 | |  | |  | |  | |  | |  | |  | |  | |  | |  | |  |
Finally, not mentioned above but certainly pertinent, is the fact that in some exposures the upland deposits
include “extremely large boulders ranging from [~0.02 m3] to [3.7 m3] (Schlee, 1957).” I have observed a
great many such boulders in a broad area south of Alexandria, Va. The conventional explanation of these
boulders, ice rafting, certainly must be called into question, given that the ice ages of the Quaternary are
associated with extreme sea-level lowstands (Lambeck and Chappell, 2001). How then could boulders have
been ice rafted to upland locations?
Some new lithological data from Northern Virginia
I present here some new lithological data bearing on the question of the origin of the upland deposits. As a
quasi-random sample of the coarser upland gravels in our former neighborhood of Hollin Hills, Fairfax
County, Va., I investigated 214 quartzite pebbles and cobbles already raked up by a neighbor (Fig. 3A). I
assume that he likely suppressed smaller pebbles and possibly set aside larger cobbles, but at least this
collection has no investigator-imposed biases. In particular, I decided to count the number fractured rocks
in each size class. Then, since many of the samples were evidently fractured along two or more intersecting
planes, I separately noted the number of rock fragments with more than one non-parallel fracture. For this
purpose, I counted as “second fractures” any planar fractures running part way through any fragment of
what was obviously once a well-rounded rock prior to the first fracture (the one which separated it from the
rest of the original rock). However, I did not count as second fractures any cracks that were parallel to the
surface of the first complete fracture, even though these occurred very commonly. In all, I noted only one
rock fragment exhibiting the slightest evidence of post-fracture re-rounding (and I counted that one as not
fractured). My results are shown in Figure 3B. Overall, 54.5% of the 32-64-mm fraction, 65% of the 64-
128-mm fraction, and 57% of the 128-256-mm fraction had at least a single fracture. The advanced surface
weathering of these rocks rules out the fractures being of very recent origin. In any event, it is clear that
the present condition and location of these rocks (high prevalence of fresh fractures and lying atop a 60 m
clay terrace) are inconsistent with fluvial deposition.
include “extremely large boulders ranging from [~0.02 m3] to [3.7 m3] (Schlee, 1957).” I have observed a
great many such boulders in a broad area south of Alexandria, Va. The conventional explanation of these
boulders, ice rafting, certainly must be called into question, given that the ice ages of the Quaternary are
associated with extreme sea-level lowstands (Lambeck and Chappell, 2001). How then could boulders have
been ice rafted to upland locations?
Some new lithological data from Northern Virginia
I present here some new lithological data bearing on the question of the origin of the upland deposits. As a
quasi-random sample of the coarser upland gravels in our former neighborhood of Hollin Hills, Fairfax
County, Va., I investigated 214 quartzite pebbles and cobbles already raked up by a neighbor (Fig. 3A). I
assume that he likely suppressed smaller pebbles and possibly set aside larger cobbles, but at least this
collection has no investigator-imposed biases. In particular, I decided to count the number fractured rocks
in each size class. Then, since many of the samples were evidently fractured along two or more intersecting
planes, I separately noted the number of rock fragments with more than one non-parallel fracture. For this
purpose, I counted as “second fractures” any planar fractures running part way through any fragment of
what was obviously once a well-rounded rock prior to the first fracture (the one which separated it from the
rest of the original rock). However, I did not count as second fractures any cracks that were parallel to the
surface of the first complete fracture, even though these occurred very commonly. In all, I noted only one
rock fragment exhibiting the slightest evidence of post-fracture re-rounding (and I counted that one as not
fractured). My results are shown in Figure 3B. Overall, 54.5% of the 32-64-mm fraction, 65% of the 64-
128-mm fraction, and 57% of the 128-256-mm fraction had at least a single fracture. The advanced surface
weathering of these rocks rules out the fractures being of very recent origin. In any event, it is clear that
the present condition and location of these rocks (high prevalence of fresh fractures and lying atop a 60 m
clay terrace) are inconsistent with fluvial deposition.
Figure 3. A: Pile of quartzite pebbles and cobbles from the upland deposits that were sorted by
size and fracture frequency. B: Histogram of the results. C: A dramatically fractured cobble
from the same residential lot in the Hollin Hills subdivision of Faifax County, Va. D: Cartoon
typical of the fracture patterns of fused silica optical fibers broken under high tensile stress.
size and fracture frequency. B: Histogram of the results. C: A dramatically fractured cobble
from the same residential lot in the Hollin Hills subdivision of Faifax County, Va. D: Cartoon
typical of the fracture patterns of fused silica optical fibers broken under high tensile stress.
Figure 3C exhibits a quartzite cobble that I noticed elsewhere in the same residential yard (not in the pile of
Fig. 3A). The fracture facies of this rock are reminiscent of the way glassy materials fracture under tensile
stress. In fact, defect-free fused-silica optical fibers can survive up to 5 Giga Pascal tensile stress before
fracturing (e.g., Bradt and Tressler, 1994); and when fracture finally occurs at such high tensile stress levels
it leaves a characteristic “mirror-mist-hackle” fracture surface, as cartooned on Figure 3D. Even though
optical fibers are typically ~50 to 500 in diameter (producing mirror-mist-hackle patterns analyzable only
under a microscope), it seems reasonable to interpret the macroscopic fracture surfaces of this
metaquartzite rock (whose fracture strength is also determined by the strength of silicon-oxygen bonds) as
possible indications of fracturing under extreme tensile stress. While there does not seem to be a mist zone
on the rock of Figure 3C, the mirror and hackle are unmistakable. In fact, the surface enclosed by the
smaller dashed ellipse is much more mirror-like than one would expect for metaquartzite, which is normally
subject to “uneven, splintery to conchoidal fracture (Chesterman and Lowe, 1978).” Unlike any other part
of this particular rock, the “mirror” area exhibits vitreous luster and is so flat that a straight edge laid across
it is within a fraction of a millimeter of the surface over the entire 7-cm width. Moreover, there is an
internal planar fracture sub-perpendicular to the “mirror” that runs about two thirds of the way through the
entire rock fragment; its intersection with the mirror surface (indicated by arrows) parallels the semi-major
axis of the ellipse defining the mirror. I believe it will one day be confirmed that these unusual features
resulted from two nearly orthogonal tensile waves having passed through this cobble, very likely at the same
moment.
N.B. Tensile waves result from reflection of pressure (shock) waves at free surfaces (e.g., Melosh, 1989).
Thus, for example, any pressure wave reflected from the surface of bed of water-saturated sediments will
propagate backward into the bed as a tensile wave. And tensile waves of sufficient intensity to rupture a
metaquartzite cobble surely occur in nature only as results of large impacts.
CALCULATION OF THE EJECTA BLANKET PROFILE
To aid in the search for proximal ejecta deposits of degraded or buried terrestrial craters, it is helpful to have
a quantitative idea of what the ejecta blanket might have looked like when it was first emplaced. An
empirical rule has emerged from target-strength-governed explosion experiments (McGetchen et al., 1973),
which by very good fortune turns out to be applicable as well to gravity-dominated impact craters ranging
from 1.3 km to 436 km (Melosh, 1989). The rule is that the heights h(r) of ejecta blankets outside of the
crater rim are typically proportional to the negative third power of the radius r from the crater center:
h(r) = f(R)r-3 (r > R), (1)
where R is the transient-crater radius and the coefficient f(R) depends in principle on specific details of the
crater in question (Melosh, 1989). The total volume of ejecta Vej present in such an ejecta blanket can be
obtained by taking the area under a surface of revolution with the radial profile defined by equation 1 by
integrating from r = R and to r = :
Whence, f(R) = RVej/. So, in essence, if one knows the crater radius R and the total volume of the ejecta
blanket, then f(R) is also known. However, the value of Vej is uncertain to the degree that the amount
substrate incorporation is not easily determined (Oberbeck,, 1975, Melosh, 1989). Thus, by using the
amount of material excavated from the crater Vexc to approximate Vej, as I will do below, I obtain a lower
limit to the actual value of f(R).
Following the general procedures adopted by Warren et al. (1996), I arrived at a value of Vexc = 4,508 km3
for the Chesapeake Bay crater. However, I defer here to a similar number given by Poag (1997): 4,300
km3. I used the final crater radius R = 45 km (Koeberl et al., 1996; Poag, 1997) to arrive at the number I
needed to calculate the (minimum) thickness of the Chesapeake Bay crater ejecta blanket:
Next, to lend “realism” to my ejecta-blanket plots based on equation 1 and the scaling coefficient of equation
3, I decided to include a profile representing the crater itself. For this purpose I approximated the crater as a
cylindrical hole in the ground of radius R having a depth adjusted to make its capacity exactly equal to the
excavated volume Vexc. The depth calculated in this way turns out to be exactly equal to the height of the
rim above the target surface. This cylindrical profile represents neither that of the transient crater nor of the
final crater. By comparison with figures in (Koeberl et al., 1996) and (Poag, 1997), the result of Figure 4
might be viewed as a fictitious (never achieved) intermediate state before the collapse of an idealized (never
realized) Chesapeake Bay crater rim.
In Figure 4, I have cartooned the depth of the crystalline basement to match the depth at which the USGS-
NASA Langley Corehole at Hampton, Va., intercepted basement granite: 626.3 m (Horton et al., 2005). And
to add further realism, I decided the give the target surface a gentle seaward slope of 0.5 m/km – just about
the same as the present day slope of the base of the upland deposits seen in the Southern Maryland section
(Fig. 5).
Except for the range r = 45±20 km, where details the final crater are badly represented, the calculated profile
Figure 4 should paint a fairly accurate picture of the lower-limit average Chesapeake Bay crater ejecta
blanket (neglecting dunes, hummocks, ridges, and rays) when it was only tens of minutes old and not yet
assaulted by the resurge of the Atlantic Ocean – which, when the return wave finally appeared, surely
carried the tallest part of the seaward rim back into the crater (Poag, 1997). The landward side, however,
should have fared far better. The volume of water to the landward at the time of the impact would have
been locally comparable to the volume of the landward solid ejecta, and this water would have been imparted
radial momentum causing most of the westward-moving wave to deflect northeastward or southward,
depending on the angle at which it struck the Blue Ridge. So the westward ejecta blanket should have
experienced little or no return wave from the west and would have been shielded from the primary Atlantic
resurge by the eastward facing blanket. Therefore, I argue that the landward crater rim should have
remained in at least partially intact for a geologically substantial period of time.
In fact, I will propose that remnants of this ejecta blanket are still there. From my juxtaposition of the
calculated Chesapeake Bay crater ejecta blanket profile on the U.S. Geological Survey sections in Figure 5, it
can be seen that the upland deposits of Southern Maryland could in principle comprise part of these
remnants. That they actually exceed the calculated ejecta depth at the longest distances is not a
contradiction, since my calculation neglects the inevitable incorporation of (soft or friable) substrate material
into the ejecta blanket (Oberbeck, 1975).
Fig. 3A). The fracture facies of this rock are reminiscent of the way glassy materials fracture under tensile
stress. In fact, defect-free fused-silica optical fibers can survive up to 5 Giga Pascal tensile stress before
fracturing (e.g., Bradt and Tressler, 1994); and when fracture finally occurs at such high tensile stress levels
it leaves a characteristic “mirror-mist-hackle” fracture surface, as cartooned on Figure 3D. Even though
optical fibers are typically ~50 to 500 in diameter (producing mirror-mist-hackle patterns analyzable only
under a microscope), it seems reasonable to interpret the macroscopic fracture surfaces of this
metaquartzite rock (whose fracture strength is also determined by the strength of silicon-oxygen bonds) as
possible indications of fracturing under extreme tensile stress. While there does not seem to be a mist zone
on the rock of Figure 3C, the mirror and hackle are unmistakable. In fact, the surface enclosed by the
smaller dashed ellipse is much more mirror-like than one would expect for metaquartzite, which is normally
subject to “uneven, splintery to conchoidal fracture (Chesterman and Lowe, 1978).” Unlike any other part
of this particular rock, the “mirror” area exhibits vitreous luster and is so flat that a straight edge laid across
it is within a fraction of a millimeter of the surface over the entire 7-cm width. Moreover, there is an
internal planar fracture sub-perpendicular to the “mirror” that runs about two thirds of the way through the
entire rock fragment; its intersection with the mirror surface (indicated by arrows) parallels the semi-major
axis of the ellipse defining the mirror. I believe it will one day be confirmed that these unusual features
resulted from two nearly orthogonal tensile waves having passed through this cobble, very likely at the same
moment.
N.B. Tensile waves result from reflection of pressure (shock) waves at free surfaces (e.g., Melosh, 1989).
Thus, for example, any pressure wave reflected from the surface of bed of water-saturated sediments will
propagate backward into the bed as a tensile wave. And tensile waves of sufficient intensity to rupture a
metaquartzite cobble surely occur in nature only as results of large impacts.
CALCULATION OF THE EJECTA BLANKET PROFILE
To aid in the search for proximal ejecta deposits of degraded or buried terrestrial craters, it is helpful to have
a quantitative idea of what the ejecta blanket might have looked like when it was first emplaced. An
empirical rule has emerged from target-strength-governed explosion experiments (McGetchen et al., 1973),
which by very good fortune turns out to be applicable as well to gravity-dominated impact craters ranging
from 1.3 km to 436 km (Melosh, 1989). The rule is that the heights h(r) of ejecta blankets outside of the
crater rim are typically proportional to the negative third power of the radius r from the crater center:
h(r) = f(R)r-3 (r > R), (1)
where R is the transient-crater radius and the coefficient f(R) depends in principle on specific details of the
crater in question (Melosh, 1989). The total volume of ejecta Vej present in such an ejecta blanket can be
obtained by taking the area under a surface of revolution with the radial profile defined by equation 1 by
integrating from r = R and to r = :
Whence, f(R) = RVej/. So, in essence, if one knows the crater radius R and the total volume of the ejecta
blanket, then f(R) is also known. However, the value of Vej is uncertain to the degree that the amount
substrate incorporation is not easily determined (Oberbeck,, 1975, Melosh, 1989). Thus, by using the
amount of material excavated from the crater Vexc to approximate Vej, as I will do below, I obtain a lower
limit to the actual value of f(R).
Following the general procedures adopted by Warren et al. (1996), I arrived at a value of Vexc = 4,508 km3
for the Chesapeake Bay crater. However, I defer here to a similar number given by Poag (1997): 4,300
km3. I used the final crater radius R = 45 km (Koeberl et al., 1996; Poag, 1997) to arrive at the number I
needed to calculate the (minimum) thickness of the Chesapeake Bay crater ejecta blanket:
Next, to lend “realism” to my ejecta-blanket plots based on equation 1 and the scaling coefficient of equation
3, I decided to include a profile representing the crater itself. For this purpose I approximated the crater as a
cylindrical hole in the ground of radius R having a depth adjusted to make its capacity exactly equal to the
excavated volume Vexc. The depth calculated in this way turns out to be exactly equal to the height of the
rim above the target surface. This cylindrical profile represents neither that of the transient crater nor of the
final crater. By comparison with figures in (Koeberl et al., 1996) and (Poag, 1997), the result of Figure 4
might be viewed as a fictitious (never achieved) intermediate state before the collapse of an idealized (never
realized) Chesapeake Bay crater rim.
In Figure 4, I have cartooned the depth of the crystalline basement to match the depth at which the USGS-
NASA Langley Corehole at Hampton, Va., intercepted basement granite: 626.3 m (Horton et al., 2005). And
to add further realism, I decided the give the target surface a gentle seaward slope of 0.5 m/km – just about
the same as the present day slope of the base of the upland deposits seen in the Southern Maryland section
(Fig. 5).
Except for the range r = 45±20 km, where details the final crater are badly represented, the calculated profile
Figure 4 should paint a fairly accurate picture of the lower-limit average Chesapeake Bay crater ejecta
blanket (neglecting dunes, hummocks, ridges, and rays) when it was only tens of minutes old and not yet
assaulted by the resurge of the Atlantic Ocean – which, when the return wave finally appeared, surely
carried the tallest part of the seaward rim back into the crater (Poag, 1997). The landward side, however,
should have fared far better. The volume of water to the landward at the time of the impact would have
been locally comparable to the volume of the landward solid ejecta, and this water would have been imparted
radial momentum causing most of the westward-moving wave to deflect northeastward or southward,
depending on the angle at which it struck the Blue Ridge. So the westward ejecta blanket should have
experienced little or no return wave from the west and would have been shielded from the primary Atlantic
resurge by the eastward facing blanket. Therefore, I argue that the landward crater rim should have
remained in at least partially intact for a geologically substantial period of time.
In fact, I will propose that remnants of this ejecta blanket are still there. From my juxtaposition of the
calculated Chesapeake Bay crater ejecta blanket profile on the U.S. Geological Survey sections in Figure 5, it
can be seen that the upland deposits of Southern Maryland could in principle comprise part of these
remnants. That they actually exceed the calculated ejecta depth at the longest distances is not a
contradiction, since my calculation neglects the inevitable incorporation of (soft or friable) substrate material
into the ejecta blanket (Oberbeck, 1975).
In Plain Sight
 |  |  |