and an average bulk density of 2.55 g/cm3 measured for the study area bedrock unit in Harrison County, Ohio (ODNR, 2002). Results from calculations of the square root energy coefficients using these velocity and density estimates are plotted in Figure 5 as a function of incidence angle. These calculations are based on assumptions that the interface is planar and that the media are isotropic.

The plots in Figure 5 show that the magnitude of the S-wave reflection coefficient is high at normal incidence (0.68). The magnitudes of the SV-wave (left) and the SH-wave (right) reflection coefficients are also seen from a qualitative standpoint to be fairly similar to each other for most incidence angles. The maximum incidence angle that is shown roughly corresponds to the likely maximum source-to-receiver offset at which a top-of-bedrock S-wave reflection would be recorded in acquired field data. Energies of reflected and refracted mode-converted waves from an incident SV-wave are relatively low for this geologic situation, and such modes only occur at small (< 8 degrees) incidence angles. The reflected mode-converted P-wave and refracted mode-converted P-wave critical angles are indicated in Figure 5. Based on the plots in Figure 5 and analyses of the data presented in this section, it can be concluded that the possible effects of reflection coefficients on the S-wave imaging potential are likely to be minor relative to the affects of noise recorded in the S-wave data components.

 


Figure 5. Plots (generated using the PSHSV computer program, Guy et al., 2003b) showing normalized square root energy coefficients as a function of incidence angle for a SV-wave (left) and a SH-wave (right) incident on the overburden (medium 1) and bedrock (medium 2) interface.

Stacked Signal Quality

An S-wave velocity field could be more accurately constructed through analysis of YY component data than from XX component data. In most cases, the average overburden S-wave velocities measured from both components were similar, when the velocity was measured at a location using both components. Small differences in the average S-wave overburden velocities (< 3 percent) were measured from corresponding XX and YY data in a relatively small number of cases (e.g., see the XX and YY component shot gathers shown in Figure 3 that were acquired with the sources located at road station 48581).

XX component and YY component supergathers (each consisting of 3 smashed CMP gathers) from three locations along line EBTravel are shown in Figure 6. The gathers are each shown before and after NMO-correction using S-wave stacking velocities calculated from YY component data (Table 3). For the XX component supergather centered at road station 48820 (Figure 6a), the event from the top-of-bedrock is slightly overcorrected (applied velocity was too low) using the YY component-derived velocity. However, traces observed on the dynamic stack function (plotted to the right of the NMO-corrected XX component gather in Figure 6a) indicate the data still stacked reasonably well at the applied velocity. The top-of-bedrock event is corrected and stacks give similar results for both components using supergathers that are centered at the other two locations (Figures 6b and 6c). Analyses conducted using XX and YY data, suggested that comparable stacks for the two components could be obtained using a velocity field derived from YY component data. It was apparent from these analyses that other factors previously discussed (e.g., optimum reflection windows and signal-to-noise ratios of acquired data) would have a greater impact than the small errors in the applied stacking velocities (i.e., < 3 percent) on the quality of constructed XX and YY component images.

EBTravel XX and YY component stacked time sections were produced with similar parameters (shown in Table 3), and the NMO corrections were applied using the same (YY component-derived) velocity field. Stacked XX and YY component data from the line were generated with NMO corrections applied at 100 percent of the stacking velocities, at 95 percent, and at 105 percent. At each velocity-field percentage, the YY component data had a higher top-of-bedrock reflection signal-to-noise ratio than the XX component data, and the YY component data showed better resolution of the bedrock horizon than the XX component data. The YY component data clearly provided a better potential than XX component data to image the top-of bedrock and to identify any disruptions in the bedrock horizon.

 


Figure 6. Line EBTravel CDP supergathers: XX component (left), and YY component (right). Gathers are shown before and after NMO correction using YY component-derived (Table 3) S-wave stacking velocities of: (a) 737 ft/s, (b) 705 ft/s, and (c) 675 ft/s. Arrows next to dynamic stack functions indicate the top-of-bedrock reflection event. This reflection is slightly over-corrected on the 48820 (a) XX component gather (although data still stack reasonably well at the applied velocity). For supergathers centered at the other two locations (b and c), this event is corrected similar on both the XX and YY components. CDP locations (CDP_X) correspond to road stations (given in feet from the western county line).

Table 3. Data processing flow for common-mode S-wave (XX and YY components) and P-wave (ZZ component) reflection data.
Processing step Description
Data reformat From SEG-Y to ProMAX format
Vibroseis correlation Correlated with pilot sweep
Geometry Defined using field notes and loaded to headers
Data truncation Records truncated to 300 ms
Trace editing Bad / noisy traces killed
Trace equalization 150 ms spatially varying window
f-k filterNon-reflection energy/linear noise suppression
CMP sort Sorted from shot gathers to midpoint gathers
Velocity analysis Integrated analysis of shot gathers, constant velocity stacks, and semblance plots for YY data; integrated shot gathers and constant velocity stacks analysis for ZZ data
NMO correction Applied based on optimum stacking velocities
Stretch mute 30 percent for line YY data, and 40 percent for ZZ data
Bandpass filter Zero-phase Ormsby filters: 50-80-160-200 Hz for YY data, and 80-120-200-240 Hz for ZZ data
AGC scaling 100 ms window
CMP ensemble / stack Summed NMO-corrected CMP gathers

 

YY DATA VERSUS ZZ DATA: FACTORS AFFECTING IMAGING POTENTIAL

Line EBTravel ZZ component shot gathers with source locations corresponding to road stations: 48580, 48638, 48726, and 48788 are shown in Figure 7a. The shot gathers are shown as unprocessed (top), with bandpass filter and AGC applied (middle), and with interpretations (bottom). The YY component shot gathers (unprocessed, processed, and interpreted) shown in Figure 3 are shown again in Figure 7b, for the purpose of comparison with ZZ component gathers. Note that the source spacing for the XX and YY data differed from that used for the ZZ data by one foot (Table 1). Multicomponent shot gathers (components: ZX, ZY, and ZZ) with the same source locations as the gathers shown in Figure 7a, are presented in Figures 8a (unprocessed) and 8b (processed). The P-wave reflections that were indicated on ZZ component records on Figure 7a are again indicated on Figure 8b. The frequency spectra of multicomponent records acquired using a Z-oriented source were similar in all cases. A bandpass frequency filter is often effective in improving the visibility of near-surface P-wave reflections (Steeples et al., 1997); the filter applied to these data was an 80-120-200-240 Hz zero-phase Ormsby filter.

Sensitivity of P- and S-Waves to Changes in Overburden Saturation

Reflected P-wave energy is indicated on the interpreted records in Figure 7a at zero-offset arrival times of about 30 ms. Depth estimates using velocity values that were derived from reflections when correlated with available drill log information (Table 2) indicate that the observed reflections are from the top-of-saturated-overburden (where dry or partially saturated unconsolidated overburden materials become fully saturated). Depth estimates from the events interpreted in Figure 7a, are also consistent with Hoffman et al. (1995), where it was reported that the groundwater within the overburden was generally within 30 feet (9.1 m) of the ground surface. No major lithologic boundaries were encountered during drilling in this depth range.

At certain locations along line EBTravel, high amplitude P-wave reflections and refractions from the top-of-saturated overburden were clearly observed, and were well separated in the frequency domain from recorded noise modes. However, at most locations along the line, the arrival times of events and the noise affecting the data made it challenging to process the data to clearly show the P-wave reflections from the top-of-saturated-overburden. P-wave reflections from geologic interfaces below the top-of-saturated-overburden, were not observed in any of the acquired records (this is subsequently explained).

Reflection and refraction analyses of data acquired in the study area indicated average P-wave velocities for unsaturated overburden around 1900 ft/s, and P-wave velocities around 5150 ft/s for the saturated overburden. Such large contrasts in unconsolidated material P-wave velocities due to water saturation have commonly been observed during near-surface reflection surveys (Miller and Xia, 1998). P-wave velocities of fully saturated unconsolidated materials are almost always equal to or greater than (except in cases of very high porosity, i.e., greater than or approximately equal to 65 percent) the velocity of water (Bradford, 2002). Acquired S-wave data (Figure 7b) contained S-wave reflections from the top-of-bedrock, but did not contain events correlating to top of saturated overburden (i.e., changes in overburden saturation could not be inferred using S-waves); this is because P- and S-waves are sensitive to different physical material properties.

 


Figure 7. Line EBTravel shot gathers: (a) ZZ component, and (b) YY component. Gathers are shown: (top) unprocessed, (middle) with a bandpass filter (ZZ: 80-120-200-240 Hz, YY: 50-80-160-200 Hz) and AGC (100 ms window) applied, and (bottom) interpreted. P-wave reflections from the top-of-saturated-overburden (a) are indicated in blue, and S-wave reflections from the top-of-bedrock (b) are indicated in yellow, with apparent NMO velocities and approximate depths given. The x-axis scales of absolute offset (AOFFSET) from the source locations (SOU_X; source locations are given in feet from the western county line) are in feet.

 


Figure 8. Line EBTravel multicomponent (source component Z, receiver components X, Y, and Z) shot gathers: (a) unprocessed, and (b) processed (bandpass filter: 80-120-200-240 Hz, and AGC gain with 100 ms window applied). P-wave reflections interpreted in Figure 7 are superimposed on the processed common-mode component (ZZ) gathers. The x-axis scales of absolute offset (AOFFSET) from the source locations (SOU_X; source locations are given in feet from the western county line) are in feet.

P-wave velocities (and amplitudes) can vary substantially with changes in pore fluid content (Domenico, 1976), primarily because of the sensitivity of bulk modulus (B) values to such changes. P-wave velocity (VP), is dependent upon B, the shear modulus (G), and bulk density (r): VP = {(B + 4/3G)/r}1/2; S-wave velocity (VS) depends upon G and r: VS = (G /r)1/2. The rigidity of ideal gases and liquids is the same (zero). Despite possible changes in VS related to changes in density or cohesion (West and Menke, 2000), the saturation of an unconsolidated material with water generally does not change the S-wave velocity appreciably relative to the change in P-wave velocity.

Reflection Coefficients: Unsaturated and Saturated Overburden Interface

Square root energy coefficients (Guy et. al., 2003b) were calculated for P- and S-waves incident on an interface between unsaturated and saturated overburden using data-derived P- and S-wave velocities, and approximate bulk densities. Recall that for common-mode reflected P- and S-waves, the terms “square root energy coefficient” and “reflection coefficient” represent the same quantities. Results are plotted in Figure 9 as a function of incidence angle, and are based on assumptions that the interface is planar, and that the media are isotropic. Representative study area media parameters used for the square root energy calculations are presented in Table 4. P-wave velocities for the unsaturated and saturated overburden of 1900 ft/s and 5150 ft/s respectively, were used for calculations of the parameters shown in Table 4. A variation in S-wave velocity across the unsaturated and saturated overburden interface was not detectable using seismic data, and an S-wave velocity of 700 ft/s was assumed as a representative value for the entire study area overburden in calculations. A contrast of 0.37 g/cm3 was assumed as a density contrast at the interface between the unsaturated and saturated overburden. This was based on average density values of 1.55 g/cm3 for dry overburden materials and 1.92 g/cm3 for wet overburden materials (Telford et al., 1990).

The plots in Figure 9 generally demonstrate that for this geologic situation:

(1) The magnitude of the P-wave reflection coefficient is much larger for all angles of incidence than the magnitude of the SV- or SH-wave reflection coefficients.

(2) The energy of the reflected and refracted mode-converted waves from an incident P- or SV-wave are relatively small for all angles of incidence.

The P-wave impedance contrast affecting the results in Figure 9 resulted from both velocity and density contrasts, but the S-wave impedance contrast affecting the results in Figure 9 resulted only from density contrasts. The magnitude of the P-wave reflection coefficient at normal incidence is 0.54, while the magnitude of the S-wave reflection coefficient at normal incidence is 0.11 (Table 4, Figure 9). Using a bedrock S-wave velocity of 2750 ft/s, an approximated bedrock P-wave velocity of 5500 ft/s (assuming a VP/VS ratio of 2.0), and an average bedrock bulk density of 2.55 g/cm3 (ODNR, 2002), normal incidence (cumulative) square root energy coefficients at the underlying overburden-bedrock interface are considered. The cumulative square root energy coefficients of reflected P- and S-waves (from incident P- and S-waves respectively) from the top-of-bedrock are 0.08 and 0.61 respectively (Table 4).

 


Figure 9. Plots (generated using the PSHSV computer program, Guy et al., 2003b) showing normalized square root energy coefficients as a function of incidence angle for a P-wave (left), a SH-wave (middle), and a SV-wave (right) incident on the interface between unsaturated overburden (medium 1) and saturated overburden (medium 2).

Table 4. Representative subsurface parameters and normal incidence P-wave and S-wave cumulative reflection coefficients for the study area subsurface lithologies and related contrasts in impedance.

General conclusions can be made based on calculations presented in this section, as follows.

  1. P-wave energy reflected from the top-of-saturated-overburden (from an incident P-wave) is large relative to the S-wave energy reflected from this interface (from an incident S-wave).

  2. P-wave energy reflected from the top-of-saturated-overburden (from an incident P-wave) is large relative to the P-wave energy reflected from the top-of-bedrock (from an incident P-wave).

  3. S-wave energy reflected from the top-of-bedrock (from an incident S-wave) is large relative to the S-wave energy reflected from the top-of- saturated-overburden (from an incident S-wave).

  4. S-wave energy reflected from the top-of-bedrock (from an incident S-wave) is large relative to the P-wave energy reflected from the top-of-bedrock (from an incident P-wave).

Noise and Optimal Reflection Windows

The air wave and high amplitude Rayleigh wave noise (in data recorded using X- and Z-oriented receivers) had arrival times at the near offsets that were similar to those of P-wave reflections from the top-of-saturated-overburden in the ZZ component data. The noise could not be sufficiently suppressed through frequency filtering without degrading the reflection signal, since the noise existed within the optimum P-wave reflection signal frequency range (Figure 7). A frequency-wavenumber (f-k) filter that was applied to the ZZ component data (prior to stacking) suppressed the air wave and the surface wave noise, but the necessity of attenuating the noise resulted in a narrow optimum window to enhance the P-wave event. The YY data S-wave event from the top-of-bedrock was not severely degraded by air wave or surface wave noise. The YY component (top-of-bedrock) S-wave reflection window was wider than the ZZ component (top-of-saturated-overburden) P-wave reflection window (Figure 7). A larger number of traces containing reflection energy could therefore be summed and stacked for the YY component data than for the ZZ component data.

Rayleigh wave noise, air wave noise, and P-wave energy reflected from the top-of-saturated-overburden (from a Z-oriented source), are strongly polarized in the vertical component, and to a lesser extent these sources of energy are also present on the horizontal component in the direction of the line (X-component). An X-directed component of the P-wave reflections in Figure 8b are observed with relatively low amplitude in X-oriented receiver gathers (top), and are not interpretable in Y-oriented receiver gathers (middle). The top-of-saturated-overburden could therefore be most effectively imaged using common-mode P-wave reflections in field gathers acquired using Z-oriented receivers. Reflected (non-converted mode) S-wave energy correlating to the top-of-bedrock, is also observed on certain gathers in Figure 8b recorded using X- and Y-oriented receivers (for instance, at about 110 ms in the 48788 ZX gather), and this resulted from S-wave generation by the Z-oriented source. As previously discussed (see above), seismic sources are not perfectly pure in a polarization sense.

Resolution Potential of ZZ Component Data

The resolution potential of common-mode S-wave reflection data acquired in the study area has already been discussed in detail with supporting modeling results in Guy et al. (2003a). This section focuses on resolution issues related to common-mode P-wave reflection data acquired in the study area. Synthetic seismograms were generated in order to investigate the resolution potential of ZZ component data relative to the field study area geology. These synthetic seismograms were generated using an acoustic finite-difference modeling code within the ProMAX (Landmark Graphics Corporation) software package.

The P-wave interval velocity model in Table 5 was used to generate synthetic data (Figure 10), and the model was constructed using velocities that were derived through the analysis of ZZ component shot gathers. The model contains P-wave velocities for the unsaturated and saturated overburden materials that were derived from NMO and refraction analyses. P-wave velocities for the bedrock and a coal seam beneath the model overburden were approximated for modeling purposes. These quantities were not measurable directly from the data, and were approximated by assuming a VP/VS ratio of 2.0 for both materials. S-wave bedrock velocities obtained from refraction measurements across the study area ranged from 2500 ft/s to 3000 ft/s, and a P-wave velocity of 5500 ft/s was specified for the model bedrock. An S-wave velocity of 2395 ft/s, was measured by Wolfe et al. (1989) for the Lower Freeport Coal (a bituminous coal located stratigraphically beneath the Upper Freeport Coal). Based upon this coal S-wave velocity value, a P-wave velocity of 4790 ft/s was specified for the model coal. Drill log data from the site were used to establish representative thicknesses for materials in the model that could not be determined from the seismic data. Synthetic seismograms generated with models containing approximations for material bulk densities did not change the main conclusions that are demonstrated using the modeling results in Figure 10.

A velocity versus depth plot from the P-wave interval velocity and the layer thickness model in Table 5 is shown in Figure 10a. The calculated arrival times of events from the model interfaces are plotted as a function of offset in Figure 10b. An uninterpreted seismogram generated using a zero-phase Ricker wavelet (with a center frequency of 150 Hz) as the surface-located source is shown in Figure 10c. A 150 Hz source wavelet resulted in a dominant peak-to-peak reflection frequency of 125 Hz. It is seen from the interpreted synthetic seismogram in Figure 10d, that the reflection from the top-of-saturated-overburden (the primary reflection) dominates the record at all source-to-receiver offsets. A low amplitude reflection from the top of the coal seam (having opposite polarity than the primary reflection) is interpretable from the synthetic data at near offsets, but reflections from the other model interfaces located beneath saturated overburden are not interpretable.

Unsaturated and saturated overburden P-wave velocities were obtained (after bandpass filter and AGC application) from the shot gather in Figure 10e (uninterpreted) and in Figure 10f (interpreted). The dominant reflection frequency of the synthetic data and field data shown in Figure 10 are comparable, and the interpretations in Figure 10d are superimposed on the field data in Figure 10f. The observed fit of the same reflection hyperbola and the same linear refraction (from the top-of-saturated overburden) to the synthetic and field data supports the interpretation of these events and their velocities in the field data. There is no evidence

Table 5. Velocity model used to generate the synthetic seismogram in Figure 10.
P-wave interval velocity (ft/sec) Lithology Layer thickness (ft)
1900Unsaturated Overburden 29.5
5150Saturated Overburden 11.5
5500Bedrock (shale) 20
4790Coal (bituminous) 7
5500Bedrock (shale) 112

 


Figure 10. Comparison of synthetic data with line EBTravel field data (ZZ component). Plots of the velocity model (Table 5) and calculated event arrival times are shown in (a) and (b) respectively. A synthetic seismogram generated using the model in (a) with a 150 Hz source is shown uninterpreted in (c), and interpreted in (d). A shot gather used as a basis for forward modeling is shown uninterpreted in (e), and interpreted in (f). A high reflection coefficient at the unsaturated and saturated overburden (primary) interface, noise, and interference prevent the interpretation of secondary events in field data (f).

suggesting the presence of additional events (from the top-of-bedrock or the coal seam) beneath the top-of-saturated-overburden in the field data. This results from many factors, including: the high amplitude surface wave noise present (after optimum bandpass frequency filtering) at near offsets in the field data, the high P-wave reflection coefficient at the unsaturated and saturated overburden interface, low reflection coefficients at interfaces located beneath the top of saturated overburden, and interference and poor resolution.

 

General conclusions can be made from the analysis of the resolution potential of ZZ component data discussed in this section, as follows:

1) P-wave energy reflected from the top-of-saturated-overburden dominates ZZ component synthetic data at all source to receiver offsets.

2) A high reflection coefficient at the top-of-saturated-overburden, lower reflection coefficients at deeper interfaces, noise, interference, and poor resolution prevent the interpretation of reflection events from below the top-of-saturated overburden in ZZ component field records.

P-Wave Versus S-Wave Resolution

Seismic wavelength (l) affects vertical and lateral resolution, and is related to wave velocity (V) and frequency (f): l = V / f. Using dry overburden average P- and S-wave velocities of 1900 ft/s and 700 ft/s respectively, and average dominant P- and S-wave data frequencies of 125 Hz and 80 Hz respectively, the quarter-wavelengths (l/4) of P- and S-waves in dry overburden are 3.8 feet and 2.2 feet respectively (Table 6). Using saturated overburden average P- and S-wave velocities of 5150 ft/s and 700 ft/s respectively, and average dominant P- and S-wave frequencies of 125 Hz and 80 Hz respectively, then l/4 values for P- and S-waves in saturated overburden are 10.3 feet and 2.2 feet respectively (Table 6). Ignoring other factors that could influence resolution potential, these calculations suggest based on consideration of wavelengths, the resolution that can be achieved using S-waves in the study area dry overburden is more than 1.7 times that which can be achieved using P-waves, and that the resolution of S-waves in the study area saturated overburden is more than 4.7 times that of P-waves (Table 6).

Data Processing

Common-mode P- and S-wave reflection data processing and imaging operations applied to each of the components (YY and ZZ) were established based on previous analyses of the data (see above). The analysis flows applied to each of these data components are described in Table 3. An event correlating as the top-of-bedrock could consistently be identified in S-wave records, and processing and imaging operations applied to the YY component data were determined for the purpose of enhancing reflections from this impedance contrast. ZZ component records contained P-wave reflections correlating to depths of the top-of-saturated-overburden, and processing of ZZ component data focused on enhancing reflections from this impedance contrast. As previously mentioned, P-wave reflections from the top-of-saturated overburden were not identifiable on as large of a percentage of shot gathers, as were the S-wave reflections from the top-of-bedrock (due to event arrival times and recorded noise characteristics). S-wave data required more accurate stacking velocities than did P-wave data for producing quality stacks. This is because a given deviation from optimum stacking velocity would represent a much larger percentage of the optimum stacking velocity in the S-wave data case, since stacking velocities for S-wave data were much lower than those for P-wave data.

Table 6. Quarter-wavelengths calculated for P- and S-waves in the study area overburden materials.
Lithology Average VP (ft/sec) Average VS (ft/sec) Average Dominant P-wave Frequency (Hz) Average Dominant S-wave Frequency (Hz) Quarter-wavelength (l/4) for
P-waves (ft)		 Quarter-wavelength (l/4) for
S-waves (ft)		 (P-waves l/4) / (S-waves l/4)
Dry Overburden1900 700 125 80 3.8 2.2 1.7
Saturated Overburden5150 700 125 80 10.3 2.2 4.7

 

For the common-mode component shot gathers, high amplitude, linear, non-reflected energy remained in records after optimum bandpass frequency filtering. An f-k (frequency-wavenumber) filter was therefore evaluated and applied to each of the common-mode component shot records (after amplitude balancing) to suppress coherent, linear non-reflection energy. This processing step was critical for suppressing high amplitude surface wave noise (having arrival times and frequency content at near offsets similar to those of P-wave reflections) recorded in ZZ component data. f-k filtering was not as critical of a process for the XX or YY data, as at the arrival times of S-wave reflections linear noise was recorded at far offsets, and could therefore be largely suppressed after NMO corrections by proper stretch muting. However, evaluation of f-k filter effects on the XX and YY component data showed that an improvement in stacked signal quality, with minimal artifact generation was obtained. The effectiveness of f-k filtering in improving records (as determined by noise suppression and reflection signal enhancement) prior to CMP sorting and stacking is demonstrated using ZZ and YY component field records from line EBTravel (Figures 11a and 11b respectively). These data were recorded with the respective source located at road station 48638, and both records contained identifiable reflections prior to application of f-k filters. Bandpass filters and AGC were applied to the gathers before generating the plots in Figure 11, in order to demonstrate non-reflected energy suppression using f-k filters across reflection signal bandwidths.

P- and S-Wave Stacked Section Imaging

Representative processed (Table 3) and stacked line EBTravel P-wave (ZZ component) and S-wave (YY component) sections are shown in Figure 12. The P-wave reflection event (with an average dominant frequency of 125 Hz) at 28 to 33 ms in Figure 12a correlates as the top-of-saturated-overburden on the P-wave depth section (Figure 12c). The appearance of this reflection event on the P-wave section has been affected from the stacking of a certain amount of noise along with reflection signal. This is due to the frequency content and related slopes and velocities of the noise relative to P-wave reflections. Noise modes could not be as effectively suppressed in ZZ component data as they could be in YY data. The S-wave reflection event (with an average dominant frequency of 80 Hz) at 105 to 115 ms in Figure 12b correlates as the top-of-bedrock on the S-wave depth section (Figure 12d). These interpretations are supported by the results obtained through shot gather analyses, modeling, and drill log data (see above).

A comparison of sections from both (ZZ and YY) components (and consideration of the results previously presented) indicates that there are specific advantages and disadvantages associated with P- and S-wave reflection data acquired in the study area, as follows.

(1) The top-of-saturated-overburden served as a detectable impedance contrast for P-waves but not for S-waves.

(2) P-wave data cannot be used to image impedance contrasts located below the saturated overburden, while images of the top-of-bedrock and any disruptions along this horizon can be constructed using S-wave reflections.

 


Figure 11. Line EBTravel ZZ component (a) and YY component (b) shot gathers (source located at road station 48638, east direction to left) and f-k spectra: (left) gathers without f-k filter applied showing reflections with zero-offset times of 32 ms (a) and 110 ms (b), and f-k amplitude spectra showing defined mute polygons, (middle) with polygons rejected to suppress noise (indicated in boxes 1 and 2), and (right) with polygons accepted (showing noise rejected through filter application). Bandpass filters and AGC gain were applied to the gathers before generating these plots in order to demonstrate f-k filter non-reflection energy suppression across reflection signal bandwidths. The x-axis scales of absolute offset from the sources are in feet.

 


Figure 12. Line EBTravel ZZ component (a) and YY component (b) time sections with fold (TR_FOLD) plots. The P-wave event at 28 to 33 ms (a) is the top-of-saturated-overburden (blue) on the depth section (c). The S-wave event at 105 to 115 ms (b) is the top-of-bedrock (yellow) on the depth section (d). The scales on the bottom x-axes of (a) and (b) show P-wave and S-wave stacking velocities respectively (velocity scales are to the right of the sections). CDP location numbers (CDP_X) correspond to road stations (units are in feet). (c) and (d) Line EBTravel ZZ component and YY component depth sections. CDP location numbers (CDP_X) correspond to road stations (given in feet from the western county line). See previous page for complete caption.

 

Insitu Physical Properties Measurement Potential Using Reflections

Seismic reflection measurements offer the potential for determining insitu elastic properties of soil and rock volumes in a cost-effective manner. Although geotechnical measurements can be made using boreholes at a sampling interval that is typically finer than seismic data resolution potential, elastic properties determined from seismic measurements have the advantage of not being potentially affected by material disturbances related to sampling or tool insertion. For these reasons the potential for estimating Poisson’s ratio (n) values from the acquired data was evaluated. It is possible in theory to produce a subsurface map showing both lateral and vertical variations in the VP/VS ratio, through the correlation of P- and S-wave reflections. This has previously been accomplished for petroleum exploration purposes, in deep earth studies that have acquired multicomponent reflection data (Garotta, 2000). It is also possible to determine variations in Poisson’s ratio (n), because the VP/VS ratio and n can be related. The relationship between VP/VS and n can be described as:

n = {(VP2 / 2VS2) - 1} / {(VP2 / VS2) - 1}.

The VP/VS ratio (and therefore n) of a geologic material is a function of many factors, including lithology, porosity, cementation, depth, age, temperature and pressure regime, and interstitial fluids (Tatham and McCormack, 1991). Typical values of VP/VS range from 1.7 to 2.0 for rocks, and are often in the range of 2.0 to 7.0 for shallow unconsolidated sediments (Hasbrouck and Padget, 1982). Because the elastic constants of geologic materials are defined as positive numbers, values of n for geologic materials are within the range of 0.0 to 0.5 (a maximum value of 0.5 corresponds to fluids with no shear strength). In petroleum exploration studies, n values around 0.05 have been measured for very hard, rigid materials, while n values between 0.25 and 0.33 have typically been measured for limestones, sandstones, and many common igneous and metamorphic rocks (Telford et al., 1990; Dobecki, 1993). For soft, shallow unconsolidated geologic materials, n values between 0.45 and 0.49 are common (Dobecki, 1993), with n values in the range of 0.496 to 0.498 typical for near-surface clayey, saturated sediments (Benjumea et al., 2001).

The potential for determining detailed variations in n for geologic materials in the study area subsurface, based on acquired data reflection information, is limited. This limitation results from the small number of reflection events in acquired data, and due to the fact that P- and S-wave reflections from similar subsurface interfaces cannot be correlated. Representative lithology values of n were able to be estimated (Table 4) however, using velocity information obtained from both reflection and refraction analyses, and by making certain assumptions. Using a P-wave velocity of 1900 ft/s (measured from a reflection from the top of saturated overburden) and a S-wave velocity of 700 ft/s (measured from a reflection from the top-of-bedrock), and by assuming that S-wave velocities above and below the top-of-saturated-overburden are the same, a VP/VS ratio of 2.7 and a n value of 0.42 are calculated for unsaturated overburden (Table 4). By using a refraction-derived P-wave velocity of 5150 ft/s, and by making the same assumption regarding S-wave velocity as that which was made in the previous example, a VP/VS ratio of 7.3 and a n value of 0.49 are calculated for saturated overburden (Table 4). Bedrock S-wave velocities could be measured from refractions in field data, however, bedrock P-wave velocities could not be directly measured from the acquired data, and calculations of a representative n for bedrock were therefore dependent upon an assumed VP/VS ratio (Table 4).

CONCLUSIONS

The effectiveness of high-resolution P- and S-wave reflection surveys for mapping a shallow stratigraphic sequence was evaluated, and factors influencing the effectiveness of common-mode reflection information for allowing near-surface horizons to be imaged, and discontinuities to be detected were determined. Results of this study demonstrate that due to differences in P- and S-wave propagation, media compressional and shear impedance contrasts, and variations in receiver sensitivity (as a function of orientation), it is necessary to consider the probable usefulness of recording and analyzing different data components/wave-type reflections for meeting the objectives of a project, prior to conducting a near-surface seismic reflection survey. Conducting a multicomponent data acquisition test phase and/or modeling based on geophysical well log information prior to a high-resolution reflection survey will have associated cost, and for this reason such work is not typically done. Information obtained through this type of pre-survey work however, will often outweigh the costs of conducting it, as data components/wave-type reflections can be selected for acquisition during the actual production-phase that will provide the best potential for meeting project objectives. Dependent upon situational factors, near-surface characterization may be able to be most effectively accomplished through the acquisition of one or multiple data components/wave-type reflections.

The risk of acquiring only one data component without first evaluating the probability that it will allow project objectives to be successfully met can be great. For example, in the study project area, successful imaging of multiple subsurface features of interest would not have been accomplished if only the traditional ZZ reflection component had been acquired. Due to the subsurface conditions and acquired data characteristics, the bedrock horizon could be effectively imaged using YY component reflection information, whereas it could not be effectively imaged or imaged at all using other acquired components. Although not effective relative to YY component data in this study area, XX component data acquisition and analysis could be worthwhile for other near-surface studies. The near-surface shear-velocity structure was such that high amplitude Love wave noise did not prevent SH-reflection imaging in the study area. In situations where detrimental Love wave noise does saturate the Y-receiver component, the XX component may allow better common-mode S-wave reflection image construction if a large percentage of incident SV-wave energy is not converted to P-wave energy at a target horizon, and if the optimum target reflection window is adequate (this will be dependent upon reflection arrival time, and characteristics of noise modes recorded on X-component receivers). Work in the petroleum industry has shown that it can be advantageous to concurrently analyze multiple S-wave reflection data components for the purpose of measuring near-surface media anisotropy and such benefits may be extended to shallow earth studies.

S-waves were insensitive to changes in overburden moisture content in the study area subsurface, whereas common-mode P-wave reflections from the top-of-saturated-overburden were recorded in ZZ component data. Event arrival times and characteristics of recorded noise modes however, made it challenging to process and use P-wave reflections. Forward modeling, shot gather analyses, and consideration of reflection coefficients indicated that common-mode P-wave reflections from impedance contrasts beneath the top-of-saturated overburden were not observed in study area due to: surface wave and airwave noise, a high-P-wave reflection coefficient at the top-of-saturated-overburden, low P-wave reflection coefficients at deeper interfaces, and interference and poor resolution. The percentage of incident P-wave energy reflected (as P-wave energy) from the top-of-bedrock was small relative to the percentage of incident S-wave energy reflected (as S-wave energy) from the top-of-bedrock. Further, the top-of-bedrock S-wave reflection in YY component data was not severely degraded by airwave or surface wave noise, and the S-wave reflection window for this event was wider than the P-wave reflection window for the top-of-saturated overburden event. Because S-waves travel at lower velocity than P-waves, it was possible to increase resolution potential by using S-wave reflections. For the study area, the resolution potential of S-waves in dry overburden was more than 1.7 times that of P-waves, and in saturated overburden the S-wave resolution potential was more than 4.7 times that of P-waves. The ability to increase resolution potential (using S-waves) during near surface studies will be dependent upon media P- and S-wave velocities, P- and S-wave frequencies that can be generated and recorded, and absorption per unit wavelength of propagating P- and S-waves in subsurface media. This study also demonstrated that concurrent analysis of P- and S-wave reflection information can allow for the estimation of insitu elastic media properties.

REFERENCES

  1. Bacharach, R. and A. Nur (1998) High-resolution shallow-seismic experiments in sand, part 1: water table, fluid flow, and saturation, Geophysics, vol. 63, pp. 1225-1233.

  2. Benjumea, B., J.A. Hunter, S.E. Pullan, R.A. Burns, R.L. Good (2001) Near-surface seismic studies to estimate potential earthquake ground motion amplification at a thick soil site in the Ottawa river valley, Canada, in: Proceedings of the Symposium on the Application of Geophysics to Engineering and Environmental Problems, 11 pp.

  3. Bradford, J.H. (2002) Depth characterization of shallow aquifers with seismic reflection, part 1 - the failure of NMO velocity analysis and quantitative error prediction, Geophysics, vol. 67, no. 1, pp. 89-97.

  4. Carr, B.J., Z. Hajnal, and A. Prugger (1998) Shear-wave studies in glacial till, Geophysics, vol. 63, pp. 1273-1284.

  5. Clark, J.C., W.J. Johnson, and W.A. Miller (1994) The application of high resolution shear wave seismic reflection surveying to hydrogeological and geotechnical investigations, in: Symposium on the Application of Geophysics to Engineering and Environmental Problems, pp. 231-245.

  6. Conroy, J. P, and E. D. Guy (2005) Borehole tomography and surface 3D radar for coal mine subsidence detection, Electronic Journal of Geotechnical Engineering, vol. 10, 20 p.

  7. Davis, R.O., and A.P.S. Selvadurai (1996) Elasticity and geomechanics, Cambridge University Press, 201 pp.

  8. Domenico, S.N. (1976) Effect of brine-gas mixture on velocity in an unconsolidated sand reservoir, Geophysics, vol. 41, pp. 882-894.

  9. Garotta, R. (2000) Shear waves from acquisition to interpretation: short course notes, distinguished instructor series no. 3, Society of Exploration Geophysicists, Tulsa, Oklahoma.

  10. Goforth, T., and C. Hayward (1992) Seismic reflection investigations of a bedrock surface buried under alluvium, Geophysics, vol. 57, pp. 1217-1227.

  11. Guy, E.D. (2004) Evaluation of near-surface converted-mode seismic reflection imaging potential, Electronic Journal of Geotechnical Engineering, vol. 9, 35 p.

  12. Guy, E.D., J.A. Nolen-Hoeksema, J.J. Daniels, T. Lefchik (2003a) High-resolution SH-wave seismic reflection investigations near a coal mine-related roadway collapse feature, Journal of Applied Geophysics, vol. 54, pp. 51-70.

  13. Guy, E.D., S.J. Radzevicius, J.P. Conroy (2003b) Computer programs for application of equations describing elastic and electromagnetic wave scattering from planar interfaces, Computers & Geosciences, vol. 29, pp. 569-575.

  14. Harris, J.B., R.D. Miller, and J. Xia (2000) Near-surface shear wave reflection surveys in the Fraser River Delta, B.C., Canada, in: Annual Meeting Society of Exploration Geophysicists Expanded Abstracts.

  15. Hasbrouck, W.P. (1990) Results of shear-wave measurements, in: Symposium on the Application of Geophysics to Engineering and Environmental Problems, Environmental and Engineering Geophysical Society, Golden, Colorado, p. 201.

  16. Hasbrouck, W.P., and Padget, N. (1982) Use of shear wave seismics in evaluation of strippable coal resources, Utah Geological and Mineral Survey Bulletin, vol. 118, pp. 203-210.

  17. Hoffman, A.G., Clark, D.M., and Bechtel, T.D., (1995) Abandoned deep mine subsidence investigation and remedial design, Interstate 70, Guernsey County, Ohio, in: Proceedings of the 46th Highway Geology Symposium, Charleston, West Virginia, 15 p.

  18. Hunter, J.A., Pullan, S.E., Burns, R.A., Gagne, R.M., and Good, R.L., (1984) Shallow seismic reflection mapping of the overburden-bedrock interface with the engineering seismograph - some simple techniques, Geophysics, vol. 49, pp. 1381-1385.

  19. Miller, G.F., and H. Pursey (1955) On the partition of energy between elastic waves in a semi-infinite solid, Proceedings of the Royal Society, A233, pp. 55-69.

  20. Miller, R.D., and D.W. Steeples (1990) A shallow seismic reflection survey in basalts of the Snake River Plain, Idaho, Geophysics, vol. 55, pp. 761-768.

  21. Miller, R.D., and J. Xia (1998) Large near-surface velocity gradients on shallow seismic reflection data, Geophysics, vol. 63, no. 4, pp. 1348-1356.

  22. Ohio Department of Natural Resources (ODNR), Division of Geological Survey (2002) Geophysical Log, Archer Township, Harrison County, Ohio, permit number 290.

  23. Sheriff, R.E., and L.P. Geldart (1982) Exploration seismology, vol. 1: history, theory, and data acquisition, Cambridge University Press, New York, 253 p.

  24. Tatham, R. H., and M. D. McCormack (1991) Multicomponent seismology in petroleum exploration: Vol. 6 of Investigations in Geophysics, Society Of Exploration Geophysicists, Tulsa, Oklahoma, 248 p.

  25. Telford, W.M., L.P. Geldart, and R.E. Sherrif (1990) Applied geophysics, second edition, Cambridge University Press, New York, 770 p.

  26. van der Veen, M., and A.G. Green (1998) Land streamer for shallow seismic data acquisition: evaluation of gimbal-mounted geophones, Geophysics, v. 63, pp. 1408-1413.

  27. West, M., and W. Menke (2000) Fluid-induced changes in shear velocity from surface waves, in: Proceedings of the Symposium on the Application of Geophysics to Engineering and Environmental Problems, pp. 21-28.

  28. Wolfe, P.J., B.H. Richard, and T.G. Holdeman (1989) Seismic waveguide characteristics of an eastern Ohio coal seam, Transactions of the Society of Mining Engineers of AIME, vol. 284, pp. 1831-1834.

 

© 2006 ejge