January 23, 2012 by Chuck Bailey
One of the courses I am teaching this term is Earth Structure & Dynamics (GEOL 323), a second-level geology course, and a required class for all Geology majors. This course combines structural geology, tectonics, geophysics, and a pinch of historical geology. Thirty-four students enrolled in this year’s class, that’s plenty, and free seats are hard to come by in the classroom.
As you might guess from the course title, I want students to be well versed at recognizing and understanding how earth structures form. What is an earth structure? Anticlines, reverse faults, half-grabens, stretching lineations, and boudinage (to name a few). This requires three-dimensional thinking, and 3D thinking requires much practice. On Friday, I projected an image of a craggy mountain side in the Needles Mountains of southwest Colorado and asked the class to 1) sketch the geology, 2) identify the salient geologic structures, and 3) work out the temporal history of the geologic events. I annotated the image such that different rocks and surface landforms are illustrated. Talus is rocky debris that has accumulated at the base of the cliffs: I want the students to “see through” the talus to visualize the underlying bedrock structures.
The limestone crops out above both the gneiss and quartzite, and it is the only one of the bunch that’s not been metamorphosed, therefore it is the youngest rock in the scene. But which of the metamorphic rocks is the oldest? The class struggled with this one. From the image alone the age relations are not discernible, but notice the quartzite is weakly metamorphosed (in contrast to the gneiss, a well-metamorphosed rock) and the original layering (bedding) is evident. Based on this information the quartzite is younger than the gneiss.
The boundaries between the different rock types are structures (referred to as geologic contacts). The limestone overlies the older metamorphic rocks along an unconformity, whereas the gneiss and the quartzite are juxtaposed across a fault (but just what type of fault?). Many students overlooked the significance of these geologic structures, but with practice, these abominations will stop!
The photo is two dimensional, but how best to help bring out the third dimension? Google Earth to the rescue—we zoomed from space into the Needle Mountains to about the same vantage point as the field photo was taken. With the tilt, pan, and zoom function it was easy to see the steep slopes of talus and that the limestone forms an erosional cap above the older metamorphic rocks. We will use overlays and tours in Google Earth to hone our visualization skills in the coming weeks. I want the Earth Structures students to translate from 2D to 3D without a hitch, by the end of the semester they’ll be dreaming in 3D!
So what is the temporal history of the geologic events? Here is my interpretation from oldest to youngest:
- formation of the protolith for the gneiss
- burial and metamorphism to form the gneiss
- uplift and erosion
- deposition of sand that forms the protolith to the quartzite
- burial and metamorphism affects the gneiss and creates the quartzite
- faulting juxtaposes the gneiss and quartzite
- deposition of the limestone
- minor tilting
- erosion to form the modern topography
- I take the picture
What the Geology 323 students worked out was the relative timing of geologic events, a fundamental skill for earth scientists. The next logical step would be to determine when, in absolute terms, these events occurred—that comes later in the semester when we delve into geochronology.
January 9, 2012 by Chuck Bailey
In my last post I discussed the Virginia earthquake that shook eastern North America on August 23rd, 2011. Here is the second part of that story: unfortunately the answer to the question I pose in the title is not particularly satisfactory. We cannot answer the question about whose fault is it (or more precisely which fault was it?) because the quake occurred along a buried and previously unrecognized fault in the Virginia Piedmont.
What caused the Quake?
Simply put, rocks in central Virginia are under stress, and on August 23rd, 2011 the stress level exceeded the local rock strength and a previously unknown fault ruptured. The maximum compressive stress that produced the earthquake was likely oriented west-northwest/east-southeast. What is the cause of that stress? Virginia lies astride eastern North America’s passive margin and is thousands of kilometers from active tectonic boundaries. The vast majority of the world’s earthquakes occur at or very near tectonic plate boundaries. The causal mechanism for the Virginia quake is not clear. As Seth Stein at Northwestern University puts it, parts of eastern North America may be better described as a “passive-aggressive” margin. We all know that passive-aggressive types are difficult to deal with and sometimes hard to understand—the Virginia Piedmont is similar. There is much left to learn about these intraplate earthquakes.
In the days after the August quake some postulated that fracking operations may have been responsible for the Virginia earthquake. Hydraulic fracturing, or fracking, is a process by which fluids are pumped into a well and pressurized until the surrounding bedrock, at a specified depth, is fractured. By increasing the fluid pressure, the effective stress on the subsurface rocks is reduced (in essence the pressurized fluid “pushes back” on the natural stresses generated from the overlying and surrounding rocks). When the effective stress exceeds the tensile strength of the rock, fracturing occurs. Hydraulic fracturing is being used to greatly increase the permeability of subsurface formations in the Appalachian basin to aid natural gas extraction. Many thousands of gas wells have been drilled in Pennsylvania, West Virginia, and Ohio over the past five years and the region has experienced nothing less than an energy boom.
Pumping fluids into the subsurface has triggered earthquakes (e.g. the Rocky Mountain Arsenal in the 1960s, the Rangely, Colorado oilfield between 1969-1973; the Basel, Switzerland geothermal field in 2006, to name a few), but these earthquakes are proximal to the injection sites (<10 km). Many localities where fracking and/or fluid injection into wells is occurring are rightly concerned about the prospect of seismic activity.
This is not the case in the central Virginia Piedmont as the nearest hydraulically fractured gas wells are nearly 200 kilometers away in West Virginia. Most hydraulic fracturing in the Appalachian basin occurs at depths of 1,000 to 2,000 meters (~4,000 to 8,000 feet); the hypocenter for the Virginia quake was considerably deeper, between 4 and 5 km depth (2.5 to 3 miles). The subsurface geology between the Appalachian basin (layered sedimentary formations) and the Virginia Piedmont (complexly deformed metamorphic rocks) is strikingly dissimilar as well. Let’s be clear, hydraulic fracturing was not responsible for the 2011 Virginia earthquake.
While the August temblor was larger than any previously recorded in the region, this quake was nothing unusual as a number of Virginia faults have slipped in the recent geologic past (remember, however, that recent for a geologist can be a million years or so). In my early years as a W&M faculty member, I worked on a project with my colleague Rick Berquist (Virginia Division of Geology and Mineral Resources) in which we examined reverse faults in the eastern Piedmont that juxtapose old igneous rocks over Neogene marine sediments. The earthquakes that caused these faults to slip occurred in the last few million years and the rupture occurred pretty much at the Earth’s surface. Although they seem novel to us, earthquakes are nothing new in Virginia.
Finding Fault at the North Anna Power Station
The shaking from the 2011 Virginia earthquake was vigorous and the nuclear reactors at the North Anna Power Station automatically shut down. Strong motion accelerometers at the North Anna site recorded peak accelerations of 2.6 m/sec2. The Earth’s gravity accelerates objects at about 9.8 m/sec2 (32 ft/sec2), so jolts from the Virginia quake accelerated the ground at just over 25% of Earth’s gravity (g). These values need to be put into perspective; most people think only of acceleration in terms of how fast an automobile ramps up its speed when the accelerator pedal is crushed. Consider a muscle car that rushes from 0 to 60 mph in 5 seconds- that’s an acceleration of 18 ft/sec2 (just 2/3s of a g) but more than sufficient to push you back in your seat. The ground acceleration coupled with the high frequency of the shaking from the Virginia quake was significant and conspired to do plenty of damage in Louisa County.
Part of the concern is that the shaking at the North Anna Power Station exceeded its original design value. I went back and examined the reports prepared by engineering firms for Vepco (the forerunner of Dominion, the power company that runs the North Anna facility) back in 1969 and 1971. In the section on Operating Basis Earthquake they write “On the basis of the seismic history of the area, it does not appear likely that the site would be subjected to ground motions above Intensity V during the life of the proposed facility.” As I noted in the previous post, the shaking experienced in central Virginia from the 2011 quake reached Intensity VII to VIII. Oops!
The report also states, “that the maximum horizontal ground acceleration… at the site due to such a shock would be no more than 12% of gravity”. The 2011 event exceeded that by a factor of two, but by all accounts the facility endured the shaking with relatively modest damage and the safety systems and procedures worked as designed. Although there were problems with the first seismic evaluation, we should be mindful that the report was written prior to the modern seismological network that has characterized the Central Virginia seismic zone over the past three decades.
In 1970, William & Mary Geology professors Steve Clement and Bruce Goodwin along with John Tyler Community College professor John Funkhouser visited the site and recognized a fault cutting through the bedrock in the excavations. At the North Anna site four approximately circular holes were excavated for the reactor foundations (although only two reactors were constructed), each hole was about 40 m (125 ft) in diameter and the total distance between the excavations from unit 1 to unit 4 was ~300 m (1000 ft). The fault, a east-northeast striking structure dipping 40˚ to 50˚ to the north, cut through all four excavations. Go figure.
The geological consultants hired to evaluate the site either never recognized the fault or chose not to disclose its presence. The original site assessment glossed over salient geological structures. Vepco’s Final Safety Analysis Report (from March 1973) stated “Faulting at the site is neither known nor is it suspected: all available information tends to confirm the continuity of the strata”.
Commenting on their 1970 visit (which occurred 3 years prior to the Final Safety Analysis Report) Steve Clement noted that “the fault was a classic example that I’d expect most undergraduate students to recognize” and Bruce Goodwin said of a Vepco representative “he’d have to have been asleep not to know there was a fault”.
A supplemental report was issued in August 1973 that focused on the geology of the now disclosed fault. The report opens with what I’d describe as a politic statement “During excavation for Units 3 and 4 at the North Anna Power Station, a minor shear zone of a type commonly found in the Piedmont was exposed. While technically a fault, since relative motion has occurred along this zone, it is not considered in any way to represent a fundamental geologic structure of the region.”
They really did not want to call it a fault. But a fault is what it is. Nobody wants a fault under a nuclear power plant.
The discussion now focused on whether the fault was a seismically capable structure. The Nuclear Regulatory Commission defines a capable fault as one that has experienced movement at or near the ground surface at least once within the past 35,000 years or movement of a recurring nature within the past 500,000 years or as a fault with instrumentally determined seismicity associated with it. Determining the capability of a fault is not necessarily easy. The supplemental Dames & Moore report is extensive and they employed an array of techniques to determine the age of faulting. Their conclusion, which was supported by a number of geologists with no vested interest in the power station, was that the fault was a minor normal fault that had reactivated an older reverse fault and that the last motion on the structure occurred ~200 million years ago. These findings were in lock-step with the 1970s-era understanding of eastern North America’s geologic history. The power station (Units 1 and 2) was built and Vepco did pay a modest fine for making false statements.
Nearly forty years on and one moderate earthquake later what can be concluded of the fault under the North Anna Power Station?
- This east-northeast striking fault is a minor structure in the architecture of the Piedmont
- The fault was not favorably oriented to have been reactivated by the stresses that produced the 2011 quake
- The gouge within the fault developed in the brittle upper crust and in the presence of abundant fluids
- The kinematics of youngest structures in the fault zone are consistent with reverse slip (not normal faulting)
- Faulting is younger than previously supposed
My conclusions are based on evaluating the existing reports, maps, and photos of the site. It would be glorious to study the exposures present in the early 1970s when the site was excavated and exposed. Today, there are more and better techniques for evaluating slip on faults as well as determining when those faults last slipped. Perhaps, the jolt from 2011 Virginia earthquake was sufficient for geologists to abandon some long-held assumptions about the passive character of eastern North America and reexamine the region anew.
December 22, 2011 by Chuck Bailey
As the year comes to a close it is a fine time to reflect on the 2011 Virginia earthquake. It’s been four months since the Virginia earthquake jolted eastern North America, and we now know more about what happened. This moderate-size (Mw=5.8) quake–felt by millions of people from Alabama to Quebec–caused significant damage in Louisa County, cracked both buildings and nerves in Washington D.C., and served notice that there is still some kick left in these ancient rocks.
What Happened on August 23rd?
At 1:51:04 p.m. (EDT) a fault ruptured at a point some 4 to 5 km (2.5 to 3 miles) below the Earth’s surface in Louisa County, Virginia (~60 km northwest of Richmond). As one side of the fault slid past the other, seismic waves radiated outward from the source area. The primary waves (P-waves) raced away at nearly 6 km/second: sweeping through Richmond 11 seconds after the quake, passing through Williamsburg in 20 seconds, and arriving at the West Coast in about 5 minutes. The primary waves were followed by shear waves and salvos of surface waves, these were the jolts that people felt. On the William & Mary campus; shaking perceptible to humans lasted about 20 seconds. At the North Anna Nuclear Power Station, 21 km from the epicenter, peak ground accelerations reached ~250 cm/sec2, more than sufficient to damage unreinforced masonry structures in the epicentral region.
The Virginia temblor was a moderate earthquake. Worldwide there have been 344 earthquakes of magnitude 5.8 or greater this year, which averages out to about one quake of this size (or larger) per day somewhere in the world. What makes this quake special is that it was the largest quake to rock the eastern United States in over a century and was felt by more people than any other quake in U.S. history. At the recent American Geophysical Union meeting, Shao and others report a seismic moment of 5.75 x 1017 Newton meters for the quake, which translates into ~35 terajoules of energy released (for comparison, World War II-era atomic bombs packed an energy punch of 50 to 90 terajoules).
The nature of seismic wave first arrivals at seismic stations helped define both the geometry and type of fault that slipped. The diagram to the right is a first motion diagram, in essence a stereographic projection that forms a visual representation of the fault style and defines two possible fault orientations for the Virginia earthquake. For the uninitiated these diagrams are confusing, geologists commonly refer to these diagrams as beachball diagrams.
Based on the pattern of first arrivals, the fault that slipped was a reverse fault striking to the north or northeast and dipping moderately either to the west or southeast. With these data alone the fault cannot be uniquely determined- it could be either plane. The P- and T- axes represent the axes of maximum contraction and extension respectively; in essence the Earth’s crust in central Virginia was shortened in an approximately east/west direction from the quake movement.
Within a day or so after the earthquake, seismologists from Virginia Tech and the U.S. Geological Survey had an array of portable seismometers installed in central Virginia. This equipment recorded hundreds of aftershocks. Most of these aftershocks were small (M= 1-3), but Louisa County residents certainly felt them.
The aftershock pattern clearly reveals the fault geometry: the earthquake occurred on a northeast—striking fault that dips about 50 to 55˚ to the southeast. Click on the link below to watch a short animation. The aftershocks are blue spheres, notice how they mostly line up neatly along a plane—that is the fault that slipped. The big red sphere (the August 23rd quake) plots off the plane, that quake was located by a regional network of seismometers and is not as accurately located as the aftershocks pinpointed by the locally deployed array of seismometers.
During the quake the southeastern side of the fault (hanging wall) was shoved upward with a maximum displacement of about 1 meter. The total rupture length along the fault was likely 5 to 10 kilometers. There was no rupture at the surface because displacement across the fault did not propagate all the way to the Earth’s surface. The 2011 earthquake occurred along a blind, and previously unrecognized, reverse fault in the Virginia Piedmont.
Geology of the Piedmont
The earthquake occurred in the Piedmont, a region of complex geology that is the metamorphic core of the Appalachian Mountain system. Some of these rocks originated far from North America and were later crushed against the continental margin during tectonic collision and faulted to their current location. In the past twenty years geologists have distinguished many different terranes in the Piedmont: terranes are blocks of crust with distinct geologic histories and are bound by major faults or tectonic sutures. The difference between terranes is well illustrated on the aeromagnetic map displayed in the animated map sequence below.
The 2011 Virginia earthquake occurred in the Chopawamsic terrane. Rocks in this terrane formed as volcanic and plutonic rocks in a continental arc during the Ordovician Period (~470 to 450 million years ago). This arc was likely outboard of ancient North America and was later accreted to the continent. In the late Paleozoic (300 to 280 million years ago), during the massive tectonic collision that created Pangaea, these rocks were squeezed and baked (deformed and metamorphosed) into gneisses and schists. The Chopawamsic terrane is bound on the northwest by the Brookneal/Shores fault zone and on the southeast by the Spotsylvania fault zone. Our kinematic studies of these fault zones indicate that they experienced simultaneous right-lateral wrenching and shortening when they were active in the Paleozoic. In essence, the Spotsylvania fault zone moved the Goochland terrane to the southwest and the Brookneal/Shores fault zone moved the Chopawamsic terrane to the southwest relative to the western Piedmont.
In the Triassic Period (220 to 195 million years ago) Piedmont terranes were fractured and broken during rifting which created sedimentary basins, such as the Culpeper and Richmond basins. This rifting ultimately opened the Atlantic Ocean. Traditionally, geologists have viewed the Piedmont as a relatively static region whose tectonic heyday was long past. Today, it is a gently rolling landscape mantled by thick soils, the product of slow erosion for millions of years and a seeming dearth of tectonic activity.
But as my colleague David Spears at the Virginia Division of Geology and Mineral Resources has pointed out, there are subtle clues in the rock structure of the central Piedmont that suggest recent tectonic activity. The 2011 quake was a not so subtle reminder that David is correct and we need to get our boots on the ground and eyes on the outcrop to study the region in more detail.
The Central Virginia Seismic Zone
The Central Virginia seismic zone is a region of moderate but persistent seismic activity. The first recorded quake occurred in 1774 near Petersburg and was felt throughout Virginia and North Carolina. The largest historical quake (prior to the 2011 temblor) in the central Virginia region took place in 1875 and is estimated to have been a magnitude 5.0. Estimating both the size and exact location of historic earthquakes is difficult. Geologists use the Modified Mercalli Intensity Scale to estimate the size of historical earthquakes based on eyewitness accounts and damage reports. This is a 12-point scale that employs roman numerals, with a II being a quake so small that only few people felt it, a IV being felt by many people indoors, a VI being felt by all with some damage to plaster and masonry, a VIII causes considerable damage to structures, a X destroys most structures and the ground is thoroughly cracked, and a XII equals total damage. The intensity of damage decreases away from the epicenter. The 1875 quake reached an intensity of VI to VII in central Virginia; the 2011 quake had a maximum intensity of VIII in Louisa County whereas in Williamsburg the quake’s intensity was a IV.
By the late 1970s a regional array of permanent seismic monitoring stations helped better locate and measure earthquakes in the southeastern United States. Over the past three decades there have been 47 quakes with a M≥2 in central Virginia (22 of those are aftershocks from the 2011 quake). These quakes are widely distributed and rarely correlate to mapped faults (see the animated map above). The focal mechanisms are consistent with slip on reverse faults at depths between 4 to 10 kilometers. At these depths, the rock is warm (70˚ to 200˚ C or 160˚ to 400˚ F), but solid and behaves in a brittle fashion when placed under stress.
There is more to tell, but my research students counseled me to curb my enthusiasm, as blog posts should not be too long. In the next post I’ll discuss the possible causes of the 2011 earthquake and tell the lurid history of finding fault at the North Anna Nuclear Power station.
December 14, 2011 by Chuck Bailey
The Alberene Dream Team spent the summer of 2011 in the field working to understand the geology of the eastern Blue Ridge. As summer turned into the fall semester, the team compiled their data and started to analyze the buckets of rocks we’d collected during the field campaign. In the Geology Department we cut rocks into small chips that were then glued to a glass slide, sawed once again, and finally ground down until the thickness of the rock is ~30 micrometers. For obvious reasons these microscope slides are called thin sections. Light typically passes through the thin sliver of rock and under the microscope the mineralogy and texture of the sample are revealed in a beautiful and colorful fashion. What follows is a report from each member of the Dream Team on the microscopic character of some eastern Blue Ridge rocks and what can be gleaned from thin section observations.
Alex Johnson- Yes, it is very Gneiss, just don’t take it for Granite
It was a strange twist of fate that I, the youngest member of the dream team, chose a thin section of the oldest rock unit in the study region—the Blue Ridge basement complex. The rock originally crystallized as a granite some one billion years ago. Large feldspar phenocrysts bear witness to the rock’s igneous heritage. But, a billion years is a long time and geologic hijinks happens—this rock is best described as a biotite-bearing granitoid gneiss. Biotite, muscovite, epidote, and sphene form a posse of metamorphic minerals that overprint the original texture. Quartz, an original igneous mineral, has recrystallized into strain-free aggregates with tidy and straight grain boundaries. Alas, the poor feldspars. To call them disturbed would be too kind, destroyed too cruel, but somewhere in between it lies. The feldspars have been saussuritized, hydrothermally altered to a fine-grained mixture of epidote, sericite, and zoisite. That’s what a billion years and a couple of orogenic events gets you.
Andrea Jensen- Soaking up the soapstone
Ultramafic rocks are characterized by very low silica content and are dominated by dark colored minerals with plentiful iron and magnesium. They are igneous rocks associated with magmas derived from the mantle deep within the Earth. The Alberene quadrangle is bursting with ultramafic rocks that were originally pyroxenites dominated by the mineral pyroxene with lesser amounts of olivine and plagioclase feldspar. These pyroxenites have since undergone low grade metamorphism and been transformed into carbonate-bearing soapstones. The soapstone in the Alberene area is so extensive that it has long been quarried—native peoples used this soft stone for heat resistant bowls and pipes; modern uses include chemically resistant counter tops, architectural flagstone, and as a medium for sculpture.
Controversy surrounds the origin of the ultramafic rocks in the eastern Blue Ridge. Some workers suggest these are differentiated mantle-derived magmas that intruded thinned continental crust during rifting, whereas other geologists posit that these rocks are altered bits of the mantle thrust into place during tectonic collision. Regardless of how they formed, these rocks have since enjoyed metamorphism. Soft minerals such as chlorite, talc and carbonate replaced original minerals.
Molly Hahn- A dark and dirty rock
The graphitic schist illustrated below forms layers within metasedimentary rocks of the Lynchburg Group. In hand sample it is a fine-grained black rock with well-developed foliation and rusty patches. In thin section, the majority of the rock is dark and opaque with bands of gray, elongated quartz grains and reddish brown biotite grains. The banding and elongation of minerals is called foliation—the result of mineral recrystallization and realignment during deformation. The dark materials are a mixture of graphite, a soft fine-grained platy mineral composed entirely of carbon, and iron oxide minerals. Just what is the source of carbon in this rock? One possible source is organic matter, perhaps the remains of microscopic organisms, that was deposited as seafloor mud in a dank anoxic environment. The Lynchburg Group is a thick sequence of mostly marine rocks deposited during the Neoproterozoic (~700 million years ago) long before complex shelled organisms had evolved. As the rock was buried and later metamorphosed the volatile elements in the organic matter were expelled and the carbon crystallized into graphite. The fine-grained iron oxide minerals likely came from oxidation of pyrite (iron-sulfide) during metamorphism and/or later weathering near the Earth’s surface.
Kevin Quinlan- Mistaken identity
The Scottsville Basin, located a few kilometers south of the Alberene quadrangle, is a body of sedimentary rocks that formed during the breakup of Pangaea some 200 million years ago. My research focuses on describing how the rocks in the basin have deformed over time. This sample came from a low outcrop in the James River and contains abundant quartz and plagioclase feldspar. These mineral grains are rounded, evidence that they were deposited as detrital sediments. Individual grains are held together by a fine-grained matrix, giving the rock a clastic texture. This texture is a good indication that the sample began as a sedimentary rock, probably sandstone. Notice, however, that the matrix between grains is dominated by brightly colored, elongate crystals. This is the mineral muscovite, formed when clay minerals in mud are heated to at least 350 °C at pressures that occur 10 to 20 kilometers below the Earth’s surface. Under these conditions the rock experiences metamorphism. Hence, the rock is best described as metamorphosed sandstone. Metamorphism in the Piedmont happened during the late Paleozoic (~300 million years ago). The Scottsville Basin formed after Paleozoic metamorphism. However, existing geologic maps place this meta-sandstone within the basin. Because we know that this rock is too old to be part of the basin, it’s reasonable to assume that the boundary of the basin has been improperly drawn on old maps. This finding will help me in my research goal of more accurately delineating the basin’s borders.
Chuck Bailey- The Monticello mystery
One of our research targets is an intriguing rock reported to crop out in the vicinity of Thomas Jefferson’s Monticello. A number of older geologic studies describe an alaskite, a felsic igneous rock composed primarily of K-feldspar (K for the element potassium) and quartz, intruding the greenstones of the Catoctin Formation. Alaskite, as a rock name, is anachronistic and modern petrologists refer to these rocks as alkali-granites. We located a string of linear outcrops of a fine-grained feldspar rich rock to the south of Monticello. The microscopic view reveals the rock is loaded with K-feldspar with lesser amounts of quartz, opaque minerals, and plagioclase. But it is the texture of the rock that tells the important story—this rock has a clastic texture, composed of rounded to angular fragments of sand grains in contact with one another. Clastic textures are diagnostic of sedimentary rocks. The Monticello ‘alaskite’ is a sedimentary rock, not an igneous rock as previously reported, that forms a layer within the sequence of Catoctin Formation lava flows. The source of this clast-supported arkosic arenite, the proper name for this rock, is a mystery as the K-feldspar could not be derived from the mafic Catoctin Formation. The older granitic basement complex, exposed to the west of the Catoctin Formation, does contain some K-feldspar, but what processes would concentrate the K-feldspar to such a degree? That’s the nature of research—as one question gets answered, others arise.
November 18, 2011 by Chuck Bailey
Fall has reached its full crescendo in Williamsburg; leaves are a riot of orange, scarlet and russet, the temperature has dropped, and frost has been sighted on more than a few pumpkins. The National Weather Service has officially declared the 2011 growing season over. The growing season ends when the first freeze occurs (i.e. the air temperature drops to or below 0˚ C).
The Weather, Climate, and Change class has taken a close look at the first freeze dates for Williamsburg. The histogram below is a time series, with 111 years of records from 1900 to 2011, that illustrates the first freeze date in Williamsburg for a given year. The earliest date for a frost was October 2nd (which occurred in both 1946 and 1947) and the latest frost occurred on December 1st (in 2009), the mean date for the first frost is just about November 1st. Are any trends evident from the time series?
Over the last 111 years Williamsburg’s first freeze occurred with equal frequency in October and November, but in the recent past the first freeze almost always occurs in November. Only once in the past 14 years did the first frost occur earlier than November 1st. What, if anything, might this mean?
Growing season, the time between the last frost in the spring and the first frost in the fall, is an important metric of a locale’s climate. In Williamsburg the mean growing season is an ample 203 days, but has varied between 156 days (in 1946) and 263 days (in 2010). The plot below is the revered (or is that reviled?) scatter plot; each orange diamond illustrates the growing season for a particular year. Do any patterns or trends jump from the graph?
The variation from year to year is significant and makes discerning any temporal patterns difficult. When in doubt it’s always fun to fit a line to a data set. With a few clicks of the mouse I ran a linear regression and produced a best-fit line for the data. The best-fit line has a gentle positive slope (0.146), indicating that the growing season in Williamsburg has increased over time. According to the regression, the growing season is increasing by ~1.5 days per decade (just over two weeks in a century). But the correlation coefficient (R2) for the best-fit line is quite low (0.06), an expected result given the wide variation in growing season length from one year to the next. Many people may understandably be skeptical of this trend.
Another way to consider these data is to take a five-year running average (incorporating the 2 previous years and 2 following years as well). This produces a smoothing effect that reduces the overall variation in growing season length. My eye is drawn to a dip in growing season in the mid-1940s and the perceptible lengthening of the growing season starting in the 1990s.
For many people the decade-scale is a handy interval in which to categorize time (e.g. the Roaring 20s, the Swinging 60s, etc.). Consider Williamsburg’s growing season averaged by decade. The 1940s take the prize for shortest mean growing season (185 days) while the 2000s come in with the longest (225 days). The thin black lines and bars (whiskers) extending above and below individual data points are a standard deviation: the longer the whisker the greater the variation.
Here’s the question I put to the students: is there a significant difference between decadal growing seasons? The standard deviations between the 1940s and 2000s don’t overlap; we can reasonably conclude that the growing season is significantly different between these decades. There is much overlap between the standard deviations in the other decades. Can we make sense of these data? Are the trends real or significant? My Weather, Climate, and Change students debated all of this.
It was a contentious debate because time series data can be complex and teasing out patterns is not trivial. But consider this: since the start of the 21st century the shortest growing season in Williamsburg has been 207 days, during the 30 year interval between 1960 and 1989 only six years (20%) had growing seasons as long as 207 days. Five of the Top Ten longest growing seasons occurred in the last decade. Since the 1980s, Williamsburg’s growing season is about 3 weeks longer than it was in the 1980s, the last frost in spring generally comes earlier and the first frost in fall arrives later. We could have an early frost next year and a short growing season, but the trend towards longer growing seasons is clear.
What the class hasn’t debated yet is just what could be causing these longer growing seasons. Is it global warming, an urban heat island effect, or maybe a decadal-scale climate oscillation? But let’s not get ahead of ourselves! Our discussion on the driving processes comes next…
November 8, 2011 by Chuck Bailey
William & Mary’s Geology Department turned 50 years old in 2011. We celebrated the half a hundred mark with a weekend wingding on campus and in the field. Nearly one hundred alums were in attendance and by my reckoning a good time was had by all.
Founded in 1961, the Geology department has graduated nearly 800 majors. The feté brought together many of these alums from across the decades, current students, departmental friends, emeriti faculty, and present-day faculty. The crowd arrived in the department on Friday night for a voluble reception.
On Saturday morning we rolled westward to the Falls of the James in Richmond for a field trip dedicated to former Professor Bruce Goodwin, who introduced legions of W&M students to this exceptional exposure of granite and its erosive fluvial features. We remembered Bruce and then cast our eyes towards the rock and riverbed to understand the geologic history of the Fall Zone. The banter on the outcrop was rich.
Back on campus that evening, participants exchanged their fleeces and boots for more upscale fashion at the banquet. As department chair the master of ceremony duties fell to me. I recounted a departmental history (although, I must admit, strict historical accuracy was tossed under the bus). Other speakers recollected past adventures and made a strong case for why the department has been a vibrant community for 50 years.
For me the weekend will be among the most memorable and satisfying I’ve had in my fifteen years at the College. To see so many students, former faculty, and friends was special. We won’t wait another 50 years for another celebration!
October 14, 2011 by Chuck Bailey
This morning I handed back the graded mid-term exam to the Geology 312- Weather, Climate, and Change class. The average (or mean) score was 78% with the high grade topping out at 92%. The grade distribution is skewed to the left (that is there is a long tail of lower grades) with a noticeable absence of high-end grades- this is a typical pattern in the larger classes that I teach. My exams are difficult and given the time constraint set by an hour-long exam, scores in the high 90’s are quite rare. Students earning scores of 85% did great. I don’t assign letter grades per se, but from the illustration below you get a sense of my thinking on that matter.
Students are creatures of habit and typically sit in the same seat throughout the semester, thus I was able to determine the average grade per row. The classroom (Tyler Hall 102) has six rows- between 9 and 12 people sit in each row. There are some discernible trends in exam scores between the rows. Row 1 (the front row) had the highest average and the row average dropped back to Row 3, there is a curious ‘grade inversion’ between Row 3 and 4, and then the grades fall off towards the back row. Now these are just averages for the rows, not everybody in the back row did poorly. A measure of grade variation per row is given by the standard deviation (shown as the gray horizontal lines), note the wider spread in grades in the back two rows. So just what is the best row to sit in? The use of averages and standard deviations will be important as the class works on their regional climate projects, discerning the variability of temperature and precipitation over time.
Speaking of averages, the cumulative precipitation for 2011, measured at William & Mary’s Keck Environmental Field Lab, is spot on for an average year. Typically, by the second week of October, Williamsburg has received just over 1,000 mm (1 meter or 39 inches) of precipitation- this is based on time series data collected between 1949 and 2010.
An average year might not seem so interesting, but how the average was reached is more informative. The plot above illustrates the weekly precipitation (red columns), the cumulative precipitation for 2011 (the thick blue line), and the cumulative yearly average (dashed purple line), which steadily climbs throughout the year. Precipitation events, such as the passage of weather fronts and thunderstorms, do not occur at an even or measured pace. Some weeks Williamsburg receives no precipitation, whereas other weeks we get drenched. By mid-August (week 33) campus had received 600 mm of precipitation for the year and was over 25% below average. Hurricane Irene (in late August) changed all that by adding over 200 mm in just under a day. As summer turned to fall, a consistent flow of southerly air has repeatedly brought rain to Williamsburg and in the process topped up W&M’s rain gauges to their long-term average. La Niña conditions are set in the equatorial Pacific Ocean, this typically brings warmer and potentially wetter conditions to the mid-Atlantic. There are still some 10 weeks left in 2011 and it remains to be seen whether precipitation for the entire year will all average out.
September 27, 2011 by Chuck Bailey
As I’ve noted in these posts before, Geology Departmental field trips are unique as they bring together the W&M geologic community in a way that staying on campus never could. The Fall Field trip took an enthusiastic crew of students and faculty to the Blue Ridge Mountains for a weekend getaway. Our timing was just right as an on-shore flow of moist air brought rain and a dreary Saturday to Williamsburg. The mountain mayhem began at Big Meadows in Shenandoah National Park. We savored the irony that Big Meadows, one of Virginia’s wettest locales with a yearly precipitation average of 130 cm (50”) per year, was dry while Williamsburg was wet.
Professors Greg Hancock and Jim Kaste got the discussion started as we pondered the flattish landscape of Big Meadows, and hiked into Hogcamp Branch to consider stream dynamics and the role that bedrock plays in water chemistry. The ascent of Bearfence Mountain took us from the basement complex (always my favorite), up through outcrops of sandstone in the Swift Run Formation, to a rocky spine of greenstone exposed along the crest of the ridge. At Rockytop Overlook we basked in late afternoon sunbeams. The scene was so sublime that some faculty broke into song, songs extolling the virtues of steep slopes and rocky tops. We made camp in the twilight, devoured bowls of chili, and reveled by the fire long into the evening.
To get a better sense of the mayhem- Check out video snippets from the field trip
On Sunday morning we shook off the dew and began with a jaunt along the Appalachian Trail to exposures of sandstones, conglomerates, and siltstones in the Weverton Formation—its depositional environment, way back in the early Cambrian period (~540 million years ago), was vigorously debated. Another hike took us to Calvary Rocks, where well-cemented quartz sandstones of the Antietam Formation revealed their secrets. Our last debate focused on the landscape—are the Blue Ridge Mountains growing, shrinking, or in some long-term steady-state? Any thoughts? (comments welcomed)
Our weekend excursion to the Blue Ridge was uplifting (pun intended!). For an old mountain range the Blue Ridge is more dynamic than you might think.
September 7, 2011 by Chuck Bailey
Hurricane Irene raked North America’s East Coast and put a kibosh on the start of William & Mary’s Fall semester. Students were sent packing to safer locales while most faculty and staff hunkered down in Williamsburg. Irene delayed opening convocation by a week, but last Friday the choir belted out a spirited version of William & Mary’s school song- James Southall Wilson’s “Our Alma Mater”- and the semester was truly under way.
For my money, the best line in the song is “Hark upon the gale!” There are varied interpretations as to what “hark upon the gale” really means- I view hark as an active rather than a passive verb. During Irene the Geology faculty did more than just listen to the gale- they were out in the gale collecting data before, during, and after the event.
On Friday afternoon the W&M campus was quiet, but the atmospheric pressure had started its telltale decline. By the wee hours of Saturday morning, August 27th the wind was blowing steadily from the east and the rain began. Professor Greg Hancock measured the cumulative precipitation at his house- in the early afternoon the rainfall rate exceeded 1” per hour, deluge-like rates! By noon the wind was coming from the northeast, and as the Hurricane’s center moved to the northeast off Virginia’s coast, the wind steadily shifted from the north to the northwest and then to the west. Sustained winds at Yorktown ranged from 15 to 23 m/sec (~30 to 50 mph) with the peak gusts, at nearly 30 m/sec (66 mph), occurring near midnight. These winds blew down trees and shut off the power to most Williamsburgers.
At his house near York River State Park Professor Jim Kaste was monitoring a small perennial stream and collecting water samples for chemical analysis. The normal base flow for the stream is quite small (<0.01 cubic feet/sec), during the height of Irene stream flow exceeded 6.5 cubic feet/sec- more than two orders of magnitude (>100x) greater than normal flow! Even a week after the storm, the stream has yet to fall to its base flow condition. His environmental geochemistry class is now running the water samples to measure the dissolved load carried by the stream and assess what fraction is derived from Irene versus groundwater.
On Sunday morning the damage from downed trees was widespread. I lost two trees to the storm, but none fell on the house- my next-door neighbor was not so fortunate. In between chain sawing and cleaning up I measured the orientation of the blown over trees in my neighborhood and then later on campus and around Williamsburg. The orientation of the downed trees tells a story about the kinematics of the wind during the storm. There is a clear bimodal pattern- the majority of the trees fell towards the southeast and southwest. Almost no trees fell toward the west, northwest, north, and northeast.
Prior to Irene’s arrival I’d hypothesized that most trees would fall towards the west and southwest as winds from the east and northeast (on the leading edge of the storm) would do the damage. As rose diagram illustrates my hypothesis was generally incorrect.
Incorrect hypotheses warrant further inquire so I decided to investigate the dynamics of the wind, leading me to calculate the cumulative wind energy for winds from various directions during the storm. Utilizing the wind data from the Yorktown Coast Guard station, I sorted the wind data into directional bins (in 10˚ increments) then for each observation determined the wind power acting over 1 square meter via this relationship:
Power (in Watts) = 1/2 x (air density) x (area) x (wind velocity^3).
I used a standard density for air of 1.225 kg/m3 and an area of 1 m2. Here is an example calculation-
1/2 x (1.22 kg/m^3) x (1 m^2) x (10 m/sec)^3 = 612 kg m^2 /sec^3 = 612 Watts
Wind measurements were made every 6 minutes, so to convert to energy I multiplied the power by 360 seconds (the number of seconds in 6 minutes). Power is energy per unit time, so to get energy from power- multiply by time. I then summed each 6-minute interval for a given bin to yield the cumulative wind energy from each direction.
Well over 60% of the cumulative wind energy came from the northeast (that is wind blowing towards the southwest) while the cumulative wind energy from the northwest is markedly lower. This is likely because the wind blew consistently from the northeast for many hours (see the wind direction plot above). The correlation between the orientation of downed trees and directional wind energy is less than stunning. Perhaps there is a better metric for understanding the downed trees.
The next step was to calculate the power (in kWatts per square meter) from the peak gusts from a particular direction. This yielded a bimodal pattern with peaks in both the southwest and southeast quadrants- much closer to the observed tree fall data. It seems reasonable that the peak winds toppled trees.
Another significant observation is that most trees fell during the last half of the storm. Although the energy and even peak gusts coming from the northeast (first half of the storm) were high, more trees fell due to winds from the northwest (last half of the storm). It had been dry for some weeks in Williamsburg and it took time for the rainfall to saturate the soil, and saturated soils are invariably less cohesive and weaker than dry soils.
Hurricane Irene was an exciting event for the William & Mary earth science community. Collectively, the storm damage to campus and the community was modest- that is a good thing. Although many were disappointed with the media’s hype of the storm, Irene tracked out just as meteorologists had predicted. In Williamsburg Irene was an asymmetric storm with copious rain in its long opening act and a tumultuous windy finish. The storm set the tone for the early days in my Weather, Climate, and Change class. As the Fall semester unfurls, Geology faculty and students will continue to work out the intricacies of Irene’s environmental impact .
So as the song says… “Hark upon the gale! Hear the thunder of our chorus”
August 23, 2011 by Chuck Bailey
This afternoon, our Geology faculty meeting was adjourned by a motion from the floor. A 20-second motion from the floor, but more to the point, a 20-second motion from the Earth. Virginia and the East Coast experienced a moderate, but widely felt earthquake at 1:51 p.m. (local time). It was quite a jolt.
The earthquake’s epicenter was about 60 km (~40 miles) northwest of Richmond, Virginia and occurred in the central Virginia seismic zone- an area of modest (or so we thought), but persistent seismic activity in the Piedmont. This region is laced with ancient faults that formed 200 to 300 million years ago when Virginia was at the frontline in an ugly collision between tectonic plates. I study these fault zones. Today’s temblor makes it clear that these faults are 1) not inactive and 2) have the potential to produce significant and damaging earthquakes. We have much to learn about the stresses that cause faults to slip this far from modern tectonic plate boundaries (in this case at the Mid-Atlantic Ridge some 3,000 km from central Virginia) and the hazards that these old, but restless, faults pose. It’s why we do research at William & Mary.
An earthquake presaging a hurricane- this could be quite a semester to study the Earth!