ARTICLE IN PRESS

International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

INTERNATIONAL SOCIETY FOR ROCK MECHANICS COMMISSION ON APPLICATION OF GEOPHYSICS TO ROCK ENGINEERING SUGGESTED METHODS FOR LAND GEOPHYSICS IN ROCK ENGINEERING

CONTENTS 1. 2. 3. 4. 5. 6. 7. 8. INTRODUCTION............................................................................................................................................ 886 SEISMIC REFRACTION................................................................................................................................ 887 SHALLOW SEISMIC REFLECTION............................................................................................................ 891 ELECTRICAL.................................................................................................................................................. 895 ELECTROMAGNETIC................................................................................................................................... 899 GROUND PENETRATING RADAR............................................................................................................ 903 GRAVITY........................................................................................................................................................ 906 RADIOMETRIC.............................................................................................................................................. 911

ACKNOWLEDGEMENT....................................................................................................................................... 913 REFERENCES........................................................................................................................................................ 914 Co-ordinator T. Takahashi (Japan)

1365-1609/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijrmms.2004.02.009

ARTICLE IN PRESS

International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

ISRM Suggested Methods for land geophysics in rock engineering

T. Takahashi*,1
Tsukuba Technical Research and Development Center, OYO Corporation, 43 Miyukigaoka, Tsukuba, Ibaraki 305-0841, Japan Accepted 19 February 2004

1. Introduction In recent years, geophysical applications have been expanding from natural resources exploration to the elds of civil engineering, disaster prevention and environmental preservation. Their applications to rock engineering are also growing signicantly. Geophysical technology itself has been advancing very rapidly. New technologies have been developed and new improvements have been accomplished in many geophysical methods. As a result, accuracy and reliability of geophysical images of underground have been increased. As advances are made in geophysics, it becomes more important to make an optimal survey plan and interpret its result correctly. There are many procedures in geophysical applications such as planning, data acquisition, data processing, interpretation and reporting. Therefore, to use geophysical methods more effectively in rock engineering projects, suggestions or recommendations for the procedures of geophysical applications are needed. To promote geophysical methods in rock engineering, the International Society for Rock Mechanics (ISRM) has, therefore, organized and published the following suggested methods for geophysics: (1) Suggested Methods for Geophysical Logging of Borehole by Commission on Standardization of Laboratory and Field Tests (1981), (2) Suggested Methods for Seismic Testing within and between boreholes by Commission on Testing Methods (1988).
*Tel.: +81-298-51-6621; fax: +81-298-51-5450. E-mail address: [email protected] (T. Takahashi). 1 Please send any written comments on this ISRM Suggested Methods to Prof. K. Sassa, President of the ISRM Commission on Application of Geophysics to Rock Engineering, 41 Ohmiya Nakabayashi, Kita-Ku, Kyoto 603-8404, Japan. 1365-1609/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijrmms.2004.02.009

In Recommendations of site investigation techniques published in 1975 by the ISRM, general recommendations on the use of geophysical methods were described in Chapter 4. This recommendation describes the general use of geophysics in various stages of the investigation, such as planning, design, construction and maintenance in rock engineering projects. Geophysical methods are divided into two categories, land geophysics and borehole geophysics, by the observation geometry used in the method. In land geophysics, the ground surface is used for its observation, while borehole geophysics uses the borehole(s) for its observation. Existing suggested methods above described the methods and the procedures used in borehole geophysics. Recently, many land geophysical surveys have also been used at various stages in civil engineering applications. Therefore, the suggested methods describing land geophysical methods and procedures used in rock engineering applications are needed for their practical and effective uses. These suggested methods describe the following seven land geophysical methods: seismic refraction, shallow seismic reection, electrical, electromagnetic, ground penetrating radar, gravity and radiometric. Although seismic refraction, electrical and radiometric methods were already described in the Recommendations mentioned above, new computer-aided technologies have been developed and used in rock engineering applications. These new technologies are included in the new suggested methods. Seismic reection, electromagnetic, ground penetrating radar and gravity methods have been recently employed in rock engineering applications such as characterization of rock mass and detection of shallow subsurface cavities. There are some other land geophysical methods such as the magnetic method. Since they have not been applied very much to rock engineering investigations as compared to seven methods above, they are not included in this version of the suggested methods.

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 887

These suggested methods are intended to assist geophysicists or rock engineers to use geophysics properly and obtain solutions required in rock engineering projects. Therefore, suggestions for all procedures from planning to reporting in geophysical applications are described in the suggested methods. Suggestions for the planning stage are especially emphasized, because it is most important how to select and integrate geophysical methods for obtaining optimal solutions to meet engineering requirements. The suggested methods are each composed of the following seven sections: (1) General: The principle and general features of the method are described. Standard techniques and their variations are also described. A schematic diagram explaining the principle and measurement congurations of the method is presented. (2) Applicability: General applicability of the method is described. Various limitations and conditions in application of the method are described in general. Typical exploration targets and recent applications of the method are also included. (3) Planning: Key items to be studied in planning a survey are listed up and notes for each item are described. Key items commonly described in each method are as follows: 1. Study of existing information and data: Information and data to be reviewed and studied in making a survey plan are recommended. It is also explained how to analyse the data. 2. Selection of the optimal technique and equipment: It is suggested how to select the optimal technique and equipment. Although it depends on the objectives and the requirements of the survey, guidelines are described. 3. Design of a survey line layout and data acquisition parameters: It is described in general how to design a survey line layout and data acquisition parameters. Standard layout and parameters are also presented. 4. The depth of investigation and resolution of the method: The depth of investigation, resolution and accuracy expected by the method are described in general. Their relationships to the method and equipment used are also outlined. (4) Field operation: In this section, the following key items to be considered in a eld operation are described: 1. Types of equipment: Types of equipment commonly used in the method and how to check them in advance of the measurement are described. There are some remarks in selection of equipment to be used. 2. Positioning of the measurement locations: There are some remarks to be noted in setting up the measurement points and lines.

3. Field parameter tests and quality control (QC) of the acquired data: There are remarks on what parameters and data should be tested before, during and after measurements in the eld. 4. Measurement procedures: Standard and commonly used measurement procedures in the method are described. There are also some remarks about eld techniques for acquiring higher quality data. (5) Data processing: Standard and commonly used data processing procedures for the method and a brief explanation are described. A schematic diagram is also presented for its better understanding. There are some remarks on data processing and the QC of the processed data. (6) Interpretation: Techniques and procedures for interpretation of the processed data are briey described. There are some remarks on key items in interpretation of the data. An example of the interpreted prole is also presented for demonstrating an example of the output of a survey. (7) Output and report: Standard outputs of the application of the method and minimum requirements for the survey report are listed. There are many factors and options to be examined in application of geophysical methods to rock engineering. Therefore, the suggested methods should not be automatically employed as a manual for geophysical applications. They should be used as a guideline for geophysicists and rock engineers to optimally apply geophysical methods to rock engineering and to obtain useful and valuable outputs for rock engineering projects.

2. Seismic refraction 2.1. General The seismic refraction method is a geophysical method to determine the subsurface velocity structure through an analysis of the seismic waves that return to the ground surface after refraction at the boundaries of subsurface layers with different seismic velocities. It has been widely used for many years in civil engineering applications. Although there are several types of seismic refraction methods depending on the survey objectives or targets, the most common methods are based on the rst arrivals of P-waves. The digital measuring equipment for seismic refraction surveying is becoming increasingly more compact and offers multi-channel recording capability. Data processing techniques increasingly employ automated analysis. In addition, seismic tomographic data processing techniques have been recently applied to seismic refraction data in order to derive more detailed velocity structures. Fig. 1 shows

ARTICLE IN PRESS
888 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

2.3.2. Study of existing information In planning a survey, a review should be made of existing information including borehole data, the topography, the geology, the depth to the water table, the degree of weathering, the possibility of thin layers and layers associated with velocity inversions, and dips of possible faults at the survey site. 2.3.3. Arrangement of the survey line(s) 1. The survey lines should be prepared in consideration of the survey objectives, the depth of investigation, the geological conditions and the topography. The location and length of the survey lines, the source and receiver intervals and the maximum offset distance between sources and receivers are basic parameters in planning a seismic refraction survey. 2. The seismic refraction method can accommodate rough topography but it is desirable to arrange survey line(s) to avoid extremely rough terrain. 3. Because the seismic refraction method derives two-dimensional (2-D) depth proles, the survey lines should be arranged perpendicular to the strike of the target geological structures and boundaries. Setting up a survey grid will provide tielines and facilitate delineation of targets in three dimensions. 4. The minimum length of the survey line is determined by the depth of the expected targets and the velocity structure. As a rule of thumb, this is generally around 510 times of the depth of investigation. If the survey line cannot be made to have sufcient length, then it is desirable to compensate for this by arranging remote shot points along the extension of the survey line or by arranging shot points in deep boreholes at the end of the survey line. 2.3.4. Intervals of source and receiver points 1. In most civil engineering applications for the seismic refraction method, the depth of investigation is within several tens of metres. In these cases, 10-m geophone intervals are usually adequate but for shallower targets, this interval can be reduced to 5 m or less. 2. During the survey, the geophones will be arranged in spreads of typically 12 or 24 geophones. These will be used to simultaneously measure the seismic waves arriving from a single source. For long lines, spreads should be run end to end. 3. Within each spread, the location of sources and the remote (far) source points should be planned so that there will be rst arrivals from each of the layers present along as much of the spread as possible. For the deepest layer (the main refractor), this coverage is mainly achieved by the remote shots. For the intermediate layers, the coverage is obtained using sources within the spread. It is

Fig. 1. Schematic diagram of the seismic refraction method (SEG Japan, 2000) [1].

the schematic diagram of the seismic refraction method using dynamite as a seismic source. 2.2. Applicability The seismic refraction method is applicable in situations where the P-wave velocity increases with depth. Since this is the usual situation in the near surface in a rock site, the method has been used widely for site characterization in road construction, dam construction and tunneling projects. The depths to the various layers can be determined and the seismic velocities estimated by the method can be utilized to determine lithology, rock strength, crack density, degree of weathering or metamorphism, and locations of fault zones. The method may not be useful in situations when the seismic velocity does not increase with depth (velocity inversions) and when the velocity and thickness of some layers are such that they do not give rise to seismic rst arrivals (the blind zone problem). 2.3. Planning 2.3.1. Planning the survey 1. The survey needs to be planned in consultation with the client so that any issues concerning land access and environmental restrictions are properly managed. 2. The client also needs to understand what is involved in conducting the survey and the likely results. Responsibilities for different aspects of the survey need to be clearly understood between the client and contractor. 3. Reporting requirements and time frames need to be established. 4. The supply of supplementary information that will assist in interpretation and validation of results should also be considered. This information can come from the use of other geophysical methods and from existing or additional drilling.

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 889

generally recommended that these source points should be at intervals of 3060 m. Additional source points can be considered when the structure is complex. In all cases, coverage needs to be obtained in forward and reverse directions along the spread. 2.3.5. Types of seismic sources 1. Dynamite is an excellent seismic source, especially for deeper exploration. Shots need to be buried to ensure maximum coupling of energy into the ground and to ensure that there is no blow-out causing surface damage and creating safety issues. As the timing of seismic signals needs to be accurate to millisecond accuracies, non-delay electric detonators of the instantaneous type are required. 2. During last two decades, powerful mechanical seismic energy devices like weight drops accelerated by rubber bands or vacuum and mini-vibrators have been developed. These devises may be used for surveys where the depth of investigation is up to several hundreds of metres. 3. In case of shallow surveys where the depth of investigation is less than 20 m, hammers and weight drops can be used as alternate, non-explosive sources. With such sources it is possible to repeat the impact and sum (stack) the results to build up signal strength. 2.4. Field operation 2.4.1. Equipment 1. Survey equipment generally consists of geophones (receivers), geophone cables with connecting takeouts, extension cables, a data acquisition system (including ampliers, display and recording facilities), and a detonating box (blaster). Geophones typically have a natural frequency of 30 Hz or less and are damped to ensure that there is not a strong resonance at the natural frequency. For a high-resolution survey with a short measurement line in a hard rock, higher natural frequency geophones or accelerometers should be used for high precision travel time measurements. 2. All instruments should undergo routine checks prior to use. Checks on the level and waveforms of geophone output, cable continuity and leakage, safety of the detonator, and functions of the data acquisition system are particularly important. 2.4.2. Positioning of the survey line 1. The locations and elevations of the survey line, the geophones and shot points need to be determined by appropriate surveying.

2. If there is more than one spread of geophones in a line, the ends of each spread should overlap so that continuity in the travel time data can be preserved. 2.4.3. Preparation for the measurement 1. To obtain good signals, all geophones should be planted rmly into the ground. They should be connected via take-outs to the geophone cable so that there is the same polarity for all geophones. 2. The instant of shot detonation starts the recording process. This time can be transmitted from the shot point via radio or via a cable. Prior to giving the shot re command to detonate the shot, the operator of the recording equipment (the seismic observer) needs to monitor the background noise level and to ensure that noise from wind, from road trafc, from drills, from aircraft, etc., is at a minimum. 2.4.4. Measurement 1. The observer needs to maintain an observers log detailing the locations of all geophones within a spread, the locations of the shots, the shot record numbers and the depth of each shot hole. 2. The observer needs to monitor amplier gains and/or ltering parameters to ensure the quality of the recorded data. If necessary, repeated measurements should be made to obtain better quality data. 2.4.5. Completion of survey At completion of the survey, all equipment needs to be retrieved and cleaned. The site needs to be rehabilitated in accordance with the clients requirements. The observer needs to ensure that all seismic data, observers logs and relevant survey information is properly archived and available for data processing and interpretation. 2.5. Data processing 2.5.1. Arrangement of the eld data All eld dataseismic recordings (shot records), observers logs and survey information need to be organized and compiled for a processing sequence of the type illustrated in Fig. 2. This is a standard processing sequence. There are other ways of processing and interpreting data, for example by using ray tracing techniques and through tomographic inversion. These alternative techniques are not yet widely practised and are not described in this document. 2.5.2. Picking rst arrivals First of all, the rst arrival times of the refracted P-waves are picked on the shot records in order to construct travel time curves. Usually, rst arrival times

ARTICLE IN PRESS
890 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

Observed waveform records

2.5.4. Checking of the travel time curves The travel time curves should be checked and corrected if necessary, on the basis of the following: 1. Reciprocity of the travel times. (Travel times between a pair of shot points should be equal.) 2. At each shot point, coincidence of intercept times from each of the refractors. 3. Parallel travel time curves when different shots provide travel times to the same refractor at the same locations. Bulk shifts in travel times can be used if it is decided that there is a constant delay, due to errors in the uphole correction. Individual travel times may need to be adjusted after further consideration of the shot records. 2.5.5. Derivation and verication of the velocity prole The number of refractors present in the travel times is determined on the basis of the number of changes in slope and the degree of parallel behaviour observed on the travel time curves. The 2-D velocity prole (depth section) under the survey line is then obtained by analysing the travel time curves using techniques such as the generalized reciprocal method (GRM) and Hagiwaras method. An intermediate step for these methods involves determining the velocities of the P-waves in each of the refracting layers present. Verication of the depth section using ray tracing to calculate synthetic travel time curves is desirable. These can be plotted on the corresponding observed travel time curves. 2.6. Interpretation

Preparation of waveforms

First arrival time picking

Preparation of travel time curves

Check and adjustment of travel time curves

Selection of travel time curves

Calculation of velocity - travel time curves

Calculation of depth - travel time curves

Velocity profile/depth section

Correction for layer dips (migration)

Verification with ray path calculations

Fig. 2. Flow chart for the standard data processing of seismic refraction data (SEG Japan, 2000) [1].

are picked with a time resolution of around 1 ms. For a high precision survey, time resolution is often less than 0.1 ms. Picking can be done manually on printed seismic records or using automatic and interactive computer techniques. An uphole correction is required to compensate for the burial of the shot.

2.5.3. Construction of the travel time curves Based on the distance along the survey line, the receiver intervals (the geophone spacing) and the rst arrival times, travel time curves are plotted with the horizontal axis being distance and the vertical axis travel time. For hard copy travel time curves, typical scales are 1 1 500 or 1000 for the distance axis and 5 or 10 ms to the cm for the vertical axis, see Fig. 3.

1. The depth section thus obtained is generally interpreted in consideration of the survey objectives, existing data and additional or supplemental proles if available. 2. If the travel time curves can be interpreted by two or more different models, it is desirable to report on all possible interpretations. These situations typically arise when hidden layers and velocity inversions are present. 3. The P-wave velocities obtained with a seismic refraction survey can be used as an indicator of rock quality for designing a construction such as a tunnel and a dam in rock engineering applications. 4. Fig. 3 shows an example of travel time curves and the resultant depth section. 2.7. Output and report Outputs of a seismic refraction survey should include at least the following: 1. location map of the survey site; 2. layout of survey lines;

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 891

Fig. 3. Example of travel time curves and the resultant velocity/depth prole (SEG Japan, 2000) [1].

3. 4. 5. 6.

observers logs; shot records (in digital form); travel time curves; velocity prole veried with ray paths.

3. Shallow seismic reection 3.1. General The shallow seismic reection method is a seismic method that delineates subsurface structures using seismic waves reected back to the ground surface from geological boundaries in the subsurface. Seismic reection proles provide high-resolution images of the subsurface that can easily be understood and interpreted. Fig. 4 shows a schematic diagram of the survey method. Seismic reection waves are recorded and processed to obtain a seismic reection prole. P- or S- (SH-) waves are used depending on the survey objectives. For shallow seismic reection, a 2-D survey is commonly used.

The survey report should describe at least the following items: 1. 2. 3. 4. outline of the survey; eld operations (including equipment used); data processing; results: J waveforms; J travel time curves; J velocity prole; 5. interpretation and discussion of the prole; 6. references.

ARTICLE IN PRESS
892 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

Fig. 4. Schematic diagram of the shallow seismic reection method (SEG Japan, 2000) [1].

3.2. Applicability 1. The shallow seismic reection method is best used in areas where the subsurface layers are nearly horizontal and where the terrain is not too rough. Therefore, the method is well suited for alluvial/ diluvial deposits and/or sedimentary rocks. The method is usually employed to characterize the subsurface stratigraphy and faults (especially buried faults), and to delineate bedrock topography underlying unconsolidated sediments. The method is sometimes used for detection of subsurface cavities such as karsts and old mine workings. 2. The depth of investigation for shallow seismic reection surveys is usually from several tens to several hundreds of metres. The method can be used from preliminary to detailed investigations stages in civil engineering and mining projects. 3.3. Planning 3.3.1. Study of existing information In planning the survey, all available existing information about the site should be collected and studied. It should include surface topography, geological maps and well log data if available, from which subsurface velocity distribution should be predicted. Land ownership and environmental factors also need to be investigated. Potential sources of noise should be identied. 3.3.2. Positioning of the survey lines The survey line(s) should be set up as straight as possible and oriented parallel to the dip direction and perpendicular to any cross cutting targets. Severe topographic features and areas where there is no surface access should be avoided if possible. A series

of parallel lines with occasional tie-lines assists with interpretation. 3.3.3. Source and receiver intervals Source and receiver intervals should be determined in consideration of the survey objectives and the depth of investigation. One or 2 m source and receiver intervals are often used for a very shallow survey and 5 or 10 m intervals are often used for a deeper survey. Once the source and receiver intervals have been determined for a particular survey, in general, they should remain constant throughout the survey so that the trace spacing and the fold can be maintained. 3.3.4. Number of recording channels The number of data recording channel is also determined in consideration of the survey objectives, the depth of investigation and accuracy of the reection image as well as the cost and efciency of the survey. The number is usually a multiple of 12. Twenty-four or 48 channel data for a very shallow survey and 48, 96 or more channel data for a deeper survey are acquired in general. 3.3.5. Type of seismic waves and sources used For very shallow surveys with targets less than 50 m depth, both P- and S-waves are often used. For deeper depths of investigation, P-waves are generally used. For a very shallow survey, the sources are commonly a sledge hammer blow for the P-waves and side-on hammering of a plank for the S-waves. For deeper surveys, a hydraulic vibrator, a weight drop or explosive sources are commonly used to generate the P-waves. 3.3.6. Instruments 1. The instruments for data acquisition consist of geophones, take-out cables (CDP cables), extension cables, a roll-along switch and a recording system.

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 893

2. The instruments should be tested with routine procedures prior to the eld operation. The tests should include measurements of the output levels and the waveforms of the test signals, cable continuity and leakage, and functions of the recording system. 3. Moving coil geophones similar to those used in deep seismic reection surveys in the oil and gas exploration eld are commonly used. However, for shallow seismic reection, higher natural frequency geophones are preferably used in order to obtain higher-resolution reection images compared to those for deep seismic reection surveys. 4. The data recording system should include a digital data acquisition unit, a display unit, a digital memory and a printer in the standard system conguration.

with environmental factors and any rehabilitation requirements. 3.4.2. Field acquisition parameter tests Prior to making measurements in the eld, the acquisition parameters should be tested. These parameters include the sampling rate, the recording length, the eld acquisition lters, the minimum and maximum sourcereceiver offsets and the stacking fold. 3.4.3. Measurement 1. Following the survey plan, the receivers (geophones) should be laid out on the ground and connected to the recording system through the CDP cables. The seismic source is then employed at each source point sequentially to generate seismic waves. The data acquisition operator (seismic observer) should monitor the recorded signals and judge the quality of the records by observing if reections correlate across the records. Noise, on the other hand, should be incoherent. If the data quality is not good then the measurements should be repeated at the same source location until better data are obtained. Vertical stacking is one way to improve the signal-to-noise ratio at the eld. Careful observers logs need to be maintained to ensure that an accurate log of recording parameters, source and geophone locations is maintained. 2. The recorded digital data are usually stored on a magnetic tape, a oppy disk or a hard disk of the recording system. Fig. 5 shows examples of the eld records.

3.3.7. Recording format It is preferable to record the data in the eld in one of the SEG standard formats such as SEG-2 and SEG-Y so that the eld data can later be easily input to standard processing systems.

3.4. Field operation 3.4.1. Preparation of the survey lines In accordance with the survey plan, the seismic lines should be pegged out on the ground surface and with the locations of the shot and geophone points accurately surveyed. Vegetation may need to be cleared to facilitate eld operations but this should be done in accordance

Fig. 5. Examples of eld records of P- and S-waves (SEG Japan, 2000) [1].

ARTICLE IN PRESS
894 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914
Original data Format conversion Survey data Data editing Statics analysis CDP sorting

3.5. Data processing 3.5.1. Preparation for data processing 1. The coordinates and elevations for all source and receiver locations have to be determined from the surveyors notes. 2. The le numbers of all shot records and their source and receiver locations have to be conrmed and edited. 3. Additional information such as drilling logs, VSP data and/or seismic refraction data collected at the same time as the reection data may need to be used as reference data in the processing of the reection data.

Bandpass filter

Gain recovery

Deconvolution filter

3.5.2. Data processing Shallow seismic reection data are generally processed using a standard CDP (or CMP) stacking method as devised for oil and gas exploration. Fig. 6 shows a standard data processing ow of shallow seismic reection data. Processing is usually undertaken by specialists of seismic data processing companies.

Static correction Velocity analysis NMO correction

Mute

Residual statics

Amplitude balancing

3.6. Interpretation
CDP stack

3.6.1. Interpretation of the shallow seismic reection prole In principle, a seismic reection prole provides a visualization of subsurface boundaries where acoustic impedance (or velocity) contrasts occur as the reection events. In order to obtain the geological section, the seismic reection prole should be interpreted in comparison with surface and borehole geological information as well as well logs and VSP if available. Fig. 7 shows an example of the interpreted depth section indicating buried faults. Interpretation can be undertaken on printed seismic proles or using interactive computing facilities.

Deconvolution filter

Coherency and bandpass filters Depth conversion

Migration

Time section

Depth section

Fig. 6. A standard data processing ow of shallow seismic reection data (SEG Japan, 2000) [1].

3.6.2. Resolution of the seismic reection prole The resolution of the seismic reection method depends on the parameters such as the source frequency, the velocities of the layers and the depth of the reections. Vertical resolution depends on the wavelength and horizontal resolution is controlled by the radius of the rst Fresnel zone, which is determined by the wavelength and the depth of the reector. Vertical resolution of a thin layer is ideally around 1 4 of the wavelength, but practically around a wavelength. However, it is well known that the thin layer can be 1 detected even if its thickness is 10 or less of the wavelength of the reection wave.

3.7. Output and report Outputs of a shallow seismic reection survey should include at least the following: 1. 2. 3. 4. 5. location map of the measurement lines; examples of eld records; examples of processed CDP or shot gathers; NMO velocity chart with statics; seismic reection time section on paper and in digital format (with or without migration); 6. seismic reection depth section on paper and in digital format;

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 895

Fig. 7. Example of an interpreted depth section (SEG Japan, 2000) [1].

7. interpreted time or depth section; 8. original survey recordings and other relevant information need to be archived. The survey report should describe at least the following: 1. 2. 3. 4. 5. 6. outline of the survey; eld operation (including the equipment used); data processing procedures; reection time and depth sections; interpretation and discussion of the sections; references.
C1

C1, C2: Current electrode P1, P2: Potential electrode Transmitter

I V
P1 Receiver P2 C2

Ground surface Equipotential surfaces Traces of current flow

Fig. 8. A schematic diagram of measurements involved in the resistivity method (SEG Japan, 2000) [1].

4. Electrical 4.1. General The electrical method delineates the subsurface structure and anomalies through the distribution of their electrical properties. There are several techniques and variations in the electrical method: the (direct current) resistivity method, self-potential method, induced polarization method, etc. Among them, the resistivity method is the most commonly used in civil engineering applications. The resistivity method includes some variations such as horizontal proling, vertical sounding and 2-D proling. Application of 2-D proling has increased in recent years. Fig. 8 shows a schematic diagram of the measurements involved in the resistivity method. The method is based on transmitting current into the ground through electrodes C1 and C2, and measuring the electrical potential with electrodes P1 and P2 to determine the electrical resistivity of the underground. 4.2. Applicability 1. The resistivity method is applicable to various investigations in the civil engineering eld, such as groundwater detection, landslide characterization, constructions of tunnels and dams, cavity detection and delineation of subsurface geological structure. Furthermore, it can be used in various stages of a civil engineering project from reconnaissance through site investigation to maintenance. 2. In characterization of geological structure, the method is used for delineation of fractured zones accompanied by faults, classication of weathering and alteration of rocks, and groundwater exploration.

ARTICLE IN PRESS
896 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

4.3. Planning 4.3.1. Review of existing information It is recommended to study as much background materials as possible on underground electrical properties as well as geological information in and around the survey area. 4.3.2. Layout of survey line(s) The layout of the survey line(s) is determined in consideration of the objectives of the investigation, topographical and geological conditions, the depth of investigation and resolution required, obstacles to measurement, cost and operational efciency, etc. Supplemental lines are planned in addition to the main lines if necessary. In the layout of the survey lines, the following points should be noted: 1. It is desirable to set up the survey lines on a site with minimal topographic change, and ideally lay the lines on a horizontal or gently sloping ground. 2. The survey lines should be laid out far from obstacles, especially articial metal structures such as metal fences and grounded power lines, because they may cause short circuits of the current and generate anomalous voltages around potential electrodes. 3. For 2-D proling, the survey line should be set up as perpendicular as possible to the strike of subsurface layers and faults. If the line cannot be set up in such way due to various restrictions in the survey area, supplemental lines should be arranged for delineating subsurface structure as correctly as possible. 4. In the layout of the survey line, the relationship between the depth of investigation and the survey line length should be taken into consideration (refer to Section 4.3.5). 5. Even for a survey along a curved line such as a tunnel route, it is desirable to arrange the survey line as straight as possible. 4.3.3. Selection of the survey method 1. The most appropriate method should be selected from vertical sounding, horizontal proling, 2-D proling and other available methods taking into consideration the scope and the stage of the survey. In the selection of the survey method, topographical and geological conditions, cost and operational efciency should also be taken into consideration. 2. The vertical sounding method is cost-effective for the investigation of horizontally stratied structure. The horizontal proling method is effective for mapping near-surface geological boundaries, buried objects, and fracture zones bearing groundwater. The 2-D proling method, combining vertical sounding and horizontal proling, is applied when the subsurface

geology is more complex. The three-dimensional (3-D) survey techniques, which have become practical in these years, can be applied at sites of very complicated geology, considering the cost and the special requirements. 4.3.4. Electrode conguration 1. Optimum electrode conguration is chosen in consideration of the method used, geological and surface conditions, and the desired output (see Table 1). 2. The Schlumberger array and the Wenner array are commonly used for vertical sounding. 3. The Wenner array, the polepole array, the pole dipole array and the dipoledipole array are suitable for horizontal proling. 4. The same electrode congurations as for horizontal proling are commonly employed for 2-D proling. 4.3.5. The depth of investigation and length of the survey line The depth of investigation of the resistivity method mostly depends on the electrode conguration and electrode spacing. As a rst approximation, the depth of investigation is a little smaller than the longest separation between the current electrodes and potential electrodes. A survey line should be much longer than the estimated depth of the survey target in order to cover the investigated area not only horizontally but also vertically. 4.3.6. Electrode spacing The maximum electrode spacing is determined from the target depth and the electrode conguration used. The minimum spacing is determined from the required spatial resolution. In horizontal proling and 2-D proling, requirements for both depth of investigation and resolution cannot be satised simultaneously. Therefore, they are usually determined bearing in mind the survey objectives, cost and operational efciency. 4.3.7. Remote electrode 1. Remote electrodes are required for the polepole array and the poledipole array. In order to improve the eld measurement efciency, it is suggested to select the location of the remote electrodes in considering access from the survey lines prior to the survey. 2. It is desirable to separate the remote electrodes from the survey line by at least 10 times the maximum electrode spacing. 4.4. Field operation 1. Field equipment for the measurements consists mainly of the resistivity meter (usually consisting of

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 Table 1 Electrode conguration factors for various electrode arrays Array type 2-electrodes Polepole Image point of apparent resistivity
C C1 a a P1 P

897

Electrode conguration factor 2pa

3-electrodes

CPP

C1 a

P1 a

P2 a

4pa

Poledipole

C1 na

P1 a

P2

2n(n+1)pa

a(n +0.5) 2

4-electrodes

Wenner

C1 a

P1 a

P2 a

C2

2pa
a

Dipoledipole

C1 a

C2 na

P2 a

P1

2n(n+1)(n+2)pa
a(n+1) 2

Schlumberger

C1 (A)

P1 (M) P2 (N) 1 L L/2

C2 (B)

p(L2l2)/4 l (L>5 l)

2.

3.

4.

5.

a power supply, a transmitter and a voltage receiver), electrodes, and cables. It is important to check whether the cables are not cut and there is proper connection between electrodes and the resistivity meter. Furthermore, it is recommended to check the ground contact resistance of every electrode. If the ground contact resistance of an electrode is extremely high, it should be lowered by inserting the electrode sticks more deeply into the ground, adding electrode sticks at the station, sprinkling water around the electrode sticks, or moving the electrode position slightly. Also, it is important to check whether any cable is directly leaking current into the ground. It is important to check repeatability of the measurements with the same electrode settings to ensure data quality. The amplitude of the measured voltage must be larger than the noise level of the resistivity meter. Another important issue concerns safety during the eld measurement. Since high-voltage current is applied to the cables and the current electrodes, it is very important to avoid any accident or injury. In the vertical sounding method, resistivity measurements are repeated using a set of electrodes with the

same midpoint but with different intervals. Since a larger electrode interval corresponds to a deeper penetration depth, a one-dimensional (1-D) resistivity model at the measurement location is obtained with the vertical sounding method. 6. In the horizontal proling method, a set of electrodes with a xed spacing is moved along the survey line to image the horizontal variation of subsurface resistivity at a similar penetration depth. 7. In the 2-D proling method, horizontal proling is repeated with various electrode spacings along the survey line. A modern resistivity meter controlled by a computer can automatically select a set of electrodes on the survey line, transmit a current and measure the potential data. 4.5. Data processing 1. Apparent resistivity, ra (O m), is estimated using the following equation for the measured electrical potential V (V) due to the injected electrical current I (A): ra GV =I ; where G is an electrode conguration factor which is calculated from the electrode spacing and

ARTICLE IN PRESS
898 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

Observed potential data

Set up an initial 2-D resistivity model

Calculate theoretical potential data

Update the 2-D resistivity model

Calculate residual (observed calculated) potential data

Residual is small ? YES

NO

Final 2-D resistivity model

Fig. 9. A standard inversion process of the 2-D resistivity proling data (SEG Japan, 2000) [1].

conguration shown in Table 1. Apparent resistivity is considered to represent average resistivity from the surface to the penetration depth. 2. For vertical sounding (VES), it is common to create an apparent resistivity versus electrode spacing curve (a VES curve; raa curve for the Wenner array) by drawing apparent resistivity values against the corresponding electrode spacing on a full logarithmic section paper. For horizontal proling, it is most common to create an apparent resistivity curve or section as a function of horizontal position. For 2-D proling, it is most common to create an apparent resistivity pseudo-section by taking the position as the horizontal axis and the electrode spacing as the vertical axis. 3. For vertical sounding, it is possible to estimate resistivity and depths of boundaries of subsurface formations from the VES curve. Data processing was previously conducted by the graphical analysis method (the curve tting method) using master curves. However, it is now commonly performed by 1-D inversion on a computer. 4. In 2-D proling, the data are inverted with a nonlinear inversion method to obtain a 2-D subsurface resistivity model. Fig. 9 shows a standard data processing ow of the 2-D resistivity proling data. 4.6. Interpretation 1. To avoid misinterpretation of the processed resistivity image, it is important to understand restrictions and limitations of the inverse method. It is recommended to refer to additional or other information about the survey site.

2. Resistivity varies with various factors such as rock type, porosity, water saturation, groundwater resistivity, clay content associated with weathering or alteration, and temperature. Lower resistivity generally indicates higher clay content or higher water content (i.e., the porosity multiplied by water saturation). 3. Geotechnical parameters such us hydraulic conductivity and porosity of the ground can be estimated from the resistivity values using empirical or experimental equations. 4. Fig. 10 shows an example of the 2-D resistivity image along a planned tunnel route in a granitic rock obtained with 2-D resistivity proling. Low resistivity portions were interpreted as high water content zones. 4.7. Output and report Outputs of an electrical (resistivity) survey should include at least the following: 1. location map; 2. survey line layout; 3. apparent resistivity curves, proles, or pseudo-sections; 4. resistivity models; 5. list of measured data. The survey report should describe at least the following items: 1. outline of the survey; 2. measurement method; 3. eld equipment;

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 899

Fig. 10. Example of a 2-D resistivity model obtained by a 2-D inversion (SEG Japan, 2000) [1].

4. eld notes describing the survey layouts and the conditions of the survey area; 5. data processing procedures; 6. apparent resistivity curves, proles, or pseudo-sections; 7. resistivity models; 8. interpretation and discussion of the resistivity models; 9. references.

5.3. Planning 5.3.1. Review of existing information 1. In order to apply the EM method most effectively, it is important to study the following items prior to the survey using existing information: J depth, size and electrical properties of the exploration target; J surface topography and available roads that affect operational efciency of the survey. 2. High-voltage power lines and man-made structures may generate noise in an EM survey. It is important to check potential noise sources in and around the survey area. 3. It is desired to examine the location of the xed EM transmitter for the CSAMT and Offset TEM measurements that may be set up outside the survey area. 5.3.2. Selecting the survey method First of all, the method used should be determined considering (a) the survey objectives and specications, (b) surface and subsurface conditions, and (c) operational cost and time. Both the CSAMT and TEM methods are suitable for deeper resistivity proling as compared to the electrical method. Since the TEM method can use non-contact sources and receivers such as a loop source and a magnetic coil, it is more suitable when a good contact of an electrode is difcult at a survey area of high-resistivity underground, for instance, snow accumulation and a rocky area. 5.3.3. Determining the survey lines and station interval 1. There are two types of measurement arrangements. One is the proling in which measurement stations are densely arrayed along a survey line. The other is the mapping in which measurement stations are

5. Electromagnetic 5.1. General The electromagnetic (EM) method is a geophysical exploration technique to delineate subsurface resistivity distribution similar to the electrical method. There are many techniques and variations in the EM method, and they can be applied for different purposes in civil engineering. Among them, these suggested methods describe the controlled-source audio-frequency magnetotelluric (CSAMT) method (Fig. 11) and the transient electromagnetic (TEM) method (Fig. 12), which are most often used. 5.2. Applicability 1. The EM method can image the subsurface resistivity distribution, which is similar to the electrical method. 2. As compared with the electrical method, better operational efciency and deeper penetration are main features of the EM method. It is applicable to site characterization for planned tunnels and dams, to detection of faults or fractured zones in rocks, etc. 3. The EM method is not suitable for the survey at an area near power lines, telephone lines, radio stations, factories, railways, power plants, power stations and urban area, because EM waves radiated from these constructions and metallic objects strongly interfere the EM measurements.

ARTICLE IN PRESS
900 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

Fig. 11. A schematic diagram of the CSAMT method (SEG Japan, 2000) [1].

Fig. 12. A schematic diagram of the TEM method (SEG Japan, 2000) [1].

spread over the survey area. The proling and the mapping can provide 2-D and 3-D resistivity images, respectively, through representing plural 1-D resistivity models or 2-D and 3-D data interpretation.

2. In the proling, the survey line should be set up as perpendicular to the strike of geological structure to be investigated. In the investigation for a linear construction such as a tunnel and a bridge, the survey

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 901

lines are usually set up along the planned construction route. When the survey line is set up parallel to the subsurface geological structure, some complementary survey lines should be arranged perpendicular to the main survey line to obtain more reliable subsurface images. 3. The length and linearity of survey lines and station intervals do not affect very much the penetration depth and the quality of the data unlike the electrical method. Therefore, planned stations can be moved if there are unexpected obstacles or noise sources close to them. 5.3.4. Determining penetration depth The penetration depth depends on the frequency used for the measurement and on the resistivity of the ground. Therefore, the frequency range used in the measurement should be determined considering the survey objectives, the depth of investigation and the expected resistivity of the ground. The penetration depth in frequency domain is usually referred to the skin depth of the plane EM wave. Fig. 13 shows the relationship between the skin depth and frequency used in the CSAMT method. The penetration depth in time domain is usually referred to a depth where a magnetic (or electric) eld from a plane impulse source takes the maximum value at a certain time. Fig. 14 shows the penetration depth as a function of time, which is used in the TEM method. 5.3.5. Setting the remote EM transmitter For deeper CSAMT and TEM surveys, a large xed transmitter (usually a grounded wire) is used as an EM

source (Figs. 11 and 12). The large transmitter should be set up at a distance of more than 35 times the skin depth from the receivers to avoid the near eld effect in the CSAMT survey. The location of the transmitter should be planned prior to the survey considering the penetration depth (the target depth) and surface conditions at the site. 5.4. Field operation 5.4.1. Equipment 1. The EM equipment consists of a transmitter, a receiver, sensors, electrodes and cables. The transmitter and receiver are special equipments designed for the methods. They should have specications that meet the requirements in accuracy and reliability for the measurement. 5.4.2. Preparation 1. Sensors and electrodes at receiving stations should be laid out on as at ground surface as possible and far away from EM noise sources such as power lines, metal fences and existing constructions. 2. For the xed transmitter source, electrode sticks or plates should be installed into the ground very rmly such that the contact resistance is less than 100 O. Since the transmitter is usually kept on the ground for a long time until the measurement is completed, it should be kept safe from accidents such as electrical shock and leakage and cutting of cables.

Fig. 13. Relationship between the skin depth of a plane wave EM eld and frequency in the CSAMT method (SEG Japan, 2000) [1].

ARTICLE IN PRESS
902 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

Fig. 14. Relationship between the penetration depth of a plane EM impulse and time in the TEM method (SEG Japan, 2000) [1].

5.4.3. Measurements 1. In the CSAMT measurement, a pair of a magnetic eld and an electric eld, which are perpendicular to each other and both horizontal, are usually measured at a station in a frequency range from a few Hz to several kHz. In the TEM measurement, a vertical magnetic eld is usually measured at a station with transient time from several microseconds to several hundreds milliseconds depending on the conguration of the measurement. 2. It should be reconrmed prior to starting the measurement that the measurement parameters are properly set up on the equipment. During the measurement, the measured data are always monitored in order to check the data quality. To ensure the reliability of the data, repeatability of the measurements should be examined. If the measured data are not stable, measurement parameters should be checked and changed if necessary. Ambient noise should also be checked. 3. During and immediately after the measurement, it is important to check the quality of the acquired data by plotting the apparent resistivity curves or voltage decay curves on site. 5.5. Data processing 1. Fig. 15 schematically illustrates a standard data processing ow of the CSAMT and TEM data. First of all, apparent resistivity values calculated

from measured data are plotted against frequency or time. Then apparent resistivity sections or maps are made as fundamental data for further processing. These apparent resistivity sections or maps are sometimes presented as the nal exploration results. 2. The apparent resistivity curves are usually used for a layered model inversion to obtain 1-D resistivity model beneath each measurement station. 3. When more detailed subsurface resistivity images are required, 2-D inversion using nite difference or nite element methods is applied to the measured data for obtaining 2-D resistivity models. 5.6. Interpretation 1. Interpretation of the CSAMT and TEM data is usually conducted with 1-D resistivity inversion to obtain layered earth models. A 2-D image is obtained by presenting plural 1-D resistivity models along a measurement line. Fig. 16 shows such an example. 2. Any inversion technique has a non-uniqueness problem. The inversion of the EM data has nonuniqueness due to what is called the equivalent resistivity layer. To avoid misinterpretation of the inversion results, a number of inversion trials with different a priori information should be conducted.

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 903

CSAMT
Survey line

TEM

Topographic section

Calculation of apparent resistivity

Apparent resistivity Apparent resistivity

CSAMT

TEM

Measuring points survey line

Survey line

Frequency f

Time t
Apparent resistivity CSAMT: Each frequency TEM: Each time

Apparent resistivity

Apparent resistivity curve

1-D interpretation CSAMT TEM

Surface Layer 1 Resistivity n1 Layer 2 Resistivity n2 Depth d1 Depth dn-1 Layer n Resistivity n n Depth dn
Vertical section of stitched 1-D model

Survey line

Multi-layer structure

Horizontal section of 1-D model Each depth or elevation

multi-layer model

Obtain the resistivity of each cell Surface Depth d1 Depth d2 Cell m-1

2-D interpretation
Survey line

Survey line

Cell m+1 Cell m

2D resistivity model

Cell n-1

Cell n

Cell n+1

Fig. 15. A schematic diagram of a standard data processing ow of the EM data (SEG Japan, 2000) [1].

5.7. Output and report Outputs of an EM survey should include at least the following: 1. location map of the survey site; 2. layout of survey lines and measurement stations; 3. apparent resistivity curves, voltage decay curves, proles, or pseudo-sections; 4. resistivity models; 5. list of measured data. The survey report should describe at least the following items: 1. outline of the survey; 2. eld operations (including equipments used); 3. eld notes describing the survey layouts and the conditions of the survey area; 4. data processing procedures;

5. apparent resistivity curves, voltage decay curves, proles, or pseudo-sections; 6. resistivity models; 7. interpretation and discussion of the resistivity models; 8. references.

6. Ground penetrating radar 6.1. General GPR is a geophysical method that uses the transmission of high-frequency EM waves for detecting subsurface objects. The operating frequency is higher than several MHz. GPR can be used on the ground surface and in boreholes. Only surface GPR is described in this document. In a GPR survey, transmitting and receiving antennae are placed on the ground surface and EM waves are injected into the ground from the transmitter

ARTICLE IN PRESS
904 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

Fig. 16. Examples of stitched 1-D resistivity model obtained with the CSAMT method at a survey area of a tunnel construction (SEG Japan, 2000) [1].

Fig. 17. A schematic diagram of the surface GPR survey (SEG Japan, 2000) [1].

antenna. Reected EM waves from subsurface geological boundaries or objects are then received by the receiving antenna and processed to image these subsurface structures (Fig. 17). 6.2. Applicability GPR is a non-destructive and non-invasive geophysical method for rapid and high-resolution imaging of subsurface objects. The penetration depth is typically several metres in soil ground. There are a variety of applications including detection of buried pipes, locating reinforcing bars in concrete, mapping voids beneath road pavements or behind tunnel linings, and inspection of roads and concrete structures. The method can also be used in archaeological investigations, monitoring the spread of contamination in the ground, and mapping faults and fracture zones in rocks. 6.3. Planning 6.3.1. Review of existing information The feasibility and the output of GPR can often be predicted by studying existing site information and by

site reconnaissance. The survey plan should be made in consideration of expected properties of subsurface materials (especially resistivity), groundwater condition, depth and characteristics of the targets, and accessibility of the equipments to the survey areas. 6.3.2. Choice of GPR antenna The centre frequency of the GPR antenna typically ranges from 30 MHz to 1.5 GHz. Resolution and penetration of GPR waves strongly depend on the antenna frequency used, as shown in Table 2, i.e., higher-frequency antenna can have higher resolution, but lower penetration. The antenna frequency used in the survey should be determined considering the following conditions: (1) Features of exploration targets such as size, depth and material type, and resolution required. (2) Properties of soil, rock and medium in which the target is buried. (3) Moisture and clay contents of the medium, which affect GPR penetration. (4) Surface unevenness and vegetation of the site and accessibility of the antenna.

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 Table 2 Relationship of GPR parameters with respect to dielectric constant, electrical conductivity of the ground, and antenna frequency GPR parameters Dielectrical constant Low Propagation velocity Attenuation Penetration Wavelength Resolution Fast High Short Long Low High Slow Low Long Short High Electrical conductivity Low Low Long High High Short Antenna frequency Low Low Long Long Low High High Short Short High Attenuation is inuenced strongly by electrical conductivity The lower the attenuation, the greater the penetration distance The shorter the wavelength, the higher the resolution Remarks 905

6.3.3. Arrangement for survey lines and their spacings For linear targets, such as buried pipes, the survey line is normally set up perpendicularly to the longitudinal direction of the expected buried pipes. The survey line spacing should be determined considering the target type, size and depth. Typically it is 12 m. For metallic targets (e.g., iron pipes), the antenna dipoles are usually oriented parallel to the longitudinal axis of the target. For non-metallic targets (e.g., gas pipes), the antenna dipoles are normally set up at the right angle to the target orientation. The layout of survey line(s) and the line spacing (s) should be determined in consideration of the following conditions: (1) (2) (3) (4) frequency and moving speed of the antenna; resolution required; surface unevenness and vegetation of the site; cost and acquisition time available.

Selection of antenna frequency is crucial for success of the survey. 6.4.2. Preparation for the survey 1. The locations of the survey lines are marked out at the site. The start and end positions of the survey lines are normally indicated in the survey plan. 2. For a rough and uneven ground, land readjustment may be needed to obtain good contact between the antenna and the ground surface. 3. The antennae are mounted on a sled or a wheeled trolley to enable them to move smoothly on the ground. 6.4.3. Field procedures 1. The GPR data acquisition system is rst set up and connected to a power supply such as a 12 V DC battery. It is recommended to warm it up for a while to stabilize the zero time position. A preliminary test measurement should be conducted prior to the survey in order to ensure that the GPR system is functioning properly and to optimize the measurement parameters. The eld acquisition parameters such as the recording time range (in nanoseconds), the sampling rate, signal gains and band-pass lters should be optimally selected in consideration of the target depth and the antenna frequency. The measuring wheel should be calibrated on site for a known distance. 2. The measured data should be monitored on the screen of the equipment during the survey for data QC. Unwanted noise or anomalous signals should be identied to remove their causes. 3. The recording time range should be adjusted for the investigation depth by measuring the propagation velocity of the EM waves in the ground using the wide-angle or the arrayed antenna measurements or by analysis of the hyperbolic diffractions generated by a discrete target. Alternatively, the velocity can be calculated from the dielectric constant of the ground, which can be measured on site using a portable

Additional survey lines or a dense grid-based survey may be required to detect smaller and multiple targets or if very accurate data processing is necessary. 6.3.4. The depth of investigation The depth of investigation of a GPR survey depends on the frequency and the power of transmitted EM waves and electrical conductivity of the ground or the medium to be investigated. The depth of investigation varies from 0.1 to 100 m, but typically from 0 to 5 m for most of civil engineering applications. Deeper penetration is possible with lower frequencies (30100 MHz) in the ground that has lower conductivity. 6.4. Field operation 6.4.1. Equipment The GPR equipment consists of a data acquisition system and antennae. Those should be chosen in consideration of the depth of the target and the accuracy required with reference to the objectives of the survey.

ARTICLE IN PRESS
906 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

dielectric meter or determined in the laboratory test using soil samples. 6.4.4. QC of the measured data The quality of the recorded data should be checked on site during or immediately after the survey by monitoring the GPR signals or proles on the screen of the equipment. Survey information, such as the positions of the survey lines, variations in surface conditions, and locations of metallic objects close to the survey line causing spurious reections, should be recorded in the eld observers notes for data processing and interpretation later on. 6.5. Data processing 1. The GPR prole (or Radargramme) is normally represented as a 2-D time section (or a depth section if velocity is known) with wiggle traces or grey/colour scales. 2. To enhance the target signals, the following digital signal processing is usually applied to the measured data on site or in house after the survey: vertical stacking, deconvolution lter, bandpass lter, moving average/background removal lter, automatic gain control (AGC) or time-variant gain (TVG). 3. In case that the data are recorded on multiple survey lines on a grid, the GPR data can be displayed in three dimensions using commercial presentation software for making its interpretation much easier. 6.6. Interpretation 1. A GPR prole can be interpreted in a similar way to the interpretation of a seismic reection prole. Subsurface anomalies such as buried pipes and cavities can be detected as diffraction patterns in the prole. 2. It is important to collect as much existing information as possible to interpret the GPR prole properly. For example, in the application of the GPR to detection of buried pipes, information relating to their locations, depths, sizes and materials should be searched prior to the interpretation. 3. If an accurate depth section is necessary, information affecting the EM wave velocity such as dielectric constants and moisture contents of the ground should be considered in the interpretation. 4. If other geophysical data such as shallow EM mapping, magnetic or resistivity data are available, they can be very useful for interpretation of the GPR prole. 5. Fig. 18 shows some examples of interpreted GPR proles.

6.7. Output and report Output of a GPR survey should include at least the following: 1. location map of the measurement lines; 2. GPR proles; 3. interpreted GPR proles. The survey report should describe at least the following: 1. 2. 3. 4. 5. 6. outline of the survey; eld operation (including the equipments used); data processing procedures; GPR proles; interpretation and discussion of the proles; references.

7. Gravity 7.1. General The gravity method is a geophysical method in which the subsurface density distribution (or the density anomaly) is estimated from the observed gravity eld. Gravity is usually measured at the ground surface using a gravity meter or a gravimeter. Gravity may also be measured in the air, at sea, or in a borehole using gravimeters designed for those environments. In this document, we will focus on and describe in detail the land gravity method. The gravity method has been widely used for delineating a base rock for earthquake resistant construction design, locating buried faults and detecting underground cavities due to civil engineering activities and old mine working. Fig. 19 shows the schematic diagram of a gravity anomaly and an example of a gravimeter. The gravity method has normally been used for large-scale subsurface imaging, for instance, for studies in Earth Sciences. The micro-gravity method using very dense measurements has recently been employed for near-surface civil engineering applications. 7.2. Applicability 1. The gravity method can be used for: (1) detection of subsurface cavities and voids due to civil engineering activities and old mine working, (2) delineation of a base rock for earthquakeresistant construction design, (3) delineation of buried faults underneath thick sediments, (4) evaluation of ground improvement such as grouting and compaction by comparing gravity anomalies measured before and after the improvement.

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 907

Fig. 18. Examples of interpreted GPR proles (SEG Japan, 2000) [1].

2. In earthquake engineering eld, the gravity method can often be employed to delineate larger-scale subsurface structures such as deeper base rocks and large faults to predict accurate earthquake ground motion. 3. The surface gravity method is used to detect lateral density changes in the ground, but in principle, it cannot be used to delineate horizontally stratied structure whose density changes only with depth. 7.3. Planning 7.3.1. Review of existing information Existing information should be studied from the following points of view: (1) dimension, depth and density of the geological or structural features to be investigated, (2) surface topography around the survey area and (3) accessibility to the site such as road conditions. 7.3.2. Resolution of detectable subsurface structures or anomalies 1. Resolution of detectable subsurface structures or anomalies with the micro-gravity method can be

around 1 m using high-density measurements under relatively low ambient noise. 2. Using more widely spaced measurement points in a very wide area, it is possible to delineate large-scale subsurface structure with the dimension of several kilometres. 3. Fig. 20 shows the relationship between the detectable dimension of the target and the amount of the gravity anomaly due to it. Since the most modern gravimeter can measure the gravity change of several microGals, subsurface features with the dimension of around 1 m can be detected by means of the microgravity method. 7.3.3. Determination of the survey area and the station spacing 1. The micro-gravity survey will normally be conducted in one of two modes, a prole and an areal survey. The former is commonly employed for investigation along the linear construction such as a railway and a road, while the latter would be used for an areal survey, for instance, for delineation of a bed rock topography beneath the investigation site 2. The survey area should be determined considering the depth of the features of interest in the survey. The

ARTICLE IN PRESS
908 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

Fig. 19. Schematic diagram of the gravity anomaly and a gravimeter (SEG Japan, 2000) [1].

Fig. 20. Relationship between dimension of the target and gravity anomaly (SEG Japan, 2000) [1].

dimension of the survey area should be 5 or 6 times the maximum depth of investigation. The station spacing should be comparable to, or smaller than, the minimum depth of the features of interest. 3. It is desirable to keep the station spacing as uniform as possible. 7.4. Field operation 7.4.1. Measurement principle 1. A gravimeter commonly used for a gravity survey can only measure relative gravity value. Therefore, in a

gravity survey, the difference of the gravity value between a station and a xed station called as a base or reference station is measured. The reference station should be either a station where the absolute gravity value is known or a station connected to another reference station where the absolute gravity value is known. 2. A measurement technique called the closed-loop method is commonly used, in which the rst and last measurements of successive measurements in a loop are made at the same station to obtain the error between these two measurements, known as the

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 909

closure error. The measurement error at each station can then be evaluated based on this closure error in the loop. Successive measurements in a loop should be normally completed within no more than about half a day to keep the error to a minimum. The closure error, mainly due to the drift of the spring in the gravimeter, should be less than 0.2 mGal. 3. In a very high accurate survey, such as detection of a very small cavity with a diameter of less than 1 m, it should be necessary to know the drift error as precisely as possible. In such a case, the successive measurements in a loop should be completed within 2 h or less and the measurements be repeated by visiting stations in the loop in a reverse order. 7.4.2. Measurement accuracy 1. A gravimeter used for a micro-gravity survey should have an accuracy of a few micro-Gals, while that for an ordinary gravity survey may have an accuracy of better than approximately 0.1 mGal. A gravimeter is a very sensitive device and must be handled with extreme care during transportation as well as during measurements. 2. The elevation of each measurement station in a micro-gravity survey must be known to the accuracy of a few millimetres or less and the accuracy of the horizontal location is required within a few tens of centimetres. For ordinary gravity survey, accuracies of the elevation and the horizontal location should be less than a few tens of centimetres and a few metres, respectively. 3. In a gravity survey over an area more than a few kilometres squared, the location of each station can be determined using a geographical map with a scale of 1:1000 to 1:25 000 corresponding to the required accuracy of the survey. Benchmarks, trig stations or other points with elevations indicated on the map can be used as reference elevation points. 7.4.3. Measurement procedures 1. To make the measurement procedure most efcient, the route and order of visits to the measurement stations should be determined at the beginning of the survey. 2. Measurement stations should be arranged at a at place without abrupt topographic changes nearby and not placed on the top or at the foot of a cliff or an embankment. 3. The measurement should be done on a rm and at surface such as a pavement and away from sources of vibrations such as trafc noise. 4. The exact station location should be marked with paint, a survey nail or a survey peg prior to the survey.

5. It is desirable to repeat measurements several times at each station to improve the accuracy and reliability of the measurement. 6. The following items should be recorded in the eld note: (1) the model name and serial number of the gravimeter used; (2) the height of the instrument from the earths surface at each measurement point; (3) each reading and its time. 7. It is desirable to make a sketch of the measurement location and its vicinity for the further terrain correction in data processing. 8. Re-measurement of any loops should be considered if the closure error is neither negligible nor acceptable. 7.5. Data processing 1. The measured gravity value at each station should be rst corrected for the gravity tide variations, the gravimeter drift and the instrument height at the measurement station. Usually, these corrections are conducted in the eld for QC. 2. The gravity anomaly value at each station is computed by applying latitude corrections, free-air corrections, Bouguer corrections, terrain corrections and atmospheric corrections to the eld data corrected as above. The coordinate system used for these corrections may be either the geodetic reference system (latitude, longitude, height) or a Cartesian system X ; Y ; Z : 3. Fig. 21 shows a standard data processing ow of gravity data. The gravity anomaly thus obtained is the subsurface density anomaly and is the object of further interpretation. 7.6. Interpretation 1. For qualitative interpretation of the subsurface density distribution, the residual gravity anomaly map is usually used. It is computed by applying the trend analysis and/or spatial lters to the original gravity anomaly data to enhance the anomalies to be extracted. 2. Depending on the requirements or the objectives of the investigation, quantitative interpretation can be used to delineate bedrock topography, to calculate the location or the depth of a cavity, or to estimate the vertical displacement of a fault. 3. The interpretation may be difcult or unreliable near the boundary of the survey area. Such problems might be avoided by integrating existing gravity data, if available, into the measured data.

ARTICLE IN PRESS
910 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

Measured Gravity Data

Correction for solar and lunar mass attraction Gravity tide correction

Correction for instrumental height

Correction for altitude from a ground surface to the sense position of a gravity meter

Drift correction

Correction for the drift of gravity meter

Gravity value at measurement point

Topographic correction

Correction for near field topographic mass attraction

Bouger correction

Correction for near field topographic mass attraction

Free-air correction

Correction for gravity difference due to the altitude difference from measurement point

Latitude correction

Correction for the normal gravity

Atmospheric correction

Correction for atmospheric mass attraction

Gravity anomaly distribution (on irregular grid)

Data interpolation in space

Transformation of irregular to regular grid data

Gravity anomaly distribution (on regular grid)

Fourier analysis (model calculation)

To obtain the target underground density structure based on spectral analysis, filter operation and model calculation

Subsurface density distribution

Fig. 21. Standard data processing ow of gravity data (SEG Japan, 2000) [1].

4. Since the gravity anomaly represents the subsurface density anomalies (density variations), the density contrast between the survey target and its surroundings should be taken into account in interpreting the survey result. 5. If available, it is important to integrate existing information into the interpretation, such as subsurface geology, geophysical images and topography around the site.

6. Fig. 22 shows an example of an interpreted gravity anomaly map for detection of subsurface cavities. 7.7. Output and report Outputs of a gravity survey should include at least the following: 1. maps of the station locations and the survey area;

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 911

2. gravity (and/or gravity anomaly) maps; 3. interpreted gravity maps; 4. measurement data. The survey report should describe at least the following items: 1. 2. 3. 4. 5. outline of the survey; eld operations (including the equipment used); data processing procedures; gravity anomaly maps; interpretation and discussion of the gravity anomaly maps; 6. references.

8. Radiometric 8.1. General The radiometric method can be used to detect subsurface geological features such as a fault or a shear

zone by measuring changes in intensity of natural radiation from the earths surface. Gamma rays are generally used in the radiometric survey for geological applications, and the term radiometric survey usually therefore refers to the gamma-ray spectrometric survey unless explicitly stated otherwise. Fig. 23 shows the schematic diagram of the gamma-ray spectrometric survey. Two types of measurement methods can be used in the gamma-ray survey: the total count method and the gamma-ray spectrometric method. The total count method measures gamma rays at all energies with no distinction between radionuclides, whereas the gammaray spectrometric method measures both the intensity and energy of radiation and can distinguish between radionuclides. The gamma-ray spectrometric method is preferred because it can provide more information than the total count method. The survey techniques are classied based on the survey operation mode into air-borne, car-borne and man-borne. The survey equipments are mounted in a helicopter or a xed-wing aircraft for the air-borne survey and in a motor vehicle for the car-borne survey. For the man-borne survey (or the portable gamma-ray survey), the operator carries the survey equipments. 8.2. Applicability 1. For civil engineering applications, the radiometric survey is applied mainly to detection of subsurface geological features such as faults or shear zones, which are often conducted for characterization of active faults in earthquake disaster prevention projects and for groundwater or hot spring developments. 2. The radiometric method is also used to investigate the distribution of rocks at the earths surface by measuring changes in gamma-ray intensity due to differences in the natural abundance of the radioisotopes in rock forming minerals.

Fig. 22. Example of an interpreted gravity anomaly map for detection of cavities (SEG Japan, 2000) [1].

Fig. 23. Schematic diagram of the radiometric method (radionuclide: Bi: bismuth, K: potassium, Tl: thallium) (SEG Japan, 2000) [1].

ARTICLE IN PRESS
912 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

8.3. Planning 8.3.1. Review of existing information Before planning a survey, all available technical references and investigation reports regarding the site should be thoroughly studied. The selection of the survey area, the survey method and the locations of measurement points or lines should be made after due consideration of the purpose of the survey and the survey targets as well as the existing information above. It is also essential to preliminarily study the geological distribution of the radioelements, the locations of potential fault zones and any possible noise sources such as man-made structures and unfavourable topographic conditions. 8.3.2. Selecting the survey method The survey method should be selected from among the air-borne, the car-borne and the man-borne method in consideration of survey objectives and requirements, time and cost of the survey, etc. The air-borne method is suitable for a very wide area survey within a short time and for a survey of the area where it is difcult to access. The car-borne is suitable for a wide area survey and the man-borne is suitable for a limited area survey. A typical application of the car-borne method would be a preliminary investigation for the selection of a further detailed investigation area in a large area. The manborne survey is, for example, useful for detection of accurate location of a buried fault. 8.3.3. Designing the survey A survey design consists of positioning the survey lines and the measurement points. The survey should be designed after due consideration of the survey objectives and targets. Additionally, the required accuracy, geological and topographic conditions and the locations of the man-made structures should be also considered. For locating a buried fault, the survey line should be set up as perpendicularly as possible to the expected strike of the fault. If there is no information about geological features of the site, the grid-based measurement lines should be set up for more accurate estimation of them. Survey lines and measurement points should be set up so as to avoid the area with abrupt topography changes and man-made constructions. 8.4. Field operation 8.4.1. Equipment The measuring equipment used in the radiometric method is generally a scintillation counter with a NaI (Tl) detector. There are two types of equipment, the total count meter and the spectrometer. The total count meter measures gamma rays at all energies with

no distinction between radionuclides, whereas the spectrometer measures both intensity and energy of radiation and can distinguish between radionuclides. The latter is commonly used in a recent radiometric survey. 8.4.2. Stationary measurement at a xed point In order to correct for time variations of gamma-ray radiation during the survey, stationary measurement at a xed point should be taken at least once or twice a day. In case of the survey over more than 1 day, stationary measurement at a xed point must be taken every day in order to correct for daily changes in the measurement. 8.4.3. Measurement 1. For the air-borne survey, the special gamma-ray detector designed for an aircraft is used. The helicopter is more preferable to civil engineering applications than the xed-wing aircraft because a helicopter can y at a lower altitude and lower speed than a xed-wing aircraft, which leads to a higherresolution survey. In the air-borne survey, navigation and altitude data must be recorded together with the gamma-ray measurements. 2. For the car-borne survey, the speed of the vehicle should be kept as slow and steady as possible to ensure good quality data. The vehicle speed should be determined considering time and accuracy of measurement. 3. For the man-borne survey, measurement should be conducted normally for 300500 s at a measurement point. The intensity of radiation due to each of major radionuclides should be recorded in the eld note for each measurement. 4. Data QC should be made during or immediately after the measurement on site. Additional or repeated measurements should be taken if necessary. 8.5. Data processing 1. The total count method provides the total gammaray intensity at each measurement point. The gammaray spectrometric method provides the gamma-ray intensity spectrum at each point. From this spectrum, the total intensity and the spectral intensities of three indexed radionuclides, 40 K (1.46 MeV), 214 Bi (1.76 MeV) and 208 Tl (2.61 MeV) can be calculated. These measurement values are compiled in a table for each survey line or measurement point. 2. The measured value should be corrected for time variation and daily change of gamma-ray radiation based on the measurement values at the xed point. If possible, climatic or surface topographic corrections should be applied. For the air-borne survey data,

ARTICLE IN PRESS
T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914 913

Fig. 24. Example of interpretation of radiometric survey proles (SEG Japan, 2000) [1].

attenuation correction must be made for removing effects of altitude changes. 3. The spatial average and standard deviation of the total gamma-ray intensities and the spectral intensities of 40 K (1.46 MeV), 214 Bi (1.76 MeV) and 208 Tl (2.61 MeV) are calculated at rst. Anomalous intensities are then detected on the basis of the mean value and the standard deviation. For the spectrometric method, the spectral ratios such as Bi/K, Tl/K and Bi/Tl are calculated. For the line survey, these calculated values are presented as a prole with distance. In case of the grid-based measurement, the calculated total or spectral intensities are normally presented as a contour map of them.

8.7. Output and report Outputs of a radiometric survey should include at least the following: 1. location map of the measurement lines or points; 2. gamma-ray total or spectral intensity or spectral ratio prole or map; 3. interpreted prole or map; 4. measurement data listed in a table. The survey report should describe at least the following: 1. 2. 3. 4. outline of the survey; eld operation (including the equipment used); data processing; gamma-ray total or spectral intensity or spectral ratio prole or map; 5. interpretation and discussion of the results; 6. references.

8.6. Interpretation The calculated total or spectral gamma-ray intensity prole maps are interpreted in terms of the survey target and objectives by referring to existing information such as a geological map. If there are other geophysical data, they can be helpful for the interpretation. Fig. 24 shows an example of interpretation of gamma-ray total intensity and spectral ratio proles obtained at a radiometric survey for detection of faults.

Acknowledgements The suggested methods were drafted by the Working Group on Standardization of Geophysical Methods for Rock Engineering, consisting of 11 members from six countries: P. Hatherly, M.S. King, H. Kusumi, B. Lehmann, L. Myer, L. Pyrak-Nolte, T. Takahashi,

ARTICLE IN PRESS
914 T. Takahashi / International Journal of Rock Mechanics & Mining Sciences 41 (2004) 885914

N.M. Toksoz, T. Uchida, T. Watanabe and C. Xu. Toru Takahashi has coordinated the working group. Major contributions in drafting and revision were received from Dr. Peter Hatherly (Australia) and Toshihiro Uchida (Japan). We would like to express special thanks to these two members for their efforts. The drafts were also greatly improved by fruitful comments and suggestions from E. Brueckl (Austria), C. Cosma (Finland), B. Lehmann (Germany), G. Borm (Germany), K. Sassa (Japan), C. Lee (Korea) and M.S. King (UK) who are the ISRM commission members. The rst drafts of the suggested methods were drafted by the sub-working group consisting of following 19 members belonging to the Japanese Committee for Rock Mechanics (JCRM) working group on Applied Geophysics for Rock Engineering: T. Aizawa, Y. Ashida, A. Chiba, K. Chida, K. Ikeda, K. Kishida, N. Kurahashi, H. Kusumi, T. Matsuoka, S. Miki, H. Suzuki, K. Suzuki, S. Tanaka, T. Tanaka, T. Takahashi, T Takeuchi, T.

Tokumaru, F. Watanabe and T. Watanabe. We would also like to thank these members for their efforts.

References
[1] SEG Japan. The guidebook of geophysical exploration methods for civil engineering. Japan: SEG Japan; 2000 [in Japanese].

Further reading Suggested Readings for details of each geophysical method

Telford WM, Geldart LP, Sheriff RE. Applied geophysics. 2nd ed. Cambridge: Cambridge University Press, 1990. Burger HR. Exploration geophysics of the shallow subsurface. Englewood Cliffs, NJ: Prentice-Hall, 1992.