See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/311805030

Two-dimensional seismic refraction tomography of a buried bedrock valley

at Hallsands beach, Devon, United Kingdom

Article  in  Environmental Geosciences · December 2016

DOI: 10.1306/eg.07131615014

CITATIONS READS

0 175

5 authors, including:

David Malone Eric Wade Peterson

Illinois State University Illinois State University
139 PUBLICATIONS   344 CITATIONS    69 PUBLICATIONS   557 CITATIONS   

SEE PROFILE SEE PROFILE

William P Anderson
Appalachian State University
31 PUBLICATIONS   248 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Karst of Carter County, KY View project

Stream Temperatures in Urbanized Headwater Streams View project

All content following this page was uploaded by David Malone on 19 May 2018.

The user has requested enhancement of the downloaded file.

1
Q:1 Two-dimensional seismic AUTHORS

2
refraction tomography of Eric B. Avalos ~ Department of
Geography–Geology, Campus Box 4400,
Illinois State University, Normal, Illinois
3
a buried bedrock valley at 61790-4400; present address: 25519 Owl
Landing Lane, Katy, Texas 77494;
4
Hallsands beach, Devon, ebavalos1@gmail.com Q:3
Eric Avalos is a production geologist at Shell
5
Q:2 United Kingdom Oil Company in Houston, Texas. He
holds B.S. and M.S. degrees from Illinois
6 Eric B. Avalos, David H. Malone, Eric W. Peterson, State University, Normal, Illinois.
7 William P. Anderson, and Roland W. Gehrels David H. Malone ~ Department of
Geography–Geology, Campus Box 4400,
Illinois State University, Normal, Illinois
61790-4400; dhmalon@ilstu.edu
8 ABSTRACT
David Malone is a university professor of
9 Two-dimensional seismic refraction tomography was used to map the geology at Illinois State University. He
10 bedrock topography beneath Hallsands beach in southwest Devon, completed his Ph.D. at the University of
11 United Kingdom. Seismic refraction data were acquired from 11 Wisconsin. His research interests include
12 spreads, 4 parallel to the beach and 7 normal to the beach, with either the structure and stratigraphy of Eocene
13 12 or 24 geophones at 5-m spacing. Eight sediment cores were used to volcanic rocks of the Absaroka Range of
14 calibrate the velocity model. The bedrock consists of metasedimentary Wyoming, the Heart Mountain Slide, three-
15 rocks that have a seismic velocity of 2100–2500 m/s and is overlain by dimensional geologic mapping, and
provenance analysis of cratonic and
16 variable amounts of gravel, peat, and muddy peat. Wood peat and
synorogenic strata using detrital zircon
17 peaty mud are differentiated within the peat as 700-m/s velocity for
geochronology.
18 wood peat and 1200-m/s velocity for peaty mud. These refraction data
19 were collected and processed in two dimensions, then imported into Eric W. Peterson ~ Department of
20 Petrel, a three-dimensional (3-D) geological modeling software pack- Geography–Geology, Campus Box 4400,
21 age. The 3-D geologic model was built using the velocity attribute of Illinois State University, Normal, Illinois
22 the seismic refraction data. These selected data points were used to 61790-4400; ewpeter@ilstu.edu
23 create 3-D horizons, surfaces, and contacts constraining the target Eric Peterson is a professor in the
24 bedrock surface from the overlying unconsolidated deposits. Geography–Geology Department at Illinois
25 The bedrock surface beneath Hallsands beach is marked by two State University, with expertise in hydrogeology.
26 paleochannels. One paleochannel occurs in the north end of the beach He holds B.S. and M.A. degrees from the
27 beneath the axis of the modern valley. A second paleochannel occurs University of South Dakota, an M.S. degree
from the University of Arkansas, and a Ph.D.
28 in the southern section of Hallsands beach centered along the axis of
from the University of Missouri.
29 a tributary valley. Bedrock occurs at a depth of approximately -10 m in
30 the southern and northern sections of the main valley. Bedrock occurs William P. Anderson ~ Department of
31 at a depth of approximately -2 m along the valley wall at the southern Geology, Appalachian State University;
32 Q:5 end of the beach east of the parking lot. Shore-perpendicular refraction andersonwp@appstate.edu
33 lines differentiate layers within the peat, whereas shore-parallel lines William P. Anderson, Jr., is professor and
34 delineate wood-peat, peaty-mud, and bedrock topography. chair of the Department of Geology at
Appalachian State University. He completed
his Ph.D. at North Carolina State
University in 1999. His research interests
Copyright ©2016. The American Association of Petroleum Geologists/Division of Environmental include groundwater–surface-water
Geosciences. All rights reserved.
interactions in both riparian and coastal
Manuscript received September 23, 2015; provisional acceptance September 23, 2015; revised manuscript
received August 17, 2016; final acceptance December 7, 2016.
settings and urban hydrology.
DOI:10.1306/eg.07131615014

Environmental Geosciences, v. 23, no. 4 (December 2016), pp. 179–193 179

 Roland W. Gehrels ~ Department of INTRODUCTION 35
Environment, York University; roland.
Q:4 gehrels@york.ac.uk The purpose of this research is to better constrain the geometry of the 36

Roland Gehrels is a Quaternary geology and bedrock surface paleotopography beneath Hallsands beach, Devon, 37
physical geography professor at the University United Kingdom, using seismic refraction tomography (Figure 1). 38
of York, United Kingdom. He completed his According to Gehrels and Long (2008) and Massey et al. (2008), 39
Ph.D. at the University of Maine. He understanding the relationship between the bedrock surface and 40
specializes in the investigation of Holocene overlying sediments will contribute to better understanding of salt- 41
sea-level rise, particularly the rate of water/freshwater mixing zone, which in turn will help constrain the 42
sea-level rise prior to and during the last two dynamics of Holocene sea-level change (Gehrels and Anderson, 43
centuries of increased greenhouse-gas
2014). 44
emissions.
Hallsands beach is located north of Hallsands village at the 45
opening of Bickerton Valley along the coast of Start Bay in southwest 46
ACKNOWLEDGMENTS Devon, United Kingdom (Figures 1, 2). The mouth of the valley is Q:6 47

Funding for this research was provided by occupied by gravel beach sediment that has accumulated since the 48

student research grants (to Avalos) from the end of the Pleistocene (Robinson, 1961). Steep cliffs of metasedi- 49

Geological Society of America and American mentary rock outcrop north and south of this beach (Ussher, 1905; 50
Geological Institute. The Department of Robinson, 1961). A small perennial stream flows along Bickerton 51
Geography–Geology provided the Valley. Under most flow conditions, this stream disappears upon 52
geophysical equipment used in this study. encountering porous and permeable beach sediments (Figure 2). An 53
We would like to thank Schlumberger Inc. intermittent stream flows through a tributary valley entering from the 54
for the donation of the Petrel software southwest. 55
package to Illinois State University. Eric The bedrock geology consists of lower Devonian schists and slates 56
Munson assisted in the field-data acquisition.
(Ussher, 1905). Bedrock outcrops display folded and faulted units 57
north and south of Bickerton Valley (Coward and McClay, 1983). Q:7 58
Overlying the bedrock are peat beds (Gehrels and Anderson, 2014) 59
that consist of phragmites peat, wood peat, and peaty mud inter- 60
bedded with muddy gravel and gravel (Figure 3). Overlying the peat 61
are sorted, rounded gravel of varying lithologies that include flint, 62
quartzite, vein quartz with lesser rhyolite, granite, and slate 63
(Dornbusch et al., 2002). The gravel beaches are locally known as 64
“shingle” beaches given the preponderance of slate clasts. 65

REFRACTION STUDIES FOR ENVIRONMENTAL PROBLEMS 66

Traditional seismic refraction methods assume a flat horizontal 67

plane to provide basic information on the geologic unit such as depth 68
to refractor and dipping or horizontal refractor (Lankston, 1989; 69
Al-Saigh and Al-Dabbagh, 2010). Despite the inaccuracy of such 70
assumptions, the lack of modern technological advances did not 71
allow for processing and interpreting increasingly complex scientific 72
scenarios. With the application of sophisticated computer processing 73
software, many of the previous limitations regarding refraction meth- 74
ods such as assuming a horizontal refractor, steeply dipping beds, 75
and complex and/or faulted geology no longer are a hindrance. 76
Refraction tomography often is used to interpret shallow sub- 77
surface geology (Min and Shin, 2006). Zhao and Xu (2010) used two- 78
dimensional (2-D) refraction tomography to map subsurface sediments 79
less than 300 m in depth offshore Kyushu, Japan. Higuera-Dı́az et al. 80

180 Two-Dimensional Seismic Refraction Tomography of a Buried Bedrock Valley

 Figure 1. Study site location in
Devon, United Kingdom, with all
refraction lines indicated; A marks
the location of boulders deposited
for erosion control. Wells and
cores indicated. Core 1-8 located
along the north valley wall.
Modified from Gehrels and
Anderson (2014).

81 (2007) used 2-D refraction tomography to test the ef- interpreted as stratigraphic variability in the peat and 95
82 fectiveness of refraction tomography to image bedrock gravel units that helped to determine fault orientation 96
83 depressions interpreted as sinkholes in karst topography and throw. Karastathis et al. (2007) used shallow 2-D 97
84 beneath unconsolidated sediments. Hirsch et al. (2008) refraction tomography to image active faults from depths 98
85 used refraction tomography to map the bedrock surface less than 30 to greater than 300 m in central Greece. Fault 99
86 beneath terrace deposits along the Bow River near Cal- interpretations were based on geometry and dip of ve- 100
87 gary, Canada. They found that refraction tomography locity layers. 101
88 delineated bedrock topography better than ground- Heincke et al. (2006) used 2-D and 3-D refraction 102
89 Q:8 penetrating radar or resistivity methods at the same lo- tomography to map mountain slope deposits in an effort 103
90 cation. Tsai et al. (2008) used three-dimensional (3-D) to better understand dynamics of catastrophic collapses 104
91 refraction and 3-D reflection to map near-surface faults as within the Swiss Alps. Zelt et al. (2006) used true 3-D 105
92 much as 30 m below a peat and gravel succession in the refraction to map a paleochannel within bedrock at a 106
93 vicinity of the San Andreas Fault Zone in southern Cal- groundwater contamination site at Hill Air Force Base, 107
94 ifornia. Lateral variations in refraction velocity were Utah. The purpose was to determine the viability of 3-D 108

Avalos et al. 181

Q:16 Figure 2. (A) Hallsands beach
at mouth of Bickerton Valley
looking south with refraction
lines, borehole, and well locations
indicated. Notice line 1004 was
run during low tide, whereas
photo was taken as tide was in-
creasing. (B) Phragmites peat and
fossil Alnus stumps exposed on
the beach at Hallsands in
Q:17 March 2007 (from Gehrels and
Anderson, 2014).

109 refraction tomography to delineate near-surface geol- Geophones were run at low tide within coarse sand 122
110 ogy in a coastal setting. Hartz et al. (2016) used seis- sediment. Maguire et al. (2011) reinterpreted pre- 123
111 mic refraction data to develop a 3-D model of Henry vious reflection data in southern England, primarily 124
112 Formation valley train outwash in southern McLean Wales. This study determined subsurface geology 125
113 County, Illinois. No similar research in coastal settings greater than 1 km in depth rather than near surface. 126
114 has as yet been published. Doody and Brooks (1986) interpreted 2-D seismic 127
refraction surveys conducted in southwest England 128
between Lizard and Start Peninsulas southwest of 129
115 PREVIOUS WORK AT HALLSANDS Hallsands. They mapped the bedrock schists present 130
here in detail but did not resolve the overlying sed- 131
116 Kelland (1975) mapped surficial sediment geome- iment units. 132
117 tries and bedrock topography offshore Hallsands With regards to near-surface marginal marine 133
118 beach into Start Bay using 2-D reflection surveying. successions, few data have been found involving 134
119 Brooks et al. (1993) mapped the geometries of bed- seismic surveys of any kind, and no data have been 135
120 rock structure in the Bristol Channel along north found involving 3-D mapping with 2-D refraction. 136
121 Devon using both 2-D reflection and refraction data. Lenham et al. (2005) mapped the bedrock geology 137

182 Two-Dimensional Seismic Refraction Tomography of a Buried Bedrock Valley

 Figure 3. Interpreted cross
section AA’ from wells 50 to 1.
MSL = mean sea level. Q:18

138 beneath Mersey Estuary in the west–central United hammer against an aluminum plate with accelerometer 168
139 Kingdom with 2-D refraction and reflection surveys trigger, five-shot stack, and a shot every two stations. 169
140 for the purpose of bridge construction. Eddies and Weather conditions were 10°C–20°C with overcast 170
141 Reynolds (1988) located what they named “buried skies with some minimal precipitation. Tidal conditions 171
142 rock valleys” at near-surface depths within Plymouth were approaching spring tides with 5-m difference 172
143 Sound and the River Tamar southwest of Hallsands, between high and low tides. During the week of data 173
144 United Kingdom; 2-D seismic reflection continuous collection, lowest tides occurred from 12:00 to 4:00 174
145 profiling was the method employed. Eddies and p.m. local time. Maximum sea wave amplitudes were 175
146 Reynolds (1988) along with Kelland (1975) were the approximately 0.75 m. Lines were run at low tide to 176
147 only studies found to locate bedrock valleys within or minimize noise from wave action. 177
148 around south Devon.
149 Gehrels and Anderson (2014) presented a novel Data Processing 178
150 method to reconstruct sea-level change in coastal fresh-
151 water back-barrier marshes. They prepared groundwater First breaks were manually picked, and velocity models 179
152 and water-table models to document sea-level positions were created with Rayfract 3.14 seismic refraction in- 180
153 between 7200 and 2400 calibrated ka. terpretation software. Figure 4 is a sample of unfiltered 181
refraction data with first breaks chosen for line 1011. All 182
lines were picked with no filtering. This was possible 183
154 METHODOLODGY given the lack of noise, because the beach gravel appears 184
to have suppressed noise from waves. Velocities were 185
155 Data Collection modeled using wavepath eikonal traveltime (WET) 186
tomography modeling (Schuster and Quintus-Bosz, 187
156 Seismic refraction data were acquired from 11 spreads 1993). Schuster and Quintus-Bosz (1993) provide an 188
157 with either 12 or 24 geophones at 5-m spacing during excellent summary of WET inversion modeling for 189
158 February 2011 (Figures 1, 2). Lines 1001–1004 were 2-D refraction processing. They describe how WET to- 190
159 run shore parallel with 24 geophones for line 1001 and mography maintains accuracy despite increased speed 191
160 12 geophones for lines 1002–1004 with an overlap of 6 and simplicity in processing and modeling compared with 192
161 geophones for each spread. Lines 1005 through 1008 other tomographic methods. 193
162 were run shore perpendicular with 12 geophones at 5-m To run WET tomography in Rayfract, software 194
163 spacing. Lines 1009 and 1010 were run shore perpen- limitations require a minimum shot every six stations, 195
164 dicular with 12 geophones at 2-m spacing. Geophones with a shot every three stations preferred. A shot was 196
165 were inserted directly into the gravel. Elevation cor- recorded every two stations to ensure complete data 197
166 rections were made using true elevation with mean sea coverage and account for bad shots. In accordance with 198
167 level (MSL) as datum. The energy source was a 5-kg Schuster and Quintus-Bosz (1993), each model was run 199

Avalos et al. 183

 Figure 4. Unfiltered refraction data with first breaks chosen for line 1011. Line 1011 was run with 24 geophones with first shot before first
geophone and last shot 20 m after last geophone, with a shot every 2 stations.

200 with a maximum of 20 iterations with MSL elevations refractor interpreted as bedrock between 2100 and 231
201 as correction for beach topography. Base elevation was 2500 m/s. Within the velocity model, velocity-contour 232
202 recorded at the parking lot at 4.5 m. Elevation correc- shape changed little between 2100 and 2500 m/s. The 233
203 tions were taken for each station that deviated from this only significant change from the 2100-m/s velocity 234
204 elevation. Model output is represented as 2-D cross contour to the 2500-m/s velocity contour is an in- 235
205 sections with contoured velocity layers within Surfer crease in depth with an increase in velocity. As such, 236
206 Golden Software. the 2100-m/s velocity contour was interpreted as 237
weathered bedrock velocity, and 2500 m/s was inter- 238
207 Velocity Calibration preted as the more competent (nonweathered) bedrock 239
velocity. 240
208 Eight cores (Gehrels and Anderson, 2014) encountered In comparison with core 8, depth of the 2100-m/s 241
209 bedrock at approximately 1-m elevation at core 8. Core velocity in line 1011 is approximately 2 m greater in 242
210 8 is located along the valley wall within the peat depth than bedrock at core 8. At core 8, bedrock is 243
211 (Figure 1). A planned spread running shore perpen- located at approximately 1-m elevation, whereas the 244
212 dicular from well 50 (within the peat) to the sea was to 2100-m/s velocity layer in line 1011 is located at ap- 245
213 be used for velocity calibration. During the two weeks proximately -1-m elevation. This would suggest that 246
214 of data collection, the peat was saturated with stand- a velocity of 2100 m/s should not be used as bedrock 247
215 ing water, preventing access to well 50. Well 50 is the velocity. Despite this difference, however, 2100-m/s 248
216 only well to encounter bedrock at approximately -8 m velocity interpreted as bedrock velocity is supported 249
217 in elevation (Figure 3). Given that well 50 was inacces- by lines 1001 and 1002, both of which intersect line 250
218 sible, and to obtain some velocity control with subsur- 1011, and both indicate that bedrock decreases in ele- 251
219 face data, line 1011 (Figure 5) was run along the valley vation from core 8 to line 1011, agreeing with the chosen 252
220 wall above the bedrock ridge approximately 10 m north velocity of 2100 m/s. 253
221 of core 8 (Figures 1, 2). Similar to determination of bedrock velocity, gravel 254
222 Line 1011 is located approximately 10 m lateral and peat velocities were calculated from field data using 255
223 offset from core 8 (Figure 1). Initial comparison of ve- beginning and ending shots as explained by Lankston 256
224 locity data in line 1011 and core 8 suggested a bedrock (1989). Gravel velocities ranged from 500 to 700 m/s, 257
225 velocity of 1800 m/s assuming bedrock topography and peat velocity was consistently 1200 m/s. The depths 258
226 does not change in the 10 m lateral offset between core of each of these velocity layers in the tomography cross 259
227 8 and line 1011. In addition, field data used to calculate sections were compared with depths in the interpreted 260
228 velocities from a traditional refraction method using cross section (Figure 3). Final interpretations of 700 m/s 261
229 beginning and ending shots as explained by Lankston for gravel and 1200 m/s for peaty mud correlate with all 262
230 (1989) indicate a range of velocities for the basement available data. 263

184 Two-Dimensional Seismic Refraction Tomography of a Buried Bedrock Valley

 Q:19 Figure 5. Line 1011 runs north of beach along valley wall to aid in bedrock velocity calibration with core data (Figures 1, 2). Bold line marks
the 2100-m/s velocity contour, which shows bedrock depth at approximately -1-m elevation. This correlates with bedrock located in core 8
at approximately 1-m elevation. Discrepancy in depths from seismic to core 8 is the result of a 10-m lateral offset from the seismic line to core
8. Note: no vertical exaggeration.

264 RESULTS AND DISCUSSION depth in the valley center. Because the line was run with 287
12 geophones at 5-m spacing resulting in a 60-m spread, 288
265 First breaks were successfully picked and models were reliable interpretations can only be made to -20-m 289
266 completed for all seismic refraction lines. Output grid depth. This results in a confident interpretation for the 290
267 files from 2-D models are represented as velocity- existence and location of a paleochannel within the 291
268 contour cross sections within Surfer Golden Soft- bedrock for line 1001 but not a confident interpretation 292
269 ware. Where lines intersect, each model shows a close regarding depths below -20 m. 293
270 approximate correlation with corresponding velocity Line 1002 (Figure 7) displays the bedrock valley 294
271 layers. Shore-parallel lines (lines 1001–1004) having elevations from approximately -2 m in the south 295
272 longer spreads imaged deeper (~15–20-m total depth) to -10 m in the north. Line 1002 is interpreted to 296
273 than shore-perpendicular lines. For imaging bedrock, display a paleochannel with similar geometry, depth, 297
274 shore-parallel lines produced deeper, smoother velocity- and location to that found in line 1001. As with line 298
275 contour profiles than shore-perpendicular lines. Shore- 1001, the location of the paleochannel is interpreted 299
276 parallel lines are perpendicular to the bedrock valley, with a high degree of confidence; however, the depth is 300
277 providing a more visually pleasing display of the valley only reliable to -20 m. The 1200-m/s velocity layer is 301
278 topography. In addition, the longer lines provide more interpreted as the peaty-mud surface. Peat layers above 302
279 data coverage, with farther offset shots that result in peaty mud are not differentiated. 303
280 deeper, more continuous tomographic images compared The paleochannel represented in lines 1001 and 304
281 with shore-perpendicular lines. 1002 should exist in line 1003 from approximately 130 305
282 A significant limitation of the seismic refraction to approximately 175 m (Figure 8). Line 1003 does dis- 306
283 Q:9 method is that depth interpretations are reliable to 0.33 play a paleochannel along the south section following 307
284 the lateral length of the line. For example, line 1001 a similar location and geometry for the modern trib- 308
285 (Figure 6) displays the 2100-m/s velocity layer at ap- utary valley in Figure 2 but not for the main valley. The 309
286 proximately -10 m at the valley edges to at least -20-m paleochannel is represented in line 1004, which was run 310

Avalos et al. 185

 Figure 6. Line 1001 with 2100-m/s velocity contour interpreted as bedrock surface. Note: no vertical exaggeration.

311 east of line 1003 in the tidal zone at low tide (Figure 9). along the coast, including a spread offshore Hallsands. 324
312 Figure 9 shows two paleochannels within the bedrock, Kelland (1975) interpreted possible relict barrier beaches 325
313 indicating that the paleochannel may have diverged as well. 326
314 between lines 1002 and 1004. To map the surface of the bedrock paleochannel, 327
315 A possible explanation for the lack of a paleo- the shore-parallel lines provide better results than shore- 328
316 channel in the north section of line 1003 is a more perpendicular lines. Shore-perpendicular lines (lines 329
317 competent layer of peat such as greater clay content 1005–1010) were run west to east (shoreward to sea). 330
318 within the peat, which resulted in a higher velocity at As such, the lines are shorter spreads run with 12 331
319 this location. A second possibility is that a relict barrier geophones, resulting in some lines not reaching bedrock 332
320 beach possibly cemented or partially cemented formed depths. Line 1005 (Figure 10) imaged peat interpreted 333
321 during a period of sea-level lowstand, resulting in a as peaty mud at approximately -3 m in the west to 334
322 higher-velocity layer. Kelland (1975) conducted seismic apparently outcrop in the east. Because no observations 335
323 reflection spreads in Start Bay, mapping paleochannels of peat outcropping within the beach were observed 336

Figure 7. Line 1002 shore parallel showing a paleochannel within bedrock similar to line 1001; 1200-m/s velocity contour interpreted as
peaty-mud surface.

186 Two-Dimensional Seismic Refraction Tomography of a Buried Bedrock Valley

 Q:20 Figure 8. Line 1003 with 2100-m/s velocity contour in bold representing contact between peat and bedrock surface. Locations A and B are
interpreted as boulders deposited in these locations for erosion control. The 1200-m/s velocity contour interpreted as the peaty-mud surface.

337 during the survey, this is considered an artifact of the to the east (seaward). Wood peat appears to increase 356
338 2-D model. Instead, peat must approach the surface from an elevation of -5 m in the west to outcrop at or 357
339 at this location and lie just beneath the surface within near the surface to the east. Given that no peat was 358
340 the gravel. Bedrock interpretation for line 1005 is in- observed exposed during data collection, peat is 359
341 consistent with line 1003 with respect to depth. The interpreted to occur just below the gravel surface at this 360
342 bedrock anomaly in line 1005 is interpreted as a possible location, similar to line 1005. 361
343 boulder or sea stack oriented west to east that was not From -5 to 10 m, line 1006 (Figure 11) does not 362
344 large enough to be located by line 1003, given that 1003 reveal any layers beneath the surface because of a ve- 363
345 was run north to south. As indicated in line 1003, locity anomaly. A possible explanation for this anomaly 364
346 boulders were deposited near this location for erosion is the presence of a shallow or perched water table. 365
347 control. The existence of a west–east-trending boulder Where the anomaly ends at 10 m into the spread also 366
348 is a viable possibility. marks the end of the surface peat and beginning of 367
349 Lines 1006, 1007, and 1009 (Figures 11–13, respec- beach gravel. A second possible explanation is the at- 368
350 tively) did not reach the interpreted bedrock velocity tenuation of the seismic wave pulse at the peat surface. 369
351 and as such are interpreted to image layers within the Once the energy source reached the more resistant 370
352 peat. All three lines show two distinct velocity layers, gravel, pulse waves were once again able to penetrate the 371
353 700 and 1200 m/s, interpreted as wood-peat and peaty- subsurface peat layer. 372
354 mud velocities, respectively. Line 1006 indicates that Line 1007 (Figure 12) shows a similar geometry to line 373
355 gravel thickens to the west (shoreward), pinching out 1006 (Figure 11), with peat occurring at approximately 374

Figure 9. Line 1004 shore parallel showing paleochannels in both north and south locations of the cross section. Line 1004 is east of line
1003, which does not show a distinct paleochannel in the north. Large structures on north end of line may be velocity artifacts given the
amount of noise at this location during data collection; however, these are interpreted as possible sea stacks.

Avalos et al. 187

 Figure 10. Line 1005 shore perpendicular along northern section of Hallsands beach.

375 -7-m elevation in the west to approximately 0-m above peaty mud, or the wood peat at this location has 385
376 elevation in the east. The 700-m/s velocity contour is similar geophysical properties to the gravel, resulting in 386
377 shown interpreted as the surface layer of wood peat. a hidden layer problem. A hidden layer occurs in seismic 387
378 Throughout the rest of the study, an anomaly exists from refraction when a subsurface unit has equal or lower 388
379 25 m to the end of the line approximately 5 m thick, velocity than the unit above. The refraction method 389
380 marked as A in line 1007 (Figure 12). This anomaly is works assuming velocity increases with depth. Gravel 390
381 interpreted to be a possible clay-rich lens or remnant of over peat over bedrock satisfies this requirement. Vari- 391
382 cemented beach gravel resulting in a higher velocity for ous layers within the peat, however, may not increase in 392
383 this location. Wood peat is interpreted to have eroded velocity with depth depending upon the competency or 393
384 away at this location, resulting in a thicker gravel layer amount of clay within each individual peat unit. The end 394

Figure 11. Line 1006 shore perpendicular west to east. The energy source does not appear to transmit well within the peat when
originated at the peat surface from -5 to 10 m along the spread.

188 Two-Dimensional Seismic Refraction Tomography of a Buried Bedrock Valley

 Figure 12. Line 1007 shore perpendicular running west to east. Location A is an area of higher velocity interpreted as a possible clay-rich
lens or remnant of cemented or partially cemented gravel beach section.

395 result is that certain layers within the peat may not be Wood peat occurs at approximately 1-m elevation in the 421
396 differentiated by refraction tomography. west, with a thickness of approximately 1 m thinning to 422
397 Line 1008 (Figure 13) satisfies the requirement for the east. Peaty mud underlies wood peat at a depth of 423
398 ideal refraction modeling of increasing velocity with 0-m elevation, with bedrock occurring from -2- to -1-m 424
399 depth, resulting in four clear individual velocity units elevation. Line 1010 occurs approximately 10 m from 425
400 delineating peat and bedrock surfaces. Line 1008 is the cliffs along the southern end of the beach where 426
401 10 m longer than line 1007 or 1006, which may explain bedrock outcrop was observed (Figures 1, 2). 427
402 why bedrock was reached, because refraction depth is
403 dependent upon the length of the survey. As with lines
404 1005 and 1006, the gravel unit decreases in thick- BEDROCK SURFACE THREE-DIMENSIONAL 428
405 ness from west to east overlying the 1200-m/s layer GEOLOGIC MODEL 429
406 interpreted as peaty mud. Bedrock topography fluc-
407 tuates some from west to east between -10 and -15 m. Output files from 2-D velocity-contour cross sections 430
408 Lines 1009 and 1010 (Figures 14, 15, respectively) were converted from grid files to X, Y, Z American 431
409 were imaged at 2.5-m geophone spacing as a result of Standard Code for Information Interchange files with an Q:10432
410 surface location between the parking lot and sea, as associated velocity attribute for each Z value. The result 433
411 indicated in Figures 1 and 2. Line 1009 (Figure 14) is a user-friendly format to convert output files into for- 434
412 imaged to a maximum depth of -5-m elevation, showing mats compatible with other software. Here, 2-D grid files 435
413 gravel, wood peat, and peaty mud in a “layer-cake” ar- from Surfer were opened with Microsoft Excel. In Excel, 436
414 rangement. Gravel exists to a depth of approximately the grid coordinates were converted from two dimensions 437
415 0 m. Wood peat occurs from approximately 1-m ele- to three dimensions for import into Petrel. A 3-D geologic 438
416 vation at 3-m thickness in the west, thinning to ap- model was created using the completed bedrock and 439
417 proximately 1 m to the east. Peaty mud exists from wood-peat surfaces as input into a final 3-D grid at 0.5-m 440
418 approximately 0- to -2-m elevation. resolution with Petrel’s 3-D grid process. For horizon 441
419 Line 1010 imaged to a depth of -7 m, with gravel modeling, the top surface was assigned erosional, and 442
420 occurring at approximately 2-m elevation (Figure 15). bottom surface was assigned base. All other surfaces were 443

Avalos et al. 189

 Figure 13. Line 1008 shore perpendicular imaging gravel, peat, and bedrock; 1200-m/s velocity layer interpreted as peaty-mud surface,
and 2100-m/s velocity layer interpreted as bedrock surface.

444 assigned conformable. Erosional instructs the model to paleochannels. One channel occurs in the north end of 464
445 truncate surfaces below the erosional surface where the the beach beneath the center of the modern main 465
446 two surfaces intersect. Base truncates all surfaces above valley (Figure 2). A second channel occurs in the 466
447 the base where the two surfaces intersect. Conformable southern section of Hallsands beach centered along 467
448 is truncated by the surfaces above and truncates those the mouth of the tributary valley (Figure 2). Bedrock 468
449 below, except for the surface assigned as base. The end occurs from approximately -10 m in the southern and 469
450 result is a geologic model with a bedrock valley, filled with northern sections of the main valley. Bedrock occurs 470
451 a peat surface and truncated by the modern erosional at approximately -2-m elevation along the modern 471
452 surface. Differentiated zones from well and core data valley wall at the southern end of the beach east of the 472
453 were created using input surfaces as boundaries. Zones parking lot. 473
454 were created between surfaces using well tops from core Wood peat and peaty mud are differentiated 474
455 and well data as data input in accordance with Petrel’s within the peat as 700-m/s velocity for wood peat 475
456 “make zones” process to obtain the final geologic inter- and 1200-m/s velocity for peaty mud. Other layers 476
457 pretation. Figure 16 displays inline and crossline cross sec- of peat were not differentiated by seismic refraction 477
458 tions through the model. The geologic model shows tomography. In general, shore-perpendicular lines dif- 478
459 a reasonable geologic interpretation that correlates with core ferentiated layers within the peat, whereas shore-parallel 479
460 data, field observations, and 2-D refraction interpretations. lines delineated wood-peat, peaty-mud, and bedrock 480
topography. 481
Using 2-D refraction data to determine sub- 482
461 CONCLUSIONS surface geological conditions may seem regressive 483
given the superior quality and nature of 3-D seismic 484
462 The bedrock valley beneath Hallsands is possibly a reflection or true 3-D refraction methods. The spe- 485
463 U-shaped valley or paleochannel floodplain with two cific equipment package used to collect data for this 486

190 Two-Dimensional Seismic Refraction Tomography of a Buried Bedrock Valley

Q:21 Figure 14. Line 1009 shore perpendicular running west to east delineates peaty mud from wood peat. Bedrock was not imaged with this
line.

Figure 15. Line 1010 shore perpendicular running west to east: 700-m/s velocity interpreted as wood-peat surface, 1200-m/s velocity
contour interpreted as peaty-mud surface, and 2100-m/s velocity contour interpreted as bedrock surface.

Avalos et al. 191

 Figure 16. Final three-dimensional geologic model of the bedrock surface at Hallsands. The English Channel is the deep area to the east
that trends north. Field observations, which helped constrain the model, were taken on the elevation of the bedrock exposed north and south
of Hallsands beach.

487 project is capable of gathering reflection or refraction REFERENCES CITED 518

488 data in both two dimensions and three dimensions. A
489 significant increase occurs, however, in the amount of Al-Saigh, N. H., and T. H. Al-Dabbagh, 2010, Identification of landslide 519
490 time, money, and equipment when collecting 2-D or slip-surface and its shear strength: A new application for shallow 520
seismic refraction method: Journal of the Geological Society of 521
491 3-D reflection. Additional costs include greater num- India, v. 76, p. 175–180, doi:10.1007/s12594-010-0092-y. 522
492 bers of cables, geophones, and data-gathering hardware Brooks, M., B. V. Hillier, and M. Miliorizos, 1993, New seismic evi- 523
493 that were beyond the available budget for this project. dence for a major geological boundary at shallow depth, N 524
Devon: Journal of the Geological Society, v. 150, p. 131–135, 525
494 Data collection and processing time increase as well
doi:10.1144/gsjgs.150.1.0130. 526
495 from time to completion measured in months rather Coward, M. P., and K. R. McClay, 1983, Thrust tectonics of S Devon: 527
496 than days. Acquiring data with 2-D refraction lines Journal of the Geological Society, v. 140, p. 215–228, doi: 528
497 alone decreases equipment cost significantly as well as 10.1144/gsjgs.140.2.0215. 529
Doody, J. J., and M. Brooks, 1986, Seismic refraction investigation of 530
498 time collecting data in the field and even more time the structural setting of the Lizard and Start complexes, SW 531
499 processing data in the lab. England: Journal of the Geological Society, v. 143, p. 135–139, 532
500 If a project involves detailing stratigraphy, the doi:10.1144/gsjgs.143.1.0135. 533
Dornbusch, U., R. B. G. Williams, C. Moses, and D. A. Robinson, 534
501 expense and time necessary for reflection are justi-
2002, Life expectancy of shingle beaches: Measuring in situ 535
502 fied, and reflection methods are the appropriate abrasion, in J. A. G. Cooper and D. W. T. Jackson, eds., Coastal 536
503 venue. For mapping a bedrock valley, however, re- Education and Research Foundation Journal of Coastal Research 537
504 fraction data have been shown as not only compa- Special Issue 36: 7th International Coastal Symposium, Tem- 538

505 rable but in many instances superior to reflection for

plepatrick, Northern Ireland, March 25–29, 2002, p. 249–255. Q:11 539
Eddies, R. D., and J. M. Reynolds, 1988, Seismic characteristics of 540
506 obtaining a good representation of bedrock topogra- buried rock-valleys in Plymouth Sound and the River Tamar: 541
507 phy (Wolfe and Richard, 1996; Hirsch et al., 2008). Proceedings of the Ussher Society, v. 7, p. 36–40. 542
508 The quality and accuracy of refraction data increase further Gehrels, W. R., and W. P. Anderson Jr., 2014, Reconstructing 543
Holocene sea-level change from coastal freshwater peat: A 544
509 given advances in refraction tomography and computer combined empirical and model-based approach: Marine Geo- 545
510 software applications in recent years. What began as an logy, v. 353, p. 140–152, doi:10.1016/j.margeo.2014.04.004. 546
511 investigation of bedrock geology to determine boundary Gehrels, W. R., and A. J. Long, 2008, Sea level is not level: The case for 547
a new approach to predicting UK sea-level rise: Geography 548
512 conditions for a hydrogeologic flow model evolved (Sheffield, England), v. 93, no. 1, p. 11–16. 549
513 into the development of a new approach for subsurface Hartz, M. A., D. H. Malone, and R. S. Nelson, 2016, Three- 550
514 mapping and interpretation. In essence, merging seismic dimensional modeling of a glacial valley train outwash deposit 551
using two-dimensional seismic refraction techniques: Geo- 552
515 refraction with seismic reflection 3-D interpretation soft-
sciences, v. 6, p. 1–15. 553
516 ware provides a tool that is effective and inexpensive for Heincke, B., H. Maurer, A. G. Green, H. Willenberg, T. Spillmann, 554
517 mapping complex subsurface bedrock geology. and L. Burlini, 2006, Characterizing an unstable mountain slope 555

192 Two-Dimensional Seismic Refraction Tomography of a Buried Bedrock Valley

556 using shallow 2D and 3D seismic tomography: Geophysics, Massey, A. C., W. R. Gehrels, D. J. Charman, G. A. Milne, W. R. Peltier, 589
557 v. 71, no. 6, p. B241–B256, doi:10.1190/1.2338823. K. Lambeck, and K. A. Selby, 2008, Relative sea-level change and 590
558 Higuera-Dı́az, C. I., P. J. Carpenter, and M. D. Thompson, 2007, postglacial isostatic adjustment along the coast of south Devon, 591
559 Identification of buried sinkholes using refraction tomography at United Kingdom: Journal of Quaternary Science, v. 23, no. 5, 592
560 Ft. Campbell Army Airfield, Kentucky: Environmental Geology, p. 415–433, doi:10.1002/jqs.1149. 593
561 v. 53, p. 805–812, doi:10.1007/s00254-007-0693-y. Min, Dong-Joo, and C. Shin, 2006, Refraction tomography using 594
562 Hirsch, M., L. R. Bentley, and P. Dietrich, 2008, A comparison of a waveform-inversion back-propagation technique: Geophysics, 595
563 electrical resistivity, ground penetrating radar and seismic re- v. 71, no. 3, p. R21–R30, doi:10.1190/1.2194522. 596
564 fraction results at a river terrace site: Journal of Environmental & Robinson, A. H. W., 1961, The hydrography of Start Bay and its re- 597
565 Engineering Geophysics, v. 13, no. 4, p. 325–333, doi:10.2113 lationship to beach changes at Hallsands: Geographical Journal, 598
566 /JEEG13.4.325. v. 127, no. 1, p. 63–77, doi:10.2307/1793197. 599
567 Karastathis, V. K., A. Ganas, J. Makris, J. Papoulia, P. Dafnis, Schuster, G. T., and A. Quintus-Bosz, 1993, Wavepath eikonal 600
568 E. Gerolymatou, and G. Drakatos, 2007, The application of traveltime inversion: Theory: Geophysics, v. 58, p. 1314–1323, 601
569 shallow seismic techniques in the study of active faults: doi:10.1190/1.1443514. Q:12
602
570 The Atalanti normal fault, central Greece: Journal of Applied Tsai, C. C., R. D. Catchings, M. R. Goldman, M. J. Rymer, 603
571 Geophysics, v. 62, p. 215–233, doi:10.1016/j.jappgeo.2006 P. Schnurle, and H. W. Chen, 2008, High-resolution seismic 604
572 .11.004. images and seismic velocities of the San Andreas Fault Zone at 605
573 Kelland, N. C., 1975, Submarine geology of Start Bay determined by Burro Flats, Southern California: Bulletin of the Seismological 606
574 continuous seismic profiling and core sampling: Journal of the Society of America, v. 98, no. 6, p. 2948–2961, doi:10.1785 607
575 Geological Society, v. 131, p. 7–17, doi:10.1144/gsjgs.131.1.0007. /0120060252. Q:13
608
576 Lankston, R. W., 1989, The seismic refraction method: A viable tool Ussher, W. A. E., 1905, Geology of the Kingsbridge and Salcombe 609
577 for mapping shallow targets into the 1990’s: Geophysics, v. 54, District: Geological Magazine, v. 2, p. 91–92, doi:10.1017 610
578 p. 1535–1542, doi:10.1190/1.1442621. /S0016756800129395. Q:14
611
579 Lenham, J. W., R. McDonald, S. Miller, and J. M. Reynolds, 2005, Wolfe, P. J., and B. H. Richard, 1996, Integrated geophysical studies of 612
580 Integrated seismic investigations across the Mersey Estuary, buried valley aquifers: Journal of Environmental & Engineering 613
581 Halton District, UK: Quarterly Journal of Engineering Geophysics, v. 1, no. 1, p. 75–84, doi:10.4133/JEEG1.1.75. 614
582 Geology and Hydrogeology, v. 38, p. 7–22, doi:10.1144 Zelt, C. A., A. Azaria, and A. Levander, 2006, 3D seismic refraction 615
583 /1470-9236/04-034. traveltime tomography at a groundwater contamination site: 616
584 Maguire, P., R. England, and A. Hardwick, 2011, LISPB Delta, Geophysics, v. 71, no. 5, p. H67–H78, doi:10.1190/1.2258094. 617
585 a lithospheric seismic profile in Britain: Analysis and inter- Zhao, Z., and J. Xu, 2010, Geological structure investigation of 618
586 pretation of the Wales and southern England section: Journal of shallow layers by the explosion seismic survey tomographic 619
587 the Geological Society, v. 168, p. 61–82, doi:10.1144/0016- technique: Journal of Environmental & Engineering Geophysics, 620
588 76492010-030. v. 15, no. 1, p. 21–28. Q:15
621

Avalos et al. 193

 AUTHOR QUERIES

AUTHOR PLEASE ANSWER ALL QUERIES 1

Q:1 Please confirm or amend short title “Two-Dimensional Seismic Refraction Tomography
of a Buried Bedrock Valley.”
Q:2 Per AAPG style, please add English unit conversions for metric values.
Q:3 Please confirm or amend the mailing address for Illinois State University.
Q:4 Please add mailing address (street name and number, city, state, zip code [plus country if
outside US]) for the affiliations at Appalachian State University and York University.
Q:5 Car park was changed to the more common American term parking lot for clarity
throughout the paper, per AAPG style.
Q:6 England was changed to United Kingdom here for consistency with the rest of the paper.
Please confirm.
Q:7 Please confirm that citation Coward, 1983 should be Coward and McClay, 1983; if these
are separate references, please ensure that both are cited in the text and included in the
References Cited section.
Q:8 Please confirm that GPR means ground-penetrating radar or amend in sentence “They
found…”
Q:9 Please confirm or amend change from one-third to 0.33 per AAPG style.
Q:10 Please confirm or amend full term for ASCII.
Q:11 Please carefully review and confirm or amend changes to reference Dornbusch et al.
2002.
Q:12 Please confirm or amend page range for reference Schuster, Quintus-Bosz, 1993.
Q:13 Please confirm or amend author list for reference Tsai et al. 2008.
Q:14 Please confirm or amend page range for reference Ussher 1905.
Q:15 Please confirm or amend journal name for reference Zhao and Xu 2010.
Q:16 If possible, can you please add a scale bar to Figure 2?
Q:17 Please confirm that written permission was received to reproduce photo from Gehrels
and Anderson, 2014.
Q:18 Please confirm or amend abbreviation definition added to Figure 3 caption.
Q:19 Per AAPG style, Figures must be able to stand alone. Please include the necessary in-
formation from Figures 1 and 2 or remove the reference.
Q:20 Please add a description for C in Figure 8 caption.
 AUTHOR QUERIES

AUTHOR PLEASE ANSWER ALL QUERIES 2

Q:21 Figures 14–16 seem blurry compared with the earlier figures; can you please provide us
with higher resolution figures if possible?

View publication stats