G : Ground-penetrating radar (GPR)

Transcription

1 G : Ground-penetrating radar (GPR) The EM methods described in GEOP424 use low frequency signals that travel in the Earth by diffusion. These methods can image resistivity of the Earth on various scales. Because they are diffusive, they will not be reflected from sharp interfaces between layers with different resistivities. Seismic exploration is the most widely used exploration method in geophysics. It uses the propagation of elastic waves in the Earth and can give detailed images of the subsurface. Can show that the best vertical resolution that can be obtained in one quarter of a wavelength. However it can be difficult to apply seismics to the study of shallow structure on the scale of 10 s of metres as high frequency signals are needed. In addition, seismic sources can be awkward to operate repeatedly. RADAR (Radio detection and ranging) was developed during WW2 to measure the distance and velocity of an aircraft. From In the 1960 s and 1970 s radar was applied to imaging subsurface structure. Frequency must be chosen carefully to ensure that (a) signals travel as waves and (b) that the depth of penetration is not limited by the skin depth phenomena. Ground penetrating radar (GPR) is now a widely used technique for imaging near surface structure. The radio wave pulses can be transmitted much more easily than in seismic exploration. Most seismic data processing techniques can be easily applied to GPR data. Calculation What will be the round-trip travel time for a reflection from a surface 5 m from the TX-RX? Assume speed of light is 3 x 10 8 m/s. Compare this with sound waves travelling at 300 m/s 1

2 G1 : Wave propagation In section C we considered how to determine if an EM signal will travel in the Earth as a wave or by diffusion. To determine if wave propagation or diffusion will dominate, can consider the σ ratio defined as r = 2πfε, where f is the frequency, µ is the permittivity of the subsurface and à is the electrical conductivity. Assume that the permittivity has the free space value µ = µ 0 = F/m If r is large, then EM diffusion will occur. If r is small then wave propagation occurs G1.1 : Velocity of propagation 1 In free space radio waves travel at the speed of light, c, where c = µ 0ε 0 and µ 0 is the permeability of free space and µ 0 is the permittivity of free space µ 0 = 4À x 10-7 H/m and µ 0 = 8.85 x F/m giving c = 3 x10 8 m/s In the Earth the permeability (µ) and permittivity (µ) do not have their free space values. This is because the atoms/molecules behave as magnetic and/or electric dipoles. The magnetic behaviour of the atoms/molecules can be expressed in terms of relative permeability (µ r ) as µ = µ r µ 0 and section C describes this topic. The relative permeability is usually close to 1, except in deposits of magnetic minerals. The relative permittivity (µ r ) is defined as ε = ε rε 0. This quantity is also called the dielectric constant (K). The permittivity will be greater than 1 if the molecules can act as electric dipoles. Water is one of the most polar molecules since the hydrogen atoms develop a positive charge and the oxygen develops a negative charge. 2

3 Can write the velocity of a radar wave in the Earth (v) in terms of the relative permittivity and permeability as c v = µ ε r Some typical values are listed below (from Davis and Annan, 1989) r Material µ r Ã(mS/m) v(m/ns) (f = 100 MHz) Air Distilled water m Fresh water m Seawater m Dry sand m Saturated sand m Limestone m Clay m Granite m Dry salt m Ice m Most materials are mixtures so need to consider a combination of properties. Topp et al., 1980 derived an empirical relationship between the dielectric constant (K a ) and the water content ( v). Figure 5 from this paper is illustrated below. 3

4 K a v = θ = θ v + 146θ v 76.7θ v x x10 K a 5.5x10 K a + 4.3x10 K a Calculation A 200 MHz GPR signal travels in the Earth with v = 0.06 m/ns. What is the wavelength? G1.2 Reflection at interfaces When a radar pulse reaches an interface between two layers, some energy will be reflected and some will be transmitted. Consider the case when the signal is normally incident on a horizontal interface In the upper layer the velocity is v 1 and in the lower layer it is v 2 If the incident wave has an amplitude A =1, then can show that the amplitude of the reflected and transmitted wave is given by the coefficients R and T. R = Ar A i v = v 2 2 v + v 1 1 T = A A t i 2v1 = v + v 2 1 If the velocity variation is solely due to changes in permittivity, then R = ε 1 ε + 1 ε ε 2 2 T = 2 1 ε ε + 2 ε 2 4

5 Example 1 : Increase in velocity with depth This example has an increase in velocity from 0.06 to 0.15 m/ns which would correspond to water saturated sediments overlying granite. Note the positive reflection coefficient as predicted by v R = v 2 2 v + v 1 1 Confirm that the round trip travel time is correctly computed. 5

6 Example 2 : Decrease in velocity with depth Velocity decreases from 0.15 to 0.06 m/ns corresponding to the water table Note the decrease in velocity that corresponds to the increase in permittivity Reflection has negative polarity as predicted by v R = v 2 2 v + v 1 1 6

7 Example 3 : Low velocity layer Low velocity layer produced by a zone of water saturated sand. The 3 layers have velocities v 1, v 2 and v 3 Note the negative polarity reflection from the top of the layer which is caused by v2 v1 a decrease in velocity according to R = v + v The second reflection comes from the base of the low velocity layer and has a positive polarity. To calculate the amplitude must consider both transmission at the top of the layer and reflection at the base. 2 1 Amplitude = 2v R = v2 + v v v 3 v + v 2 v v2 + v 2 Note that the middle term can be positive or negative. However the first and third terms (transmission on downward and upward paths) are both positive. Thus the polarity of the second reflection is determined by the velocity change where it is reflected. 7

8 Can you confirm the relative magnitude of these two reflections for the numerical values used in this case? In the animation you can also see reverberation (multiple internal reflections) in the low velocity layer. As the layer gets thinner, the reflections arrive closer in time. Can show that the λ two reflections can only be distinguished if t > where t is the thickness of the 4 layer. Calculation A 100 MHz survey takes place in a region where dry sand has v = 0.15 m/ns. What is the thinnest layer that could be detected? G1.3 Factors that reduce the amplitude of radar waves The amplitude of radar waves will be reduced by two primary effects (a) Geometric spreading. As the wave spreads out to cover a larger area, conservation of energy requires that the amplitude decreases. This is very similar to the way that ripples on a pond decrease in amplitude. (b) Attenuation through the skin depth effect. This is the most important effect and you should always compute the skin depth from the frequency and conductivity. At high frequency in a high conductivity medium, the skin depth can be just centimetres (see table above). 8

9 G2 : Travel time curves for GPR G2.1 Half-space with variable TX-RX offset Air wave Ground wave G2.2 Single interface with variable TX-RX offset G2.2.1 Forward problem Air wave, Ground wave, Reflection, normal moveout Example from Figures 8.16 and 8.17 from Burger et al., G Inverse problem (a) With a single zero offset trace we only have a measure of the travel time at normal incidence (t 0 ). Unfortunately with just once data point, we cannot determine 2 variables (d 1 and v 1 ). (b) Plot a graph of NMO as a function of x 2 and from slope can determine v 1 Only valid when x < d (c) Plot a graph of x 2 -t 2. From the slope can determine v 1 Figure 8.20 (Burger) 9

10 G2.3 Single interface with constant TX-RX offset profiling Air wave, Ground wave Reflection Since t = d/v we cannot resolve d and v with this approach G2.4 Diffraction with constant TX-RX offset profiling Air wave, Ground wave Reflection Example from Figures 8.19 (Burger) G2.5 Variable depth with constant offset profiling As basement shallows, the travel time decreases Diffraction from corners Note that the reflection does not occur directly below the TX-RX. Can cause some distortion of the geometry. Process called migration used to remove this effect. 10

11 G3 : Data collection and analysis techniques for GPR data G3.1 Instrumentation Frequency range 10 MHZ to 1 GHz Low frequencies used for deeper imaging, but with reduced resolution from the longer wavelengths. Ground and airborne systems developed Sensors and Software, PulseEkko-100 system Ramac GPR system 11

12 Ramac GPR system mounted in a sled G3.2 Choice of frequency Vertical resolution controlled by frequency. Smallest layer that can be detected is a thickness of ¼ wavelength. Thus to image small structures a short wavelength is needed. However shorter wavelengths require a higher frequency. Since attenuation increases with frequency, this represents a trade-off in terms of resolution and depth of penetration. Illustrated in Figure 8.14 (50, 100, 200 MHz) 12

13 G3.3 Survey geometry G3.4 Velocity analysis G3.3.1 Walk away test G Buried object G3.3.3 Analysis of reflections See G2 for details 13

14 G4 : Applications of ground-penetrating radar G4.1 Geotechnical and environmental applications of GPR The following examples of GPR data are from Sensors and Software Ltd. are thanked for permission to use these figures. G4.1.1 Soil water content From Soil water content of a California vineyard Velocity of the ground wave decreases as the water content of the soil increases. This allows for non-invasive mapping in locations such as in a vineyard. 14

15 The radar measurements give good agreement with other methods of measuring soil-water content. Details in Hubbard et al., The Leading Edge, (2002) 15

16 G4.1.2 Water table depth From G4.1.3 Contaminant mapping From 16

17 G4.1.4 Contaminant remediation From G4.1.5 Underground storage tanks From 17

18 G4.1.6 Saltwater intrusion From G4.1.7 Depth to bedrock From Davis and Annan, Geophysical Prospecting, (1989) 18

19 G4.2 Archaeology

20 Detection of an archaeological cave in the Biblical city of Nysa (Shomron, Israel) using a GPR survey, from the time of Joshua Bin-Nun (around 1300 BC). This cave is known as the "Mikve" of the city. 20

21 G4.3 Forensics Locating clandestine burials G4.4 Concrete and rebar From 21

22 G4.5 Location of buried utilities G4.6 Military applications Applications include landmine clearance, unexploded ordinance detection and location of clandestine tunnels and bunkers. 22

23 G4.7 Sedimentology Real time studies of sedimentation in a gravel bar in Bangladesh from Best et al., (2003) 23

24 G4.8 Mining Example from Davis and Annan (1989) of GPR data collected in a mine. Fractures can be identified, as well as a dyke. 24

25 GPR Data collected in a potash mine 25

26 G4.9 Glaciology D4.9.1 Measuring ice thickness and bedrock topography Figure from Martin Sharp (EAS) Problems: off-nadir returns (migration), internal reflections, scattering from crevasses, absorption by water Resolution/penetration trade-off in frequency selection Map internal reflectors, bed reflection power as indicators of thermal structure 26

27 Figures from Martin Sharp (EAS) 27

28 28

29 Photos from Martin Sharp and Jeff Kavanaugh (EAS) Antarctica One of the first applications of radar for measuring ice thickness was described by Walford (1964). Also called radio-echo sounding (RES). Radio waves travel at m/ns. General review of radar applications in glaciology given by Plewes and Hubbard (2001). Extensive data base collected in Antarctica in the 1960 s and 1970s using airborne radar system. More than 400,000 km of profile data collected Scott Polar Research Institute, NSF and Technical University of Denmark. Figures below are from 29

30 Recent compilation of ice thickness data in the BEDMAP project 30

31 31

32 Icepod / Rosetta project Looking at the Ross Ice Shelf with Airborne radar and other geophysical techniques 32

33 Canadian Arctic Radar section from John Evans Glacier, Ellesmere Island. For details see Copland and Sharp (2001). Measuring ice thickness can confirm space geodetic studies of glacier volume. 33

34 G4.9.2 Mapping internal structure of ice sheets and glaciers. Internal reflections believed to be due to isochronous layers. Layers of dust from volcanic eruptions can give strong reflections. Example below is taken from the Fletcher Promontory, Antarctica and the undulations in the isochronous layers are not correlated with bedrock topography (Vaughan et al., 1999). One feature occurs at an ice divide and reflects non-linear rheology of ice. Mapping this internal structure helps with interpretation of ice core data. The internal reflections can also used to understand flow pattern of glacier. 34

35 Vaughan et al., Nature, (1999) 35

36 Figure from Martin Sharp (EAS) Local-scale snow accumulation variability on the Greenland ice sheet from ground-penetrating radar (GPR) G4.9.3 Determine basal conditions Sub-glacial lakes first identified on the basis of character of a flat basal reflection, and flat ice surface. Lake Vostock was detected from both airborne and satellite radar. 36

37 More information Animation : Elsewhere the amplitude of the basal reflection can be used to study the composition of sub-glacial sediments. Reflections from a frozen base or one containing free water are quite different. This parameter is important for understanding how easily the glacier or ice sheet can move. For examples see Holt et al., (2006) who used airborne radar to infer that the bed of the Taylor Glacier is frozen. 37

38 G4.9.4 Crevasse detection Photos from Antarctica New Zealand in

39 G5 References Arcone SA (1996), High resolution of glacial ice stratigraphy: A ground-penetrating radar study of Pegasus Runway, McMurdo Station, Antarctica. Geophysics. 61(6): Best JL, PJ Ashworth, CS Bristow, and J Roden, Three-Dimensional Sedimentary Architecture of a Large, Mid-Channel Sand Braid Bar, Jamuna River, Bangladesh, J. of Sedimentary Research, : Copland L and Sharp M Mapping thermal and hydrological conditions beneath a polythermal glacier using radio echo sounding. Journal of Glaciology, 47, Davis JL, and P Annan, Ground-penetrating radar for high resolution mapping of soil and rock stratigraphy, Geophysical Prospecting, 37, , Grote, K., S. Hubbard, Y. Rubin, GPR monitoring of volumetric water content in soils applied to highway construction and maintenance, The Leading Edge, May, Holt JW, ME Peters, DL Morse, DD Blankenship, LE Lindzey, JL Kavanaugh and KM Cuffey, Identifying and characterizing subsurface echoes in airborne radar sounding data from a high clutter environment in Taylor Valley, Antarctica, 11 th International conference of Ground Penetrating radar, June 19-22, Columbus, Ohio, USA, Hubbard S, K Grote, Y Rubin, Mapping the volumetric soil water content of a California vineyard using high frequency GPR ground wave data, The Leading Edge, , June Huisman JA, S Hubbard, JD Redman and AP Annan, Measuring soil water content with groundpenetrating radar : A review, Vadose Zone Journal, 2(4), , Plewes LA and B Hubbard, A review of the use of radio echo sounding in glaciology, Progress in Physical Geography 2001; 25; 203, DOI: / Roth K, R Schulin, H Flühler, and W Attinger (1990), Calibration of Time Domain Reflectometry for Water Content Measurement Using a Composite Dielectric Approach, Water Resour. Res., 26(10), Topp GC, Davis JL, Annan AP, Electromagnetic determination of soil water content: measurements in coaxial transmission lines. Water Resources Research 16(3), Vaughan DG, HFJ Corr, CSM Doake, and ED Waddington (1999), Distortion of isochronous layers in ice revealed by ground-penetrating radar. Nature. 398: Walford MER (1964), Radio-echo sounding through an ice shelf. Nature. 204: Welch BC, WT Pfeffer, JT Harper and NF Humphrey (1998), Mapping subglacial surfaces of temperate valley glaciers by two-pass migration of radio-echo sounding survey data, Journal of Glaciology. 44: Winebrenner DP, BE Smith, GA Catania, HB Conway and CF Raymond (2003), Radiofrequency attenuation between Siple Dome, West Antarctica, from wide-angle and profiling radar observations. Annals of Glaciology. 37: