Plate 1. View looking south from the northern end of Beaver Dam reservoir, which serves as a secondary water reserve for Loudon County, Virginia. According to signs posted around the shore, it was drained for maintenance. This photo shows an outcrop of Jurassic diabase that is unusually leucocratic (light colored minerals).

Plate 2. Geologic map of the area around Beaver Dam Reservoir. The outcrop seen in Plate 1 is a high-titanium, quartz-normative tholeiitic diabase, which occurs in dikes and differentiated sheets throughout the map area labeled as Jdh on the map (Jurassic age). The high titanium content and available quartz indicate that this intrusive rock originated within a larger magma chamber (deeper within the crust) and multiple intrusions were emplaced over a period of time as the chemistry evolved. Tholeiitic magma is associated with mid-ocean ridges, which indicates a mantle source rather than melted crustal rocks. A different diabase (Jdg on the map) is younger and was injected into the pre-existing diabase (Jdg) as granophyre, an intrusive rock that indicates a highly evolved magma chamber. These intrusive rocks were injected along bedding planes, faults, and fractures within the Jurassic-Triassic Bull Run Siltstone (JTrtm on the map). Two other Jurassic diabase units (Jdl and Jd) are indicated by circled areas where they cut across Triassic sedimentary rocks (lower area where Jd cuts JTrtm) and Jurassic diabase (upper circled area where Jdl cuts Jdh).

Plate 3. View of the reservoir showing the man-made shoreline and boulders scattered on a sandy-muddy substrate. The title of this post is intended as geological humor disguised as a newspaper headline. I wondered why the reservoir was so low.

Plate 4. This image (approximately 2 feet across) reveals a medium-sized crystalline texture and color similar to granite, which was the initial field identification. An important difference between these rocks and granite is the lower quartz and alkali feldspar content, which isn’t visible in this exposure. Note at least one set of joints, indicated by the weathered “X” rotated slightly to the left, in the center of the photo; then let your eye go down a little and you will notice another X, this time rotated to the right. This second X is repeated throughout the exposure. This type of joint is associated with fracturing of the rock as it cools and pressure is reduced because of erosion of overlying rocks. The suggestion of multiple patterns (I admit it isn’t that obvious) implies several steps in cooling; however, interpreting joints is complex, requiring many detailed measurements, and beyond this post. I just wanted to mention it because joints tell us about the geologic history of a rock after its formation, millions of years later.

Plate 5. This photo shows a large outcrop at the high-water mark of the reservoir. Note the angular structure of the outcrop (lighter rock at the upper left of the image) and the pieces broken off during construction of the reservoir.

Plate 6. View looking north towards the outflow gate of Beaver Dam reservoir, showing the weirs and maximum water level (about fifteen feet above present level). The low water level suggests that evaporation and ground-seepage (not much with the subsurface comprising diabase) exceed local run-off. This is the reason for the post title. Rainfall has been low enough that the water level keeps dropping, even after the reservoir was emptied. Is it a drought?

The last post reported a generic Jurassic diabase (Jd) west of the Bull Run fault (BRF), less than 10 miles NNW of Beaver Dam reservoir, which was intruded into Precambrian sediments. However, Jd also occurs as small intrusions throughout Loudon County (not shown). In other words, it wasn’t emplaced as sills or sheets, but rather filled fault and other fractures in older rocks. Using general principles of stratigraphy (e.g. Steno’s Laws), we can speculate about what we’ve seen in these two recent field trips.

The unspecified Jurassic diabase dike we saw at Morven Park (Jd in Plate 1 of the last post) cuts through Proterozoic sediments but wasn’t seen east of Bull Run fault in that area. The younger (Triassic) sedimentary rocks in the area of Beaver Dam reservoir (JTrtm in Plate 2), as well as Jurassic intrusive rocks (e.g. Jdh on Plate 2) are cut by dikes of diabase (Jd and Jdl, circled areas in Plate 2) that suggest the continuous chemical fractionation of a magma chamber, which produced smaller amounts of magma that had less space to fill.

I propose that magma with a composition like most of the world’s ocean floor formed (tholeiitic basalt, or Mid-Ocean-Ridge Basalt–MORB) beneath the oldest rocks in the area (more than 500 million years old) when Pangea was stretched by upper mantle convection during the early Jurassic (about 200 million years ago), sending tentacles of molten rock to fill every weak point in the overlying rock, sheets and dikes were created, possibly even laccoliths, between layers of sediments. This stage created the large area of diabase in the study area (Jdh in Plate 2). As the magma chamber lost material and cooled, it injected smaller volumes of magma into even smaller fissures and weak points in the overlying rock, including earlier diabase. These late-stage injections are seen as dikes of Jd and Jdl in Plate 2.

Finally, the crust throughout this area stretched to the breaking point and Bull Run fault formed, with the east side sliding downward and to the east. All of the Paleozoic and Mesozoic rocks on the west side of Bull run fault, including diabase sills and dikes, were eroded by wind, water, and ice, leaving only the final, highly fractionated late-stage magmatic dikes (e.g. the granophyre of Plate 2) protruding out of the Precambrian sediments. The deep source of all of these diabase plutonic rocks remains buried deep beneath western Loudon County.

Finally … the only Jurassic diabase I found on the USGS geologic map of Loudon County occurs as thin exposures parallel to Bull Run fault and within a mile of it, which suggests that BRF defines the western limit of the fault zone associated with the break-up of Pangea.

That’s my story and I’m sticking to it …

Plate 1. Geologic map of study area in northern Virginia (see inset map). The area is bisected by the Bull Run Fault (BRF, dash line), which separates older metasedimentary rocks from younger sedimentary rocks. The west side of BRF has moved upward relative to the east side, following the regional trend along the east coast of North America (dash line in inset map). The inset photo shows how BRF appears today, forming a topographic rise with less than 100 feet of relief. The study area (blue circle) is located on the western part of Morven Park, which is the location of a mansion (inset photo) occupied by an early twentieth century governor of Virginia, now a historic site, museum, and equestrian park. The image is approximately 4.5 miles across.

Plate 2. photos of an outcrop of Catoctin Formation metamorphic rocks from the southern end of the study area (see Plate 1). These rocks are between 1 by and 540 my old; they were originally basalts, tuffs, sandstone, siltstone; before being buried and metamorphosed into their current lithologies. This plate shows the outcrop along strike (A) and along dip (B), revealing sedimentary structures and grain sizes that suggest this is a cross-bedded sandstone with intercalated siltstone. These sediments (and associated volcanics not seen in the study area) were deposited when proto-North America collided with proto-Europe during the late Proterozoic. They were then buried deeply enough to be altered but not so much to become gneiss, or melt to become igneous rocks. Their current orientation has a strike of 35 east of north, and a dip of approximately 30 degrees, but this deformation was not caused during the collision that formed Pangea.

Plate 3. Close-up of the outcrop in Plate 2. The top of the sequence is a bed 12 inches thick. Below this is are several cross-bedded layers (identified by the lines that dip to the left) that are discontinuous, and intercalated with thin, massive (no lamination) beds. The lowest visible beds are lenticular in this view. This kind of cross-bedding suggests that these sandy sediments were deposited in a river, where one-directional currents create uniform cross-beds. There is no evidence of gravel and the sand is fairly well sorted (as best as I could tell from the outcrop), suggesting that this was not near the source but in an alluvial fan. The heterogeneous lithologies of the Catoctin Formation are likely due to delta switching, i.e., the main channel moving across a relatively flat area before entering either a sea or lake. There are no fossils in these rocks because they predate the appearance of shell-forming invertebrates like clams, snails, etc. There were no land plants either, so their organic carbon content is practically zero.

Plate 4. Close-up from the outcrop in Plate 3, showing lenses of white minerals within laminated, slightly folded sandstone beds. Such lenticular bedding is common in metamorphic rocks because of the high heat and pressure caused by deep burial. Incompatible elements (e.g., calcium or silicon) are squeezed out of the rocks and form blebs of new minerals, such as calcite (excess calcium) or quartz (excess silicon). There were no shell-forming animals (invertebrates form their shells of calcium minerals) when these sediments were deposited, but carbonate rocks have been produced by abiotic processes as long as 4 billion years ago; not to mention algal mats created by stromatolites. Marble (metamorphosed carbonate rock) is reported as lenses within the Catoctin Formation. These lenses would have been originally deposited as either algal mats or chemical sediments.

Plate 5. Close-up from Plate 3, showing details of one of the lenses. Note the lamination in the sandstone and transition between the two mineralogies where they are in contact in middle of the photo. I didn’t use acid to test for calcite (it fizzes under dilute HCL) because the motto of Rocks and (no) Roads is to use our eyes and available information. However, note the white rectangle at the top-center of the photo; it is very similar to the shape of calcite crystals. Other such shapes are visible if you open the image, zoom in, and pan around. I am going with this being a lens of calcite, crystallized from mineralogically incompatible calcium, because that is consistent with the official description of the Catoctin Formation lithologies.

Plate 6. Photos of phyllite found loose on the ridge west of the Bull Run Fault (see Plate 1). (A) The sheen of this sample is caused by aligned muscovite minerals during burial. (B) A close-up reveals the platy texture typical of phyllite, which typically forms from shale and is intermediate in metamorphic grade between slate and schist. The chemistry of these rocks suggests that they were originally deposited as tuff (a fine-grained volcanic deposit) rather than mud. Of course, once the ash settled it would have been transported by rivers and become intercalated with the sandstone seen in Plate 2. This sample shows no sign of stream transport (e.g. rounded into a cobble), so it is probably a remnant of eroded, overlying (i.e. younger) volcanic sediments, after the quiescent period represented by the older sediments from Plate 2.

Plate 7. Photo looking north at an outcrop of the Jurassic diabase (age ~175 my) indicated in Plate 1, exposed by a stream that cuts across the Bull Run Fault. These are the youngest rocks within the region after rifting of Pangea had begun. Deeply buried rocks melt and some of the magma rises, following joints and weak lines in the overlying, solid rocks. Diabase, which has a composition similar to basalt, is formed like this although it often feeds volcanoes. The continental crust was stretched thin and fractured, allowing the diabase to work its way towards the surface. It only appears as lenses like those seen in Plate 2 in this area.

Plate 8. This image shows the typical growth for trees along the rocky crest of the hill seen in Plate 1. These large trees started out growing in fractures in the rock and, in some kind of enhanced biochemical and physical weathering, thrived and grew to be tall trees, probably almost a hundred years old. I’ve seen trees growing from large joints in rocks before, but never a forest that looks like it is set in concrete. There is absolutely no soil on these ridges but that didn’t stop Mother Nature.

Plate 9. Cross-section across Loudon County, Virginia, from west to east. Bull Run Fault became active during the rifting of Pangea as the supercontinent stretched. The younger rocks on the east (right side of BRF) slipped down this fault surface, leaving the older rocks several thousand feet higher than where they belong stratigraphically (beneath the younger rocks). The arrows indicate the relative movement along BRF. The current erosional surface is indicated by the blue, horizontal line. This displacement has juxtaposed sedimentary and volcanic rocks that were created during the closing of an ocean basin (e.g. Plates 2-6), more than 600 my ago, with sedimentary and igneous rocks emplaced during the opening of another ocean basin, about 200 my ago (previous post ). The diabase intrusions (Plate 7) were emplaced during the latter event, cutting through rocks that weren’t much older than them (if you call 20 my a short time interval). The earth’s crust moves slowly over a semi-molten mantle, but it never stops moving back and forth; and the result is there to be seen if we’re looking for it.

The confluence of Sugarland Run (left) and the Potomac River (right). This is where today’s trip ends. The trees haven’t recovered their foliage yet because it is late March. We’re going to start upriver about a mile and end up here.

The last post covered the area inside the blue ellipse. We encountered 200 Ma shales and sandstones of the Balls Bluff shale. we expect to see similar rocks today, but we’ll be entering the Potomac flood plain. The red ellipse is where we are in this post. The numbers will be referred to later as the locations where photos were taken. The dashed lines for “rock” and “gravel” are approximate locations where the bed of Sugarland Run changed composition. It should be interesting.

Here at Site 1, the bed consists of large, angular boulders with rounded corners. These rocks didn’t travel far, probably eroded from now-gone cliffs like we saw upstream. This location, as with similar rocky transits we saw upstream, represents a point where the stream flows over an exposed ledge of bedrock. This is very common for streams in this area.

View looking downstream (north) at Site 1. Note the dramatic change in stream bed composition. The bar on the right consists of silt and gravel. Note also the eroded, soft bank on the left. We have entered the ancestral Potomac flood plain.

At Site 2 (see map above for location) gravel bars like this were found where a smaller stream entered Sugarland Run. It is probable that the current stream is cutting through ancient sediments because there is no source for gravel like this anywhere around. These are recycled deposits.

Confluence of Sugarland Run and a side channel at Site 2. Channels like this criss-cross the ancient flood plain. These are larger than those we saw further upstream on the Potomac in a previous post.

This photo from Site 2 is the last appearance of bedrock in the stream bed. Note the flat surface across the stream that tilts slightly towards the camera. This is a bedding surface for the Balls Bluff Siltstone. Upstream, these rocks form low cliffs and are tilted away from the modern stream. It is likely that overlying beds have been eroded after tens-of-millions of years by the ancestral Potomac River. The change in dip suggests, further, that there is a structural feature between this location and a mile upstream. There is some evidence for a fault that runs along Sugarland Run several miles upstream. It was probably part of regional adjustment during uplift over the last 200 my.

As with other streams on the Potomac River flood plain, there has been rapid erosion. This example from Site 3 can be dated by the age of the tree. I don’t know how old it is, but it is certainly less than a century. What is unusual is that this erosion is occurring inside a bend. Usually streams cut on the inside of a meander and deposit point bars on the outside. We’ve seen this at every scale in previous posts. From what I’ve read there has been rapid erosion in the last few decades because of urbanization. We saw an extreme example in the last post. The field data suggests that Sugarland Run is widening but not meandering. This is not a natural process in unconsolidated sediments like these. The ancestral Sugarland Run certainly does meander (see map above), but this rapid erosion unaccompanied by channel migration is not natural.

There are several small lakes near the modern Potomac River, such as this one (just north of the Site 3 label in the map above). Sugarland Run passes it within 100 yards, through unconsolidated muddy sediments. Features like this are difficult to understand because the age relationship between the stream and lake cannot be unambiguously identified through radiometric dating. Both developed in sediments of the same age, older than either feature. These lakes (see map above) don’t look like oxbow lakes. Given the common occurrence of depressions throughout the area, which form small ponds and lakes during the wet season, the geological fact that the Potomac floodplain has wandered across a wide swath of the area (see for example a previous post), and the lack of any outflow to a modern stream (see map), it is probable that these lakes represent undulations in the ancient flood plain and Sugarland Run is younger. It just happened to miss the lake as it cut down through the soft sediments without meandering.

This meander at Site 3 shows how Sugarland Run is becoming incised rather than following a typical meandering trajectory, as at Horseshoe Bend on the Colorado River. The scale is drastically smaller but the processes are similar; the stream lacks the energy to erode the banks and becomes “trapped”, so it cuts downward as the upriver source is uplifted relative to the outflow. In addition, this small stream appears to be widening, as seen in the eroded tree on the bank in a previous photo.

Another interesting feature we saw between Sites 3 and 4 was a couple of elevated flat surfaces like this one, seen in the center-left of the photo, about halfway between the current stream bed and surface. These benches were small in area (less than 100 feet) and at the current water level of the stream. My best guess (a common occurrence in geology) is that they were point bars when Sugarland Run was smaller and are relict features on the modern Potomac flood plain.

Here we are about 100 yards from the Potomac. There is no delta associated with Sugarland Run but there is a bar at its mouth (see first photo).

Sugarland Run is an intermediate-sized stream flowing into the Potomac River. Goose Creek is one of the larger ones, which supplies drinking water for the region, whereas Horsepen Run is a small one. Despite the difference in flow between these tributaries, they display similar geomorphic features (e.g. meandering, point bars, gravel and muddy beds, recent erosion and entrenchment) because they all cross the wide, ancient Potomac floodplain composed of mixed sediment types. The modern Potomac River itself is less than four-million years old although there is evidence of the ancestral river flowing though this area back 20 my. The supply of sediment has decreased over the eons as the ancestral Appalachian Mountains eroded, so we don’t see the kind of sedimentation today that would have been occurring several million years ago.

The sediments being eroded by modern streams like Sugarland Run record a climate and topography very different from what we see today. However, the physical processes were the same and the landscape was shaped, ultimately, by geological processes occurring deep within the earth’s crust. These same constraints produced the ice age that is closing in our times and associated fluctuations in sea level, adding nuances and new themes to the unfolding story of our Earth.

This adventure followed Sugarland Run (aka Creek) a mile or so, where it flows through a wide spot between ledges of sandstone. The area has been developed for a long time and the creek is crossed by weirs (white water in this photo). I don’t know if they were to maintain water levels or as roads.

This geologic map from Rock-D shows the starting point our walk (blue circle), which ends at Route 1582 (Algonkian Pkwy). The pink rock seen to the left of the creek (tan color running N-S) is Balls Bluff siltstone. It consists of of predominantly shale but we’ll see some coarser sediments today. These sediments were originally deposited in lakes during the early stages of rifting of the supercontinent Pangaea about 220 Ma (million years ago), when the modern Atlantic Ocean was first opening. We encountered this rock at its type locale, and again in Goose Creek.

This is our first glance of Balls Bluff siltstone. Note how it holds up the ridge that borders Sugarland Run.

Exposure of cross-bedded siltstone about twenty feet above the creek level. Note that it is dipping to the west at less than 30 degrees west. The strike follows the regional trend of ~30 degrees northeast. This is the orientation (rotated over the last 200 my) of the rift that tore Pangaea apart.

This close-up of the previous image reveals a well-sorted sandy texture. There are no large pebbles or angular rock fragments visible and it was rough to the touch. Like sandpaper.

This is a nice view looking up-section. The total thickness represented in this side-creek/drainage channel is more than 50 feet, which is not available in exposures elsewhere in the area. The lower part (note the stream bed near camera) is blocky sandstone whereas shale predominates up the section (noted by slopes rather than ledges). The sandstone/siltstone beds become thinner up-section but are present.

We didn’t only encounter 200 Ma lake sediments along Sugarland run; this photo shows a fire hydrant and a road that has been eroded by recent erosion. Apparently, someone wanted to keep the hydrant because the bank has been stabilized with blocks of riprap. Note the rounded, angular boulders lining the creek bed. They were eroded from the nearby cliffs.

The stream bed is defined by periodic rapids (see above photo) and pools of quiet water, as seen in this image. The flood plain is a couple hundred yards across here and the creek is meandering in soft sediments. The underlying rock is not uniform, which leads to this alternating pattern.

This photo shows the thin-bedded coarser sediments (upper left) at the top of the section.

The thick-bedded sediment in the lower section can be seen in this image to consist of both cross-bedded layers (lower right) and massive beds (just above the center and left-lower). This variability could be due to thin layers of mud, which weathers easily distributed irregularly when the sediments were deposited; or the thick layers could be channel deposits, for example. This kind of variation at the outcrop-scale suggests a dynamic environment; a likely scenario is rapidly changing channels at a delta where a stream originally entered a lake. These channels can change during a single flow event (e.g. a heavy rain) or every few years.

Many of the processes recorded in these rocks are at work today. This image was taken from a gravel point bar, looking upstream. The gravel clasts were rounded and probably were transported many miles from upstream, in the recent past. They are not from the original sediment, but were broken loose from exposed rocks within the last few million years and transported to this location. Note the large, fallen tree forming a temporary dam in the upper part of the image.

Because of the obstruction seen in the previous photo, flow and sediment delivery downstream is reduced temporarily (until the tree rots and collapses). The creek bed is exposed, showing multiple channels that predate the obstruction. Creeks never follow a single channel like a canal. Sugarland Run is a braided creek at this locality. This morphology is dynamic because of high sediment load and can change rapidly, unlike a more stable anastomosing river. The Potomac is an example of the latter. Of course, a river can change character in different sections of its channel, which we’ll see in my next post.

Much of Sugarland Run’s banks are deeply eroded, reflecting rapid erosion because of regional uplift. The previous image shows this on the east (right) bank. However, erosion and deposition in braided streams occurs at many time and space scales. This photo shows erosion of the stream bed on very short time and spatial scales, probably in response to the reduced sediment flow caused by the fallen tree obstructing sediment more than water.

At the end of our walk we met Algonkian Parkway, where the Balls Bluff sandy sediments have been removed (or were absent to begin with), creating a wide flood plain near a point where another creek joins Sugarland run (not shown). This major boulevard follows a natural rise to the south, the dividing point between highlands and the ancient flood plain of the Potomac River.

Today’s walk followed a braided stream about a mile between ridges supported by sandy sediments that were deposited about 200 Ma in lakes, when the supercontinent Pangaea began to split apart. It is very likely that this stream is following an ancient fault zone associated with that event. The orientation of these rocks and their lack of folding supports the inference that this area was being stretched and the rocks, which were still buried many miles beneath the surface, fractured to accommodate the crustal extension. In fact, intrusive rocks cut through the Balls Bluff sediments elsewhere in the area.

The sedimentary processes occurring along Sugarland Run today are not that different from when these sediments were first deposited in lakes more than 200 my ago. There is one critical difference in their depositional regimes, however; this region is experiencing uplift today whereas this was a sinking basin when Pangaea was torn asunder. Consequently, the original sediments were fine-grained, eroding from distant mountains whereas the gravel and boulder seen today is the crumbling remains of those ancient sediments.

We’ll see what happens when Sugarland Run reaches the Potomac next time …

I wrote a post about streams traversing the Potomac River flood plain in a previous post. Horsepen Run is a meandering stream that has cut down several feet across the gently undulating sediments blanketing the mile-wide flood plain at this point on the Potomac (Fig. 1).

Figure 1. Map of Horsepen Run area. The reference to Fig. 2 is from the original post.

It had rained for 24 hours prior to this field trip, and then temperatures plummeted to well below freezing. It never snowed but there was some sleet. We are going to see some interesting features that resulted from this unique event. The watershed for Horsepen Run is rocky, with bed rock never more than a few feet beneath the surface, except on the floodplain (Fig. 1).

Figure 2. Shallow ditch along the path leading to the Potomac River, just entering the black-circled area in Fig. 1.

When I first came across the curious ice structures in Fig. 2, I thought someone had ridden a bicycle along the ice for fun. I didn’t figure it out until later.

Figure 3. Meander in Horsepen Run. The surface appears to be ice. Note the ice lying along the opposite shoreline. A curious structure.

Figure 4. The surface of Horsepen Run where it empties into the Potomac is solid ice. We tossed a branch out and it broke, indicating the ice is several inches thick.

Figure 5. Close-up of Horsepen Run at the Potomac. Note the broken ice about one inch thick lying along the opposite bank.

It’s time to put it all together. The water level during the steady rain was elevated so close to the stream’s outlet. The temperature was low enough (~10 F) to freeze the surface while water continued to flow beneath the ice. Even though the temperature remained very low, when the floodwater ran out from under the ice, it cracked and collapsed like a pane of glass. The newly exposed subsurface water at the lowered level then froze. This process occurred even in a few inches of water (Fig. 2).

The ice sheets (Figs. 4 and 5) serve as a high-water marker that melted away with the next thaw, serving as markers of how much water can collect during a light rain on nonporous soils and rock.

Geology integrates rock and soil with the atmosphere and hydrosphere into a holistic system that can surprise us at every change in the weather.

We had a chance this week to see what the Nature Conservancy does with our donations. They buy land and either maintain it or return it to the state or local government as public parks.

FIGURE 1. This post takes us back to the Potomac River, where we hiked around Fraser Preserve, a plot of land owned by the Nature Conservancy and open to the public. The photo above shows a stream flowing under an old concrete bridge as it cuts its way to the Potomac River. Downcutting here was similar to what we’ve seen elsewhere along the VA side.

FIGURE 2. The blue dot marks the gravel road that leads to the trail we took, which is indicated by the dashed line. The light-green area to the left is the schist (1000-511 Ma) we’ve seen along this stretch of the Potomac. The darker area to its right is a metagraywacke from the same era. These rocks were originally deposited in an ocean trench where ocean crust was being subducted beneath continental crust. They were buried along with the ocean crust and deformed into medium-grade metamorphic rocks.

We left the nature preserve and briefly entered Seneca Regional Park (north of the road marked DCWA in Fig. 2) to get access to the Potomac River.

FIGURE 3. A side channel of the Potomac River with a gravel bed and pristine water flowing over outcrops of schist. The elongate dark areas are lenses of schist on the river bed. Similar features were observed at River Bend park and discussed in a previous post. The banks here are gravel and would be a great place to cool off on a hot summer day.

FIGURE 4. This is a view of an abandoned channel of the Potomac River, taken from the top of a steep bank, probably 40 feet above the river. This area is about a half mile downstream of Fig. 3. The water seen through the foliage is part of a cut-off lake that is active only during high-water. Similar features have been seen further upstream but with some water flow year round, as discussed previously.

FIGURE 5. Close-up of a small exposure of metagraywacke along the access road at the blue dot in Fig. 2. Note the thin bedding and striations aligned perpendicular to the hillside. These are probably sole marks that indicate the flow direction in the original sediments. These rocks appear as lenses within the larger volume of schist, which was originally deep-sea mud. Imagine submarine flows flowing down the steep face of a submarine fan as turbidites.

This is a short post because we have seen most of these rocks and geomorphic features before. The novel feature that prompted me to write this was the wide floodplain (at least 300 yards across) and totally abandoned channel (Fig. 4). I also haven’t seen such clear water with no mud deposited at the shoreline. This location isn’t far from the narrow chasm that created Great Falls, where the river turns southward. I also noted a large number of steep gullies that appeared with no warning, indicating recent erosion from the surrounding hills, which are a couple hundred feet above river level. It seemed that some of the higher ridges were supported by cobblestones rather than bedrock, a feature we noted in Claude Moore Park that suggests ancient point bars.

Photographs can’t capture the complex topography of Fraser Preserve, especially with such colorful foliage interfering, so I encourage anyone who has the time to get out and see this beautiful landscape for themselves.

One final note: I support Nature Conservancy in their efforts to preserve natural lands and keep rampant development in check…

After so many posts from the volcanic island of Oahu, you wouldn’t think there was much left, but I couldn’t overlook the most famous volcano of all, although technically Diamond Head (Fig. 1) is a tuff cone like Koko Crater. This brief post is going to examine the internal structure of one of its limbs, on the seaward side.

The lighter color of the rocks in Fig. 2, compared to what we saw at Koko Crater or elsewhere on Oahu, suggests that the underlying magma chamber was depleted of mafic minerals. Dark hues associated with basalt are caused by minerals like plagioclase feldspar, amphibole and pyroxene, and biotite mica. The lighter color of the road cut (fresh and unweathered) suggests that the magma contained felsic minerals like albite and orthoclase feldspar, quartz, and muscovite mica. I could be completely wrong about this but there is no doubt that the rocks in Fig. 2 are not dark gray or black…

My hypothesis is consistent with what is known about the crystallization sequence of minerals from a melt and the resulting viscosity of igneous rocks. Mafic minerals and the lava they form have low viscosity and flow readily, as we’ve all seen in videos of eruptions on the island of Hawaii. These magmas bubble, flow, shoot fire into the air, and release pressure easily. However, felsic minerals (especially quartz) are sticky and have high viscosity, which causes them to resist flow, contain gasses, and eventually explode spectacularly (e.g. Mt. St. Helens).

I think the Diamond Head vent (i.e. volcano) tapped a part of the magma chamber that had already lost most of its mafic minerals, but it wasn’t as explosive as Mount St. Helens.

Another surprising feature I saw along the seaward margin of the Diamond Head tuff cone was a set of vertical joints filled with reddish rock (Fig. 4).

It is important to remember that the entire island of Oahu was constructed by magma escaping through innumerable fissures like those seen in Fig. 4, at first creating thick lava sequences deep beneath the Pacific Ocean’s surface, then flowing through breaks in the jumbled mass of previous flows. By the time the pile of basalt reached the water’s surface to form Oahu, the magma chamber was running out of gas (so to speak), and the lava was thicker and more viscous.

Diamond Head and Koko Craters were the result of these last gasps.

This post concludes my visit to Oahu, an island that rose from the sea less than five million years ago, formed by a huge magma chamber that was created when the Pacific plate slid over an upper mantle hot spot so concentrated that it melted ocean crust an constructed the Hawaiian archipelago, more than 1500 miles long.

I encourage anyone reading this post to explore the amazing story of this new land as it was populated by plants and animals, culminating in the incredible story of how Polynesian culture reached this remote land…

For this post, we went inside Koko Crater (Fig. 1) on the north side (Fig. 2), where the cone was breached, allowing easy access. A road had been constructed and the interior is now filled with a botanical garden and an equestrian center.

Parking is just outside the crater and a trail leads inside (Fig. 3), where a three-mile trail goes around the periphery. We didn’t have time to complete the circuit, so we settled for entering the main crater (see Fig. 2), where the walls were visible but not accessible for close examination (Fig. 4). However, the lower parts that were visible were covered with coarse debris less than 6 inches in diameter. There were some large boulders of vesicular basalt lying around, but they were loose and could have come from anywhere.

The extreme weathering seen on the inner slope in Fig. 4 suggests that the cap rock at least has a different composition, even if it is built from layers of ash. It is important to remember that tuff cones like Koko crater don’t continually erupt for centuries or millennia; they are local phenomena that vent part of the magma chamber that underlies a truly massive volcano like Ko’olau caldera (see Fig. 2). Thus, they are only active for a while, although dating is a problem for such short time scales.

It is important to note that Koko crater as we see it today has been eroded and the interior filled with breccia and ash during and between eruptions. We can’t say how much time passed between the layers seen in the middle of Fig. 5, but it could be hours to weeks. Most tuff cones are active for a couple of months, so the active period of Koko was on the order of a few years. Volcanic vents can produce a lot of ash very quickly.

This is my last post from Koko Crater. I didn’t have time to climb the 1048 steps to its summit, and I’m pretty sure my knees are glad.

In a nutshell, a vent formed along a fracture zone associated with the Ko’olau volcanic system and spewed ash and minimal lava flows onto the surface, where they interacted with the nearby shoreline, all of it lasting only a few decades at most.

Erosion has been minimal so we see Koko pretty much the way Pele left it….

These volcanic rocks were erupted more than 1.7 million years after the devastating collapse of the Ko’olau Caldera. It was probably a last gasp to release pressure within the magma chamber. Exact dating of these younger tuff cones is problematic, but they were all created within about a 50 ky window.

The kinds of volcanic tephra produced by the original Ko’olau volcano and the younger tuff cones shows a tendency towards more ash and less basalt. Certainly, Koko, Diamond Head, and this unnamed crater were part of monogenetic fields. These cones degassed a part of the magma chamber then became dormant; others appeared to perform the same function in another part of the chamber, part of the overall development of the Ko’olau volcano.

With the extinction of the fires here and at Koko Crater, it is safe to say that Ko’olau is dead and the goddess Pele has moved to her new home in the Kilauea volcano, on Hawaii…

This is a quick post to summarize some effects on the beaches of the windward side of Oahu, where basalt rocks of the Ko’olau volcano don’t protect the coast. Kailua isn’t far from Honolulu (Fig. 1) and the climate is similar. The wetter coastline, such as at Nu’uanu park, doesn’t extend this far. Kailua is a broad flat area, unlike the deep valleys of the north shore. The caldera is set back much further from the coast and the basalt is buried.

There is nothing surprising about what we saw at Kailua beach. Beach erosion is ubiquitous around the world; for example, it takes years for scarps like that seen in Fig. 3 to recover from a tropical storm in the Atlantic Basin. Recent studies suggest that in general, sandy beaches are being eroded; the proximate cause is a lack of sediment or increased wave energy, but the root cause most-often blamed is climate change and sea level rise.

We need to stop blaming the climate and reconsider all of the dams and diversions we’ve constructed on rivers that feed the world’s beaches, and ill-considered engineering projects completed in coastal zones.

The beach isn’t a play pool…