When gas prices crept up to $4/gallon in the United States (it’s happened before, it will happen again) we started seeing a lot of questions about the oil supply, driven mainly by concern about what is driving the prices up. There is not room enough to adequately describe all the regional and local processes that control the price of gasoline, but there are some relatively straight-forward principles involved. We will discuss several in this and following chapters.

            Q: Is the world running out of oil?

            – Mark M.

            A: The answer is yes and no. I thought you’d like that.

            Yes, the world is running out of “sweet” crude – the easy-to-get stuff, the “low-hanging fruit” of the hydrocarbon world.

            No, there is plenty of oil still left out there – surprisingly huge amounts, in fact – but it’s the hard-to-get stuff.  It is more difficult to find (and therefore more expensive) and a lot more dangerous to extract, and therefore even more expensive still… but there are a LOT of unexploited hydrocarbons still out there.  The April 9, 2012 issue of Time Magazine (http://content.time.com/time/magazine/0,9263,7601120409,00.html ) gives a good summary of why there is a lot of new oil coming online – yet gas prices will still inevitably go up. However, even this particular article is not complete, largely missing the more recent Fracking Revolution, and unfortunately you also need a subscription to even read it.

            Sigh. These journalists – always feeling like they need to feed their families!

            In 1956, a US Geological Survey geologist named M. King Hubbert published a paper describing the life history of an oil play (Hubbert, 1962). From discovery to full exploitation, the production of a given oil field quickly ramps up, reaches a peak… and then declines in a roughly symmetric curve as you pump most of the good stuff out. If you put together a production curve for all the major fields in the United States, it makes sense that they would all, in aggregate, behave in a similar fashion: US production would ramp up, reach a peak, and then peter out. Hubbert was the first to reason this out, and his prediction in the late 1950’s is called the Hubbert Curve.

            This curve has proven remarkably accurate – at least insofar as its build up and decline shape is concerned. It has been consistently incorrect, however, as US production seems to always be greater than the Hubbert prediction. More recently, the fracking revolution has totally upended the Hubbert curve. Hubbert also had some fierce detractors, including the chief geologist (and later USGS Director, Vincent McKelvey), who all disagreed with his reasoning.

            As we watch prices rising it might be tempting to either get depressed – or buy stocks in alternative energy sources. However, there were several things that this famous curve didn’t factor in, and those things include supply-and-demand laws (basic economics), geopolitics (the narrow Strait of Hormuz, for instance), and… the incredible ingenuity of the human mind. It doesn’t take a lot of smart guys, either, if there is money driving the train. Yes, Capitalism. If you have 7 billion people on a planet, it just takes a few with the vision and the determination to find a way. In most cases like this (a notable exception being everyone other than Wall Street during the Great Recession), everyone benefits.

            For this chapter, I will only discuss enhancing production – stretching that Hubbert curve out. I worked as a young physicist for Getty Oil Company in Bakersfield, California. It was a huge improvement over fighting forest fires and washing dishes to pay for college. Getty (formerly Tidewater Oil Company) had been sitting on vast stretches of the Kern Oil Field since it had initially been exploited in the First World War era. By the time I went to work for them, the area called Oildale was an ecological ruin: the ground everywhere was reddish brown, and there was sparse, always sick vegetation. Early wells in the ’20’s and ’30’s would “blow out” as the drill reached the producing horizon (which was quite shallow, ranging from as little as just 500-900 feet, or 150-275 meters in depth), and oil would blast out all over everything until the pressure dropped. The field was originally found by following natural seeps decades before geophysical technology was developed. The Roustabouts – the guys who built and then worked the wooden derricks at the time – would rush to build a dam to save at least some of the Black Gold that came gushing out. The Kern field by the late 1960’s was well past the maximum production point. The overlying ground pressure had long since squeezed out the low-viscosity, easy stuff, and no longer squeezed anything. Pumps were producing ever-dwindling results, sometimes down to a barrel or two a day. Since the pumps ran on diesel fuel, it quickly became uneconomic to keep these running, and large parts of the field had already been shut down by the time I arrived.

            But several brilliant geologists did some core-drilling, initially to see if there were places where they could put more or better pumps. They were astounded to learn that the “nearly dead” Kern Field had produced only about 15% of its entire contents since 1915!  The other 85%, the black stuff that remained in the ground and showed up in those cores as sandy tar was too viscous to work its way out. In other words, it was too “gooey” for the straining pumps to move efficiently. Today we call this Heavy Crude. Some equally brilliant engineers figured out that if you could heat the stuff, the viscosity would drop, and it would flow out much more easily.

            There are two problems with this:

A. You have to somehow heat it up, and then…

B. You somehow have to keep it warm or it will turn solid and clog your pipes.

            There are essentially three ways to move the heavy crude sludge to the surface: water flood, steam flood and fire flood. Then move it down pipes afterwards? That’s a whole different ball-game, but it is solvable.

            1. Water flood is a push-pull operation. You start by assuming there was SOME viscosity left in the oil in the producing horizon. You simply do just that: you flood a chosen well with water, and push the oil out to other wells that would be pumping furiously to help pull the oil in. This has been used very successfully in a failing oil field near Taft, California. To do this, of course, you would have to set cement and screens on your well to allow fluid movement only at the producing horizon, and case (or block out) the other horizons – or you will lose all your water. You also need a substantial supply of water, not something easily found in an arid agricultural region like the southern San Joaquin Valley.

            2. With steam flood, something was added to the push-pull: heat designed to lower the viscosity of the gooey stuff. The engineers built huge steam generators in several locations around the Kern Field. These things were monstrous: they looked like huge silver cans lying on their sides – cans the size of a diesel locomotive. The engineers would heat water to 500 degrees F, and pump it at 500 psi down the production pipes for 5 days – the so-called 5/5/5 (and why this part of the story is not metric). They would then let it “stew and cook” for two days, then would un-cap those same production pipes and just let the steam blow off for several days. Then they would start the pumps going again. For a surprisingly long time afterwards, the once-abandoned production wells would easily produce 50 – 300 barrels (8 to 50 cubic meters)/day. There are actually two versions of steam flood: a steady push like fire flood (below) and an episodic version that I’ve just described. This episodic version can work with a single well instead of multiple wells, and is called “Huff and Puff.”

            3. With fire flood, the ante (and the risk) was ratcheted up another notch. More wells are needed. After carefully imaging the producing horizon with seismic methods, engineers would drill a circle of wells around a central well. In the central well, you would ignite a fire in the oil at the producing horizon, then push air down into this well at phenomenal rates, up to 900 cfm (cubic feet per minute, or 25 cubic meters per minute). The heat and the pressure would “loosen up” the heavy crude around that starter well, and drive it towards the surrounding peripheral wells. There was one immediate problem: the producing horizon is never quite uniform, which meant that the fire reached one ring-well before the others. To keep this from blowing out explosively, the engineers put thermocouples down each of the peripheral wells. When one got too hot, they would shut down and cap that well, and the firewall would follow the pressure gradient to the other wells instead.

            I’ve seen drill core from these Kern Field wells before and after a fire flood passed through them: beforehand the cores looked like cylinders of black, sandy tar. Afterwards, the cores looked like solid white beach sand!         

            I said earlier that there was a separate problem, and my first introduction to this was startling. An engineer at a fire flood operation gave us a tour of the field. First, he opened a spigot on a pipe and out burbled a tan-colored (and SO₂-rotten-egg-smelling) goop into a bucket. We watched it for a moment, and then he led us away to see the huge fans used to drive the air down the central well to keep the combustion going. After an hour or two, he steered us back to the bucket. He asked me to lift it. With difficulty I did. “Turn it over,” he said. Gingerly, using gloves because it was still quite hot, I turned it over.

            And nothing poured out. That tan goop had turned solid long before it had dropped back to room temperature. Scratch that bucket.

            So, what do you do about this? It turns out that about the only thing you can do – besides adding some very expensive, environmentally-iffy chemicals like they do in the Athabasca Tar-Sands – is to make a slurry of this stuff. Mix the heavy crude with sweet crude from a different nearby field into a solid-liquid slurry, and it would then move down the pipeline to a refinery without clogging that pipeline. A slurry is like Bubble Tea – that sweet Japanese drink with liquid surrounding little balls of semi-solid tapioca. You can easily get it through the straw – but not if the tapioca settles on the bottom!

            As you might expect, “cracking” this heavy-crude-light-crude slurry was a lot more complicated than cracking the sweet crude into diesel and gasoline. So… this oil was technically accessible, but more expensive to turn into gasoline.

            We thus get back to that old physics conservation law: ain’t no such thing as a free lunch.

            There is a newer, more controversial technology called “fracking” (see the later chapter specifically with that name). This is another way to get certain “tight” rock formations – rocks like dark, carbon-rich shale formed in an ancient swamp – to surrender some of that carbon. The problem with “tight” rock is that it has very low porosity, so the oil, while abundant, cannot move around within nor escape it easily, no matter how much you heat it. Think of trying to squeeze water out of a towel… then think of trying to squeeze water out of a piece of leather.  It just can’t get out…

            Unless you shatter the leather. And that ain’t no free lunch, either.

            The next part of this explanation will address yet another example of human ingenuity: It’s titled “The Hubbert Curve – how we cheat it.”

            From T.S. Elliot’s poem “The Hollow Men” (conclusion):

This is the way the world ends

This is the way the world ends

This is the way the world ends

Not with a bang but a whimper.

            Actually, it will probably be a slow bang.

            At the Ask-a-Geologist desk, we have received quite a number of end-of-the-world queries. These could be consolidated into a single sentence with two parts:

            Q: (1) Are mass extinctions real, and (2) will another one happen soon?

            A: There has been accumulating evidence over the past century that animal life on Earth has been decimated repeatedly. The biggest documented extinction events (though not all) are:

• ~440 million years ago (the demise of the Bryozoa, among other fossil species, marking the end of the Ordovician period),
• 251 million years ago (the “Great Permian Extinction” or “Great Dying” that saw the disappearance of over 95% of all genera living at the time including the Trilobites),
• 219 million years ago (the end of the Carnian stage in the late Triassic period, coincident with the appearance of the huge, ~85-km Manicougan craters in Quebec, Canada),
• 65 million years ago (the end of the Cretaceous period and with it most of the dinosaurs), sometimes called the Chicxulub event, named for a village in northern Yucatan, Mexico.
• 33 million years ago (The demise of the Cassidaria family of mollusks near the end of the Eocene, after ancestors of the horse first appeared),
• 1 million years ago (the boundary between the Pliocene and the Pleistocene epochs).
• 40,000 and…
• 12,000 years ago.

            What could possibly cause all these extinctions?

            In the past century geologists have come to realize that the Earth’s crust doesn’t change gradually, but instead it apparently evolves episodically – something usually called “punctuated equilibrium.” The “punctuations” generally take one of two general forms: asteroid or comet impacts, and episodic convulsions of the Earth’s deep interior.

            Let’s consider the first possible reason for extinctions: asteroids or comets.

            For some time now, astronomers have known about a 26-to-30-million-year cycle of our Solar System – that it oscillates in and out of the plane of the galaxy as it revolves around a supermassive black hole at the galactic core (Sagittarius-A* in the center of the Milky Way). There is a very rough (in other words, very arguable) periodicity in asteroid impacts mapped in the Earth’s crust. Some scientists put this periodicity at about 26 million years.

            The thinking goes something like this: as the Solar System passes through the plane of the Galaxy, there are more close approaches by other stars, which disturb the previously-stable orbits of Oort belt objects. These are icy planetesimals orbiting far beyond Kuiper Belt objects like Pluto and Sedna, reaching out to 50,000 astronomical units from the Sun (up to a light year away). The Oort belt is where most of the comets come from. Thus, a disturbance out at this distance could send one or more into the inner Solar System. These may directly impact the Earth, or may disturb or deflect one or more asteroids orbiting between Mars and Jupiter. Asteroids are far more common in the mid-to-inner Solar System, but comets generally have a much high relative velocity with respect to Earth. Since kinetic energy goes as the mass times the velocity squared, a comet could potentially do quite a bit more damage for the same size/mass if it impacted the Earth.

            For more than a century now, scientists have been aware of these extinctions in the paleontological record. The cause of the great Permian Extinction of 251 million years ago is still not fully understood, but may be related to huge seafloor craters now known to exist off the northwest coast of Australia (Bedoubt) or the Falkland Islands east of Argentina. The extinction of the dinosaurs 65 million years ago actually has a ‘smoking gun’: a huge, 150-to-180-km crater now lying beneath the northern edge of the modern Yucatan Peninsula of Mexico. There is other evidence: ginormous tsunami deposits elsewhere in the Caribbean including Haiti, a tektite strewn field throughout the American southeast, and distinctive fragments found in Montana, and in eastern Pacific Ocean deep-sea drill cores thought to be from this impact.

            Keep in mind that the Earth’s crust is a very dynamic place; while we see thousands of craters on the Moon, we see few on the Earth. Careful mapping has identified only 184 asteroid-impact craters on the Earth, even counting the tiny recent ones like Wabar in Saudi Arabia, and Henbury in Australia. The Earth’s crust is evolving constantly because of plate tectonics and weathering, so any evidence of impacts is steadily being erased.

            The second possible reason for mass extinctions: gargantuan volcanic eruptions.

            An article in EOS, the Transactions of the American Geophysical Union (Rampino, et al., 2013), points out that there have been roughly cyclic episodes of large igneous provinces (LIP’s). The Deccan Traps, making up much of western India, is just one of these provinces: kilometers-thick, near-continent-sized basalt flows all erupted over a fairly short window of time (geologically speaking). A vast basalt province in Siberia called the Siberian Traps, and the huge Columbia River basalts are among the others. These are thought to be the result of large upwelling mantle plumes; for scale imagine the US east of the Mississippi River being covered by miles-thick flows of basaltic lava.

            The geologic record shows these LIP’s to have occurred around

• 390 million years ago
• 295 million years ago
• 251-250 million years ago (the Siberian Traps)
• 200 million years ago
• 185 million years ago
• 135 million years ago
• 100 million years ago
• 65 million years ago (the Deccan Traps occurred close in time to the Chicxulub impact, causing some confusion about a possible connection between the two events)
• 30 million years ago
• 17-14 million years ago (the Columbia River Basalt province).

            From these ages, frequency-domain filtering (and your eye if you plotted them out) suggests an apparent rough cyclicity of 28-to-35 million years, especially prominent starting 135 million years ago.

            When volcanic centers this size erupt, there is a huge degassing process associated with it: sulfur dioxide and vast amounts of carbon dioxide are released. When Mount Pinatubo erupted in 1992, it sent a proportionally smaller cloud of SO₂ into the stratosphere – and the Earth’s average temperature cooled for two years afterwards. And that’s just from what happens in the stratosphere, from a single point volcano.

            Could there be a third reason for mass extinctions?

            Around 40,000 years ago, most of the large animals of Australia abruptly disappeared. These included the rhino-sized, wombat-like marsupials called Diprotodons, giant 200-kg kangaroos, a goanna bigger than the modern Komodo dragon, a giant goose-like bird twice the size of the emu, and many others. These animals had survived at least two episodes of climate change prior to 40,000 years ago. In North America about 12,000 years ago, most of the large “charismatic megafauna” of North America (mammoths, giant sloths, camels, cave bears, saber-tooth tigers, etc.) suddenly disappeared. In both cases, these mass extinction events (and a more recent event on Madagascar that is still very much on-going) correlate closely with the arrival of the human species in these regions. The implication of overhunting is hard to miss here. As the human population surges past 7 billion today, the largest mass extinction in the past 65 million years is fully underway, and the Passenger Pigeon is just one of the first and best-known victims (Kolbert, 2014). Habitat loss, overhunting, and accelerating climate change are the proximate mechanisms for this current and stunningly rapid mass extinction event.

            The End of Things As We Know It

            There are Near Earth Objects (NEO’s) out there that NASA and the US Air Force are monitoring (the number keeps growing, but at least the search process is now automated). Based on their known sizes (we can generally only see the big ones) and what happened at Chicxulub 66 million years ago, many of these could wipe out human civilization as we currently know it… not if, but when, one hits us.

            If Yellowstone (just one of several known supervolcanoes) unzipped tomorrow, it would cover the eastern two thirds of the United States with a vast blanket of ash, suffocating most living things within the first several hundred kilometers in all directions. The gas-release and the peripheral consequences would devastate the entire planet. To put things in perspective, the last eruption 640,000 years ago left an off-white layer 20 meters (65 feet) thick called the Pearlette Ash Formation near Colorado Springs… 800 miles away. Nothing survived under this ash blanket. I have personally pulled a camel’s tooth from the bottom of this formation.

            However, the real problem may be even more imminent. The greatest mass extinction in the history of the planet is underway right now (Kolbert, 2014). The levels of carbon dioxide in the atmosphere crossed over 400 ppm in 2013 (Showstack, 2013), and the rise is accelerating.

            It reached 415 ppm in 2019 (https://www.sciencealert.com/it-s-official-atmospheric-co2-just-exceeded-415-ppm-for-first-time-in-human-history ).

  As Pogo said, “We has met the enemy, and it is us.”  

What can we predict, exactly?

According to my calculations, the 6th grade means students are around 11-12 years old. If so, then the Rising Generation is full of people a lot smarter than I was at that age. The question below from Ask-a-Geologist is just one of many like it:  

Q: Dear Geologist,

Our name is Arianah and Cray and we are sixth grade students at Preston Middle School in fort Collins, Colorado. We are currently learning about how the Earth’s surface changes over time. We are curious about earthquakes. We have a couple questions for you. Is there a common time when earthquakes happen during the day? Also, why did you become a geologist?

Yours sincerely, Arianah and Cray 😀

A: 1. Earthquakes are essentially random. We understand why they happen, we understand where they happen, but we do NOT understand WHEN they will happen. There are always aftershocks following a main event, of course, but the main event cannot be predicted. Extensive research has shown that there is no correlation between earthquakes and certain times of the day or external * events – for instance there is no correlation with either the location of the Sun, or of the Moon, or with tides (alignments of celestial bodies, which cause neap tides or spring tides, is called syzygy). Some of the brightest minds on this planet have been searching for more than a half century for some evidence that main event earthquakes can be predicted, but without success. They can be forecast #, but not predicted.  

2. I was a solid-state physicist with a masters degree, and realized that if I didn’t do something drastic, I would be stuck inside a laboratory all my life with radioactive sources and high-pressure cells. This was brought very much to my attention one day when I had a high-pressure cell blow out and spew Cobalt-60 all over the inside of our lab, and had to call in a special Spill Team. Also, by this time physics as a profession was drifting into a dead end with string theory and meta-philosophy about the un-testable multiverse, and I saw relatively little value to humanity to spending billions of dollars to see if another exotic particle existed. I checked out break-offs of physics, including astrophysics, hydro-geophysics, weather physics, and geophysics, and found the last one to be very exciting. It also got me out into exotic places, like the Venezuelan jungle, the southeastern Alaska wilderness, inside an erupting volcano in Kamchatka, the Empty Quarter of Saudi Arabia, etc. Geoscience gives me amazing opportunities to visit these places and many more. My wife and five kids came with me to many of them – and all ended up being multilingual.

But even more interesting to me is the opportunity to be a detective – to be the first to discover something beneath the ground or the seafloor. I was the first to say in detail where the groundwater lies beneath the huge San Pedro Basin in Arizona and Sonora, Mexico, a host to one of four major migratory bird flyways, and a marine geophysical technology I developed was the first to map where titanium sands lay hidden beneath the seafloor off the coast of South Africa. That is ever so cool.  

* It has been shown that if you inject fluids into certain formations (e.g., deep sediments northeast of Denver, CO), you can trigger swarms of micro-earthquakes. Basically this is the ground shuddering as it tries to equilibrate and adjust itself to a slightly new stress regime. However these sorts of events are so small that they are almost never felt.They really are not earthquakes as the general public understands earthquakes. 

# A forecast is not a prediction: a forecast means there is an X percent chance that there will be a magnitude Y event on the Z fault zone in northern California within the next 30 years. This is very, very different from saying that there will be a Magnitude Y event at Z location on X day – that would be a prediction. Science can’t do that.

~~~~~

          We get a LOT of questions about volcanoes, including how to “know if she’ll blow.” There are a number of ways we can track magma movement at depth, including monitoring deformation and tracking “LP’s” – long-period (low-frequency) seismic tremors indicative of deep fluid movement. At late stages of unrest, we will start seeing “Vt’s” – short-period volcanic seismic evidence of rocks breaking – and often dramatic increases in Carbon-14-depleted Carbon Dioxide and Hydrogen Sulfide (burnt-match smell) gases. There is a good possibility that we can detect very early movement of magma at 30 – 40 km depths using Magnetotelluric systems, something I proposed many years ago, but there hasn’t been enough funding to try this. As I write this, edifice deformation still currently reaches out the longest time ahead of all these detection systems to give us warning of an impending eruption up to months ahead of time.      

  The term “deformation” is used by specialists in ground movement in the geosciences; these guys call themselves “geodesists”. Geodesists measure movement as a component of strain along an active fault, to try to get a sense of the elastic energy accumulating that could lead to an earthquake. Deformation is also used in volcanology to look for – and then track – inflation in a volcanic edifice. Deformation is measured in several ways:

            1. Surveying the ground with high precision. This has been done at Yellowstone since the mid-1920’s, and those early data have helped us get a much better sense of how the huge caldera moves and breathes over time. It’s not at all unusual to find a section rising 20 cm (7.5 inches) in a few weeks. Typically, another part of the caldera will be deflating at the same time.

            2. Deploying tilt-meters. Originally these were long tubes of water laid out over the ground. If the ground under the flank of a volcano started tilting, a very tiny vertical movement would show up in amplified displacement of water in vertical tubes at the end of the long horizontal tube. Modern tiltmeters are ultra-sensitive cylinders placed in a vertical hole in the volcanic rock, then packed in with sand. The signal from these devices and all the following systems is generally telemetered back to a recording and monitoring station.

            3. Radar satellites – this approach is called InSAR (Interferometric Synthetic Aperture Radar). If two radar images can be captured over the same volcano, spaced days or weeks apart, they can be used to make interferograms. These look like colored Moiré patterns (see figures 47 and 48 below). They are generated with prodigious mathematical calculations to geometrically correct, and then ratio each individual pixel to another taken on a different satellite pass. These complex geometric corrections are sometimes called “rubber sheeting”, and the final result will show inflation over the surface of a volcano and its environs. Each rainbow-colored ring-set represents one radar wavelength (typically 5 – 15 centimeters) of uplift. These often form spectacular bulls-eyes centered over an inflating volcano or a deflating caldera, and I’ve seen several gorgeous examples at Ngiragongo volcano, in Central Africa; at Pavlof, Akutan, Okmok, Shishaldin, and many other volcanoes in the Aleutians (figure 47), as well as Mauna Loa (figure 48) and Kilauea volcanoes in Hawai’i.

            4. Gravity level-lines. This is like the precision survey leveling discussed earlier, but is done by making repeat measurements with three gravity meters over a line of stations every six months or so. It saves you having to traverse an extremely rugged volcanic field on foot to survey it (you can use a helicopter). All other things (including the water table) being equal, an inflating volcano will show up as a decrease in the gravity field – the station site that the gravimeter is sitting on is being moved upwards. If the station moves farther away from the Earth’s center, the pull of gravity falls off as 1/distance to the center of the Earth squared. For years I did this kind of survey every six months to monitor magma moving into the Harrat Rahat volcanic field, located east of Madinah al-Munawarrah (“Medina”) in Saudi Arabia. Seismic telemetry also showed small earthquakes associated with this magma movement into the Earth’s crust at the same time. The events died out before an eruption, and before I moved back to the United States, causing a lot of people to breathe a collective sigh of relief. This kind of on-again-off-again restive behavior is not at all unusual for a volcano. This same volcanic field erupted in 1256 AD and nearly wiped out the city of Medina at that time, so we take these things very seriously.

            5. Telemetered GPS. These use the same GPS satellites that you and I utilize in our cars or when hiking, but the precision measurements made by geodesists use different signals (called “p-code”) from the same satellites. This gives far higher precision when averaged for a fixed location and the data are stacked over a period of time. If two GPS stations on opposite sides of a volcano are moving apart… then something is filling the edifice in between (see figure 48).

            6. We also instrument volcanoes with sensitive analog seismic sensors, and ultra-sensitive broadband seismic sensors. Some of these data are telemetered, some are recorded and just stored in the instrument box on a small hard-drive over the winter months until retrieved the following summer. That is, unless bears decide to play ball with one (which happens). One over-winter seismic network campaign at Katmai volcano in Alaska found 5 of 11 of our very expensive stations had been trashed by bears before the geodesists could get back to retrieve them.

            GPS is a fascinating field, and applies all through and far beyond the earth sciences. A brief run-down might be useful here.

            The Global Positioning System was first envisioned by DARPA – the Defense Advanced Research Projects Agency of the Department of Defense – during the 1980’s. Navigation at that time was complex and difficult, and getting any sort of location precision over vast distances including oceans was very important to some people. Korean Air flight 007 was shot down by Russian warplanes in 1983 over what proved to be a small navigation error. Also, people targeting ballistic missiles would like to place them exactly on top of a hardened missile silo.

            In the late 1980’s I was a young scientist working in the Venezuelan jungle, where our main form of navigation was to use 1:250,000-scale side-looking airborne radar (SLAR) maps. These were assembled by flight strips – radar images taken from high altitudes – and it was not unusual to find splice errors as large as 3 kilometers between strips. That means I could be standing on a single rock – and according to the SLAR map, half of the rock was 2 miles away from the other half of the rock. It gives the expression “huge strides” quite a different meaning. I have been on a helicopter traveling for an hour over trackless forest using a half-meter-sized, million-scale roadmap of the country (except there are no roads in the jungle) and crudely-penciled lines with the azimuth, and distance in flight-minutes to the target site that we wanted to visit. If that helicopter’s fuel line had a single bug in it, we would have dropped down into the trees. Even assuming we had survived such a crash (the incident statistics gave me a 50% chance of surviving this sort of crash), how would you call in a rescue helicopter? How in the world would you even tell rescuers where you were, when the wreck-site could not be seen beneath the jungle canopy?!? Your location and its accuracy become a life-and-death issue very quickly.

            I first began using a GPS device in the early 1990’s in Saudi Arabia. In the northern reaches of the country there is a vast plain (the al-Jalamid Plain) that is dead flat for hundreds of kilometers in all directions. Some of our staff had accidentally strayed across the Iraqi border, because there is no way to know where the boundary line (arbitrarily drawn by the British almost a century earlier) actually was. The first GPS units were incredibly slow, the size of a Betty Crocker cookbook, and didn’t always work – but the idea fascinated me. With a radio, I could then tell people where I was within a hundred meters or so at that time; we could now find someone and perhaps save a life.

            Since then, hand-held GPS devices have shriveled to coin size (I have one on my wrist, tracking my steps and bike-rides), and many have maps built in. You can program them, collect precise tracks of where you walked or flew… the list of bells and whistles goes on and on.

            But how do they work? What actually is out (or up) there?

            The American GPS constellation has at any given time about 24 active satellites and a few loitering backup spares, and each one transmits a very faint signal on two frequencies – a digital signal for hand-held device use and another digital carrier (the “p-code”) that is used for more precision location acquisition. I’m talking about centimeter-size precision here. Both GPS frequencies were originally encrypted by the US military. Because the satellites belonged to them, for a long time the signals were deliberately “fuzzed” – this was called Selective Availability, or SA for short. If you had the digital key and a certain type of book-sized device, you could get relatively precise locations – within 10’s of meters. But the US Department of Defense didn’t want someone else using those same signals to pop an artillery round on top of an American military outpost. Even to this day, if you try to use a receiver and go faster than a commercial airliner (as in: a ballistic missile) it won’t work. It has a built-in fail-safe.

            In the meantime, the rest of the world has become incredibly dependent on the American GPS constellation. Aircraft you fly in now use them to navigate – even use them to land with in bad weather. I could never summarize adequately all the ways and places where GPS is used right now. As I write this, the Russians have their own satellite constellation called GLONASS, the European Union has one called GALILEO, the Chinese have the BeiDou Navigation Satellite System (BDS), and the Indians have the Indian Regional Navigation Satellite System (IRNSS) in orbit. In each case, these were launched to avoid dependence on the United States for anything that was mission-critical. Except creativity, perhaps.

            If you are surveying – or trying to see if two points on the opposite sides of a volcano are moving apart from each other (uh-oh), then you need great precision. With modern GPS technology, this precision can now be better than a centimeter horizontally and 2-3 centimeters vertically. In part this difference in horizontal vs. vertical precision is because for horizontal solutions you can subtract the atmosphere effect from two different near-horizon satellites – and also triangulate better. For vertical elevations, you have only satellites in one direction (above you). You cannot detect a GPS signal from beneath your receiver because the signal would have to travel all the way through the Earth.

            GPS signals all use similar frequencies, but the signals are encoded differently to separate the satellites. As implied earlier, signals from each satellite are encoded, so you can’t use one for a ballistic missile guidance system unless you own the codes. As I said, above a certain aircraft speed, GPS won’t work.

            The Russians and Chinese certainly didn’t want to be dependent on some signal that the Americans could fuzz – or even turn off. So, despite their crushing economic difficulties, they turned the best Russian minds onto building their own GPS constellation. The GLONASS signal is not encoded, and the energy transmitted is greater, so the signal-to-noise ratio is 5 times or 15 db better. Because of this, the signal penetrates tree canopy, so I could now use it in the trackless jungle where I used to work. The GLONASS system also uses 3 different frequencies, so you can reduce ambiguities and better calculate the complex differential atmospheric corrections. The GALILEO, BDS, and IRNSS systems are similar.

            These five GNSS (Global Navigation Satellite Systems) are so precise that they even routinely calculate and correct for relativistic effects (called “frame dragging”)! There are also huge atmosphere effects that must be compensated for – dense air masses here and differentially ionized layers there. GLONASS even works on new American and European hand-held devices when the GPS signals are poor due to a restricted view of the satellite constellation – if you’ve ever been in steep canyons in southern Utah or New York City, you know what I mean here. Not wanting to be left behind, the American version of GNSS (and the only one that should technically be called “GPS”) is being upgraded.

            All five of these GNSS systems use L-band frequencies to resolve ambiguities and increase precision – and penetrate the ionosphere. What does L-band mean? Look at your personal car’s GPS antenna-fin, and the smallest dimension on it will give you an idea of the wavelength for L-band. That’s the size of your optimum-reception GPS antenna, by the way.

            The navigation problem is more than just triangulation. Three satellites near the horizon would serve for this; two would give you two possible location solutions, and three would reduce this to only one possible solution. But there are four unknowns in this complicated system, since you must also measure how long the stretch of space and air that each signal must travel through. The precision of your timing thus becomes utterly critical, because the speed of light is so huge (~300,000 km/second), and hand-held GNSS devices cannot carry $100,000 maser clocks. Thus, you must use a 4th satellite to help solve for the 4th unknown: 3 for position, 1 for a clock reference for your receiver. This part is basic high-school matrix algebra that I learned as a sophomore.

            There are a few more complications. Because of the 3D density variations of the Earth’s atmosphere, you can get even better precision if you also reference a ground station. This will give you even better differential distance calculations – to do good back-corrections for the changing satellite orbits, the complex and varying atmosphere, rain, and snow cover, etc. However, during the Tohoku earthquake in early 2011, all of Japan jerked eastward about 10 centimeters/4 inches. Geodesists couldn’t see the whole shift with really great precision – because all their reference stations also jerked eastward.

            So back to the case in hand – how does this help volcanologists? As I wrote earlier, if two telemetered GNSS receivers are moving away from each other, and there is a volcano in between them (this has been happening for years now with Mauna Loa, the largest volcano on Earth), then you are being given a warning that something is moving into the volcano’s edifice – the space in between the stations.

            In 1989 we didn’t have such a warning before Redoubt volcano in Cook Inlet of Alaska erupted. A KLM Boeing 747 flew right into the ash cloud – and lost all four engines in rapid succession. I’ve listened to a recording of the captain’s voice as she tried to guide her flight crew in Dutch and talk with flight control in Anchorage in English. Her voice rose steadily a full octave before she finally yelled “Anchorage we have lost all four engines, we are in a fall. We can use all the help you can offer.” They somehow managed to restart two of the engines, and made a rough landing at Anchorage International airport. No lives were lost – but the repairs to make that Boeing 747 flight-worthy again cost $80 million.

            To put that in perspective, when I served as chief scientist for volcano hazards for the US Geological Survey, my entire science team’s annual budget was less than $20 million. There are over 1,600 volcanoes around the world.

            There is another interesting GNSS application that you might find fascinating – I sure did. When Mount St Helens erupted on 1 October, 2004, we had just a week of accelerating racket on our seismic network beforehand for a warning. The extrusion was first seen on October 12 – and by pure luck I got the first photo of the new “spine” from a helicopter orbiting the steaming and fractured Crater Glacier. This dacite (63% – 68% silica) extrusion was 700 degrees C where it was coming up from the talus slope at its base. It came out like a tube of squeezed gray toothpaste. In fact, it resembled the back of a whale, so that became the name for the first of several domes that would be extruded: The Whale. It moved south through crumbling talus and ice until it hit the remaining south rim of the 1980 eruption. The geodesists wondered when it actually reached that wall – “When Did the Whale Hit the Wall?” A check of a GPS station on the other side, on the outside south slope of the volcano, answered the question. On November 17, 2004, that station suddenly started moving southward. Was it an effect of snow on the antenna? No, because the only direction it moved was south – by about 10 cm. The entire crater wall was shoved southward by about 4 inches by the growing dacite dome hitting it on the other side.

            I’ll never forget the elation of scientists using GPS technology to answer a real question about an erupting volcano. But GNSS systems provide us more than just answers to our scientific curiosity.

            In 2005 a sharp-eyed geodesist then based in Anchorage, Alaska, was routinely checking data from several GPS units installed on Augustine volcano in the middle of Cook Inlet, south of Anchorage. This had erupted in 1979 and nearly killed David Johnston, one of our brightest young geologists, who was then killed just a year later during the May 18, 1980 lateral blast, the opening eruption salvo of Mount St Helens.

            In August 2005 our Anchorage-based geodesist noticed some differential movement between two GPS stations – the first subtle signal that inflation was starting – and notified the Alaska Volcano Observatory Scientist-in-Charge. A daily check-and-monitoring effort was started. Sure enough, the signal was real, showing above all the background noise – and it was continuing. Federal and State Emergency entities, along with the FAA (Federal Aviation Administration), were put on notice. In Late December the first VT’s – rock-breaking signals – started appearing on the seismometers. As they accelerated in frequency and amplitude, the USGS issued a warning: an eruption is imminent in hours or days. One day later, on January 16, 2006, Augustine erupted, and dusted Anchorage with ash. International flights were cancelled or re-routed for three days – but not a single aircraft was damaged, and not a single life was lost.

            Yeah! This deformation stuff really works! And more to the point, it saves live.

Yep, the lights are definitely gonna go out.

This year we celebrate the formal 50^th anniversary of chaos theory and the end of deterministic predictability. The latter is the wonderfully intuitive idea that if you understand the physics principles and the starting parameters, you could then predict where a system would be at any time in the future. It’s seems intuitive: set up a row of dominos and you know exactly how things will end when you tip the first one, right? 

That everything-is-predictable-from-the-beginning-conditions idea, however, got us into the Viet Nam war.

Around that same time in 1963 a meteorologist named Edward Lorenz showed that this idea was a failure, at least for most systems beyond what you could set up on a table. The reason? Most “systems” (like the weather, or transform faults, or power grids) are not linear, nor are they simple, like a row of dominos. A slight change in an initial parameter in your weather model (the temperature at one of millions of points over the tropical Atlantic, the amount of dust sent westward by a sandstorm in the Sahara a week earlier, the number of sunspots on the approaching limb of the Sun, etc.) and your final calculated outcome for the Atlantic Hurricane season can be totally different from what your computer had calculated just an hour earlier.

DANG, you say. I really need to know when a hurricane/tornado is going to hit! I’m running a BUSINESS here for heaven’s sake.

Well, don’t give up hope – we can’t predict earthquakes, but meteorologists have made huge progress in the past two decades. Nate Silver in his book “The Signal and the Noise” points out that massive new computing power, coupled with the vaster integrative capacity of thousands of human minds, have together contributed to huge progress in predicting weather out for more than a week at a time. They correctly and precisely predicted the landfalls of Hurricanes Katrina and Sandy. For human reasons, however, very few people took the warnings for Katrina seriously and thus many people died as a result.

However, weather forecasters still can’t predict when a hurricane will start (though they know when the hurricane season will “light up” their boards), nor can they predict very well the power of these monsters when they hit. The damage usually comes not from the wind, but from the so-called low-pressure storm surge that lifts the ocean up 5 or 10 meters as the eye of the storm approaches land. Think: a 20-foot wall of ocean pouring in on your neighborhood at 20 miles per hour. As an interesting anecdote, Nate Silver shows that local television weather people have a truly abysmal predictive record, far worse than the NOAA weather forecasters, whose data they can easily access for free. How could this possibly be? The reason for this is very human: no forecaster wants to get flogged for UNDER estimating the likelihood of precipitation.

Let’s go back again to the foundations of predicting things. This is, basically, prophesying. Deterministic predictability actually does hold, at least theoretically. The problem is having ALL the parameter data PRECISELY correct in your weather model. It is also important to have a computing grid fine enough that when you do your calculations the temperature and pressure on any given point is not that different from that of any adjacent point. In other words, so the point-to-point behavior can be treated mathematically as approximately linear. Some of Lorenz’ earlier computer models used to try to predict the weather gave different results when run more than once. What? But everything input was the same! Not quite, it turned out. The starting numbers were returned to the computation with only the third decimal place retained – in other words the numbers were rounded up. 26.2653 became 26.265 – and the final results were startlingly different. It took Lorenz awhile to realize this, but there was a big clue down there in the minute decimals.

Classical physics teaches that given the current state of a system, all future states can be calculated. It seemed to work in the 19^th Century: it was used to predict the orbits of planets and comets, and slight perturbations successfully guided the search for Uranus and then Neptune (and in 1930 a small perturbation in the orbit of Neptune led to the discovery of Pluto, though that case is arguable). 

However, back in the 1880’s, Henri Poincaré was studying the three-body problem, in which three bodies continuously influence each other in celestial mechanics in complex and overlapping ways. Poincaré noticed “…that small differences in the initial conditions produce very great ones in the final phenomena.” He concluded that prediction is impossible for three bodies orbiting in space. Contemporaries thought they just had a data quality problem, but the root was much deeper than that.

So chaos theory, but without that name, preceded Lorenz by nearly a century. Chaos theory, by the way, has a common metaphor that is fairly widespread: the so-called “The Butterfly Effect”. This stems from the title of Lorenz’s 1972 presentation to the American Association for the Advancement of Science: “Does the flap of a butterfly’s wings in Brazil set off a tornado in Texas?” This is also called ‘sensitive dependence on initial conditions’, and it’s a trademark characteristic of a complex non-linear system. On the other hand, the trademark behavior of a chaotic system is apparent randomness – but this is deceiving. Determinism actually works, but you have to know ALL the initial data and ALL the force actors to high precision.

~~~~~

Well, what has all this got to do with the electric grid in the title? In the United States there are really three quasi-independent power grids: The Eastern Interconnection for the entire eastern US to about the Kansas-Colorado border, the Western Interconnection from there to the Pacific coast… and the Texas grid. We always knew Texas would insist on being different. It may surprise you to know that these grids are the largest engineering structures ever built, and consist of thousands of energy sources from coal-fired power plants, to the huge Bonneville and Grand Coulee Dam hydro-electrical generators, feeding ultimately to billions of power outlets in our homes. These systems affect virtually every aspect of our day-to-day lives. If you are reading this, it means your grid is working.

However, within each of these domains – and increasingly across their boundaries – a perturbation in one place will cascade across the rest of the network with usually unpredictable consequences.

While there are power generating stations everywhere throughout the three grids, there are powerful sources of irregularity in the entire system. Wind energy sources can drop suddenly, and the growing solar input systems are diurnal (they produce nothing at night), or a power plant may go offline for maintenance. Furthermore, a Coronal Mass Ejection (see  http://jeffwynn.blogspot.com/2012/01/cme-events-how-they-affect-your-life.html) can send a huge bolus of charged particles at our planet. The Earth’s magnetic field is a pretty good defensive barrier, but it can be – and has been – beaten down to the ground. When that happens there are huge telluric currents set up – vast flows of electricity along the ground. When this hits a power substation it can cause huge shorts in the giant accumulators. If you’ve never seen a power transformer “pop”, then you are in for a spectacular surprise as long as you are not next to it. I’ve watched video of a tornado approaching Oklahoma City, and its approach is marked distinctly by bright flashes as these pole-top transformers explode.

When a small transformer like this goes down, it blacks out a part of the network and is repairable within a few days at most. When a larger accumulator explodes in a power substation, it’s a different matter, and there will be huge surges of power coming in on the grid to try to compensate for its loss. Enough of these kinds of events and the instability they bring will cause vast areas to go down. 

The most famous of these events happened in the summer of 1965, when New York City was blacked out. Interestingly, there was a huge surge of births in the area precisely 9 months later. More recently, a CME shut down the Canadian provinces of Quebec and Ontario, when they experienced a huge and long-lasting blackout in the middle of winter. If electricity is your source of heat, this could be a life-threatening event. If you survive, your water pipes will freeze and burst, and you will have heck to pay when it warms up again.

When these surge-and-sag events happen, human operators jump in and try to stop the cascading failure from propagating. But they don’t always succeed, in part because the entire grid is fundamentally a non-linear system, sensitive to the tiniest things. In other words, we can’t predict ahead what is going to happen to our home power supply, because there are too many variables involved and we don’t understand the behavior of the system except in statistical ways.

But the human and growing automation reasons for grid instability are perhaps the most interesting – and the least predictable. Thousands of induction motors in air conditioners can all surge at once and drag down (“brown out”) the entire system when a sudden heat wave hits California, or New York, or any other major collection of humanity. As more and more renewable energy sources come online, the points of failure and surge grow even further. Newer smart appliances just add to this mix because human control steadily diminishes.

It’s perhaps not really surprising, then, that Chinese military hackers have turned their attention to the North American power grid, and have persistently probed the computer control systems monitoring and adjusting against just these sorts of failures.

Yes, chaos theory rules our world. Another way to say this is that our small part of the universe is chock full of nonlinear systems, including especially humans, and nonlinear systems are very hard to forecast.

~~~~~ 

One final quote, this time from the famous mathematician Pierre Simon Laplace: “An Intelligence which could comprehend all the forces by which nature is animated and the respective situation of the beings who compose it – an Intelligence sufficiently vast to submit these data to analysis… for It, nothing would be uncertain and the future, as the past, would be present to It’s eyes.”

~~~~~

Repeatedly I have had questions about climate change addressed to me, both electronically in Ask-A-Geologist, and verbally from acquaintances  There are a lot of things floating around in the “news media” about climate change. A lot of this is correct, some of it is foo-foo, and far too much of it is deliberate obfuscation by people who have an agenda.

Q: Hi!

Oxygen in our atmosphere was created by small creatures who had just invented a new process:  photosynthesis.

The waste product of photosynthesis is oxygen.

After an unfathomable number of  years, too much O2 built up in the atmosphere, changing the greenhouse gas, methane, into carbon dioxide, which isn’t such a strong greenhouse gas.  This caused the earth to cool off to the point where the first known ice age began, the huronic, I think.  It lasted for millions of years and enveloped the entire earth with ice.

Question 1: how did these photosynthetic creatures survive in an iced-over earth?

Question 2: what caused the end of this very long ice age?

 Just curious – Thank you!

 – Susan K

A: Your question suggests you are well along in studying this topic. I’ve stored a number of closely related questions and answers on this blog, and by way of a long answer, some of these may help:

http://askageologist.blogspot.com/2013/07/climate-change-is-it-real.html

http://askageologist.blogspot.com/2013/04/castastrophes-and-mass-extinctions.html

http://askageologist.blogspot.com/2012/06/snowball-earth-faint-young-sun-paradox.html

The medium-length answer: Our atmosphere passed through the oxygenation transition around 2.5 billion years ago, and it certainly involved photosynthesis – stromatolites (fossil algal clumps) have been found dating as far back as 3.3 billion years. However, there are also suggestions that mantle out-gassing, tectonics, and oceanic current-shifts may have contributed. There really was a Snowball Earth episode, and there have been a series of cold-warm cycles since then. Scientists have been exploring – with limited data considering how long ago it was and what little evidence remains – what could have caused these events. Suggestions range from the fairly mundane to the exotic: asteroid impact, tectonic change that interrupted oceanic current flows, etc. The Chicxulub asteroid event 65 million years ago certainly knocked the oxygen levels in the atmosphere down dramatically, requiring millions of years to recover. This is almost certainly why bar-headed geese can easily fly over Mount Everest: birds have evolved a truly advanced respiratory system (including hollow bones) since that event. Evolution is well-documented to speed up under environmental stress.

Some short answers:

1. If there is one thing certain about microbial life, it is that it can survive almost anything. Microbes have been found many kilometers deep in the Earth, despite the fact that temperatures steadily increase with depth due to radio-isotope decay in the Mantle and Core of the Earth (the temperature rises to typically 60 C at 4,000 meters depth).

2. There are a lot of variables that may have been involved in the recurring cool-warming cycles, including the fact that the Sun has steadily grown in luminosity during its 5-billion-year lifetime, as well as tectonics, and methane-emitting life forms. Likely a combination of these – and probably additional factors – led to out-of-control feedback loops that dead-ended in climate extremes before the atmosphere eventually  recovered.