Radar Methods General Overview

Transcription

1 Environmental and Exploration Geophysics II Radar Methods General Overview tom.h.wilson Department of Geology and Geography West Virginia University Morgantown, WV Brown (2004) 1

2 Neidell (1990) Sheriff (1996) Neidell (1990) 2

3 Brown (2004) Sheriff (1996) 3

4 The radar band is loosely taken to extend from approximately 0.1cm to 100cm. The microwave region is often used for surface imaging from airborne or satellite platform. Radar image of the earth s surface at 5.4cm or 5.6 GHz. Seneca Rocks 4

5 Ground penetrating radar (GPR) systems often operate in the tens of MHz to GHz region of the spectrum. 25MHz = 12m wavelength (40ns) 50MHz = 6m (20ns) 100MHz = 3m (10ns) 1GHz = 0.3m (1ns) Times in nanoseconds represent the time it takes light to travel through 1 wavelength in a vacuum. Visual wavelength image Shuttle Imaging Radar - SIR A λ ~ 25cm Sabins,

6 Sabins, 1996 Ground Surveys GPR mono-static and bi-static transmitter-receiver configurations. Note similarity to coincident source-receiver and offset source receiver configurations discussed in the context of seismic methods Daniels, J., 1989 & Sensors and Software 6

7 Spectral and temporal characteristics of the GPR wavelet. Does this sound familiar! Sensors & Software Inc. - Ekko Updates As with seismic data, reflection arrival times are 2-way times and depth equals ½ the two-way time x average velocity. Velocity in air is approximately equal to the velocity of light in a vacuum: c. c = 3 x 10 8 m/sec = 9.84 x 10 8 f/s or approximately 1 foot per nanosecond. 1 nanosecond is 10-9th seconds. This also corresponds to or 0.3 m/nanosecond Thinking in terms of two-way times, it takes 2ns to travel 1 ft. 7

8 In general the velocity of the radar wave is defined as V = c ε r Valid only when electrical conductivity is small where c is the velocity of light in a vacuum (or air), and ε r is the electric permitivity of the material through which the radar wave travels. Examples of ε r (see Daniels) are 81 for water 6 for unsaturated sand 20 for saturated sand The presence of water has a significant effect on velocity. Typical velocities V = c ε r c ~ 1ft/ns in air v ~ 1/2 to 1/3rd ft/ns in unsaturated sand v ~ 1/3rd to 1/5th ft/ns in saturated sand ε is proportional to conductivity σ - materials of relatively high conductivity have slower velocity than less conductive materials. 8

9 A r = A s e α r In our discussions of seismic we recognized absorption as an important process affecting the ability of the seismic wave to penetrate beneath the earth s surface. High attenuation coefficient α produces rapid decay of seismic wave amplitude with distance traveled (r). The same process controls the ability of electromagnetic waves to penetrate beneath the earth s surface. The expression controlling attenuation is a function of several quantities, the most important of which are conductivity and permitivity. Attenuation of electromagnetic waves is controlled by the propagation factor which has real and imaginary parts. The real part α (the attenuation coefficient) illustrates the influence of permitivity and conductivity on absorption. α = ω με ( 2 σ 1+ ( ) ωε 2 1) Note in this equation that increases of σ translate into increased attenuation. Also note that increases of angular frequency (ω=2πf) will increase attenuation. 9

10 The display of radar waves shows considerable similarity to that of seismic data Diffraction events are commonly produced by heterogeneity in the electrical properties of subsurface materials 10

11 The diffraction response can be used - as you would have guessed to determine velocity. Remember the ray path geometry for the diffraction event? * For coincident source and receiver acquisition 11

12 * d = 2 x + z 2 2 d x + z t = = V x + z Vt = 4x+ 4z Vt 4x = 4z Vt x = 4z z 2z a = V b= z Slope of Asymtote = a 2 = b V 12

13 Average Velocity = 1/2 the reciprocal of the slope Note that the 0.2 m/ns velocities in the sand dune complex is pretty high compared to the above. Sensors & Software Inc. - Ekko Updates 13

14 Direct airwave arrival Direct arrival through surface medium Reflection hyperbola Smith and Jol, 1995 The characteristics of a common midpoint gather from a GPR data set look very similar to those from a seismic CMP gather. Thinning layer response and resolution considerations. Daniels, J.,

15 Horizontal Resolution: The Fresnel Zone The Fresnel Zone Radius R f R f 2 lλ λ = Depth = 2000 Velocity = dominant frequency = 50 two-way time = dominant wavelength = 270 Fresnel zone radius = Approximation = An approximation R f V 4 4t f c 15

16 Topographic variations must also be compensated for. Daniels, J., 1989 West Pearl Queen Field Area 16

17 Surface along the GPR line shown below was very irregular so that apparent structure in the section below is often the result of relief across features in the surface sand dune complex. GPR data is often collected by pulling the GPR unit across the surface. Subsurface scans are made at regular intervals, but since the unit is often pulled at varying speeds across the surface, the records are adjusted to portray constant spacing between records. This process s referred to as rubbersheeting. Daniels, J.,

18 Smith and Jol, 1995, AG Smith and Jol, 1995, AG 18

19 Increased frequency and bandwidth reduce the dominant period and duration of the wavelet Comparison of the 25MHz and 100 MHz records Smith and Jol, 1995, AG 19

20 A r = A s e α r We also expect to see decreased depth of penetration (i.e. increased attenuation) for higher frequency wavelets and components of the GPR signal. Smith and Jol, 1995, AG Sensors & Software Inc. - Ekko Updates 20

21 In the acquisition of GPR data one must worry about overhead reflections. Daniels, J., and tree branches! Daniels, J.,

22 GPR unit Visit Sensors & Software Inc. Smart Cart Sensors & Software Inc. Salt Water Infiltration 22

23 Pulse EKKO bistatic Sensors and Software - Locating underground storage tanks 23

24 Sensors and Software Locating unexploded ordnance Sensors and Software forensic applications 24

25 Time slice map from 3D data volume of radar data. This is a surface of equal travel time. Disruptions in the reflection pattern are associated with the waste pit. Green et al., 1999, LE Bright red areas define the location of the landfill; the orange objects represent gravel bodies. The brownish-pink lobes are high reflectivity objects of unknown origin. The view at bottom profiles the underside of the landfill and gravel bodies. Green et al., 1999, LE 25

26 Sensors and Software Avalanche Assessment Sensors & Software Inc. Glaciers, Ice and Snow 26

27 Sensors & Software Inc. Plastic and metal pipes 27

28 Sensors & Software Inc. - Ekko Updates 28