Deterministic Predictability and the Power Grid

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 the 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.”

~~~~~

This is the Way the World Ends

From T.S. Elliot's The Hollow Men:

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. 

~~~~~

Catastrophes and Mass Extinctions – Are they periodic?

Q: 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: 

Are mass extinctions real, and 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 extinction events:

• ~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” 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 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 horse ancestors first appeared),
• 2.6 million years ago (the boundary between the Pliocene and the Pleistocene epochs)…
• …and 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, either, but instead it apparently evolves episodically. This takes two general forms: asteroid or comet impacts, and episodic convulsions of the Earth’s deep interior. 

The first possible reason for extinctions: asteroids or comets.

For some time astronomers have known about a 26-30 million year cycle of the Solar System, oscillating in and out of the plane of the galaxy as it revolves around a supermassive black hole at its 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. 

The thinking goes something like this: as the Solar System passes through the plane of the Galaxy, there are 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 such as Pluto and Sedna, reaching out to 50,000 astronomical units from the Sun (up to a light year). 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 the mass times the velocity squared, a comet could potentially do quite a bit more damage for the same size if it impacted the Earth.

For more than a century scientists have been aware of these extinctions in the paleontological record. The cause of the great Permian Extinction of 250 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 eastern Pacific ocean deep-sea drill cores.

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 170+ asteroid-impact craters on the Earth, even counting the tiny ones like Wabar in Saudi Arabia, and Henbury in Australia. The Earth’s crust is evolving constantly because of plate tectonics and weathering, so evidence of impacts is steadily being erased.

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

A recent article in EOS, the Transactions of the American Geophysical Union (Rampino and Prokoph, EOS 94, No 12, 19 March 2013, p. 113-114), 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 one of these provinces: kilometers-thick, continent-sized basalt flows all erupted over a fairly short window of time. 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 eastern US 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
• 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 to the Chicxulub impact, causing some confusion about the relative effects of 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.

Could there be a third reason for mass extinctions? 

Around 40,000 years ago, most of the large animals of Australia abruptly disappeared. These include 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 the first and most obvious victim. 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 65 million years ago, most 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 all living things. The gas and 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’) thick called the Pearlette Ash Formation near Colorado Springs… 800 miles away. I have personally pulled a camel’s tooth from the bottom of this formation.

However, the problem may be more imminent. 

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

~~~~~

Home Invasions and Practical Statistics

I was awakened a few days ago around 4am by a thump downstairs. This is about the time that the newspaper is delivered to our front porch, so I rolled over and went back to sleep.

But a thunk sound below my bedroom at 1:30am a few days later was a different thing, and I got up and did a systematic search of my 3-level house. In truth, I was pretty sure the sound was a book falling off an unbalanced stack downstairs, or perhaps even a weird dream. How I did that search speaks more to my “weaponized” (NOT firearms) house and my background: I'm a Taiho-Jutsu sensei. Of the Japanese martial arts, this one is sometimes called “Jujitsu on steroids.”

Oh BOY! There may be a 1% chance I can USE this!

But first, back to home invasions. The expression alone conjures terrible images from lurid newspaper reports. If you are only robbed and beaten up, you are fortunate. Home invasion differs from a burglary in its violent intent, and this distinction appears to be largely American.

Home invasions are real, if very rare. Statistics on these are notoriously hard to find, in part because of different crime classifications in different jurisdictions (in many places these are classified separately as homicide, rape, kidnaping, etc.). However, the impact on the human psyche is not unlike the impact of shark attacks. As in most of these “news” events, the real killers are rarely mentioned:

• fishermen wantonly kill millions of sharks every year just for their fins; 
• handguns in the United States kill 1,130 times more people per year than lightning does, and 
• 31,672 more people are killed by handguns in the United States each year than are killed by sharks.

To put things in perspective:

HANDGUN DEATHS:
In 2011, there were 8,583 murders in the United States committed with handguns. However, there were 31,672 TOTAL handgun deaths - all the rest (73%) were suicides, with the occasional handgun accident lumped in. This is comparable to motor vehicle deaths in the US.

There is a message here: having a handgun ready to blast someone in your home is statistically a far greater threat to you than the statistical chance of any possible safety it may offer. Read on.

SHARK ATTACK DEATHS:
There were 118 reported shark attacks world-wide in 2011, with 17 fatalities. There were 47 “unprovoked” shark attacks in the United States in 2012, with just 1 fatality. You have to wonder about the “unprovoked” caveat: do a lot of people “mess” with sharks? The vast majority of shark attacks occur in Florida. One human death to several million shark deaths is a very unsportsmanlike ratio, IMHO.

LIGHTNING DEATHS:
There were 28 Lightning fatalities in the United States in 2012, and some 24,000 lightning deaths worldwide. Again, you have to wonder at the disparity in the statistics, as the US represents 6% of the world population. There are many possible reasons, including regions like the Pacific Northwest where lightning is a very rare thing, or perhaps people live in flimsier houses elsewhere.

Back again to home invasions. Are you at risk of one? A pattern of dealing in illegal and “recreational” drugs will dramatically increase your potential likelihood for being targeted. Living in Anacostia, Maryland, or south Chicago will also dramatically increase your risk, but this is probably no surprise to anyone in the US.

What can you do about it?

A HANDGUN IN THE HOME:
The reality here is dramatically different than what the National Rifle Association would have you believe. If a woman carries a handgun, there is an 80% statistical likelihood that it will be used against HER: in other words, she will be either shot with her own weapon, or pistol-whipped with it. That’s a pretty large statistical number - basically it is a high probability.


Another take-away: don't carry a pistol in your handbag.

If you’ve ever been to a gun-range, the following observation will be obvious:
It takes slow, calm focus and concentration to hit a human target at 5 meters (16 feet) distance. Imagine trying to do that in the dark, when charged with adrenaline, and both the shooter and the target are moving.

There is a video on YouTube where a traffic stop ends with a perp jumping out of a stolen car and firing several shots at the police officer. The dashboard cam shows that the officer then proceeds to fire at least 12 rounds at the perp, who is running in a straight line towards the nearby forest. NOT ONE BULLET ON EITHER SIDE HIT ITS TARGET.

In a home invasion situation, a handgun is an excellent means to poke holes in your house - and probably several neighboring houses at the same time. It will likely NOT protect you.

IS THERE A BETTER ALTERNATIVE?  YES

1. Actually, there are several. There are two excellent and inexpensive tools for this, in fact, and used together they are pretty effective:  The first is a cell-phone that has been charging near your bed. USE IT. Call 911 and the entire conversation will be recorded. 
2. The second tool is a Mag-Light. This has a steel shaft, and a six-C-cell version is the sine qua non, if hard to find. Held on your shoulder, it can be used to both momentarily blind an attacker in the dark and strike a devastating blow. It works like an ASP baton, but it is significantly heavier.
3. There is also a Taser, if you can afford one (and the necessary training that goes with it). A Taser has been shown to be nearly 100% effective in disabling a targeted human being for up to several minutes after a single zap. There is at least one case, however, where a very angry man fired weapons at police and nearby civilians continuously after taking 42 separate bullet "hits" to his body. Only the 43rd bullet, which severed his upper spine, stopped his murderous rampage.

So...should you then search for the invader and pound him? 

NO! Searching for the invader in your home, no matter HOW many stripes you have on your black belt, is statistically stupid - tactically, medically, and legally*.

Instead, with the cell phone in one hand, you can provide your address to the 911 Operator, and then maintain a running recording of what you see - perhaps including the only court-defensible and relevant description of any intruder that you might encounter. Your objective, however, is to GET SAFELY TO YOUR FRONT DOOR, UNLOCK IT FOR THE POLICE, AND THEN GET OUT.

Leave it to the professionals to deal with the intruder. They can use a Taser (far more disabling than a handgun), pepper spray (which you would only inflict yourself with if you tried to use it without some training and experience), and a handgun if absolutely necessary.


Then stand back and watch. Use your phone to video the "perp walk" for your family, friends, and local TV station.

* Stupid Tactically, Medically, and Legally:

1. First, compare yourself against a trained pair of completely awake and alert police officers. You lose majorly in this comparison.
2. Second, a significant number of people, inexperienced in using a handgun in complex circumstances, end up injuring themselves with their own weapon (3rd degree burns, lacerated hands, even self-inflicted gunshot wounds). 
3. Third, unless you have an unlimited bank account, you should expect that ANY use of a firearm on another human being - for whatever reason - will require hiring an attorney, and then months of court appearances and thinking about them.

~~~~~

Geomorphology Part II

Giant Worms, Gold in Your Hair, and Another Way Volcanoes Kill

The fun and utility of geomorphology comes from what land-forms can tell us that we didn’t already know. An example of this is the Palouse region of western Washington State. Here we find hills composed of excellent sandy soil, covered with wheat farms that seem to have no “normal” (that is dendritic, or branching) water drainage pattern. When you drive through them on highway 91 it just looks “wrong” to a geologist. The area hosts the giant Palouse earthworm or Washington giant earthworm (Driloleirus americanus). There are other odd bits of unusual evidence lying around: boulder “erratics” – huge chunks of granite located in unexpected places. My wife reports seeing a huge granite boulder on the south side of a 1,100-ft-high (340 meter) ridge above the Columbia River, many kilometers from the source location of identical rock where they must have come from. How did these huge, Volkswagen-bus-sized erratics get moved such a long distance and then lifted over that ridge? 

Only in the 20th Century did geologists gain access to aircraft and air photos, and noticed that the Palouse “hills” looked like giant versions of ripple marks that you can see in stream beds. Further mapping and thinking led to the then-astounding conclusion that the Palouse represented the remnants of a gigantic series of floods in relatively recent history. The scale of the first flood was beyond anything that anyone could even begin to comprehend. As recently as 12,000 years ago, an ice dam formed at Glacial Lake Missoula (Idaho-Montana). Because it was an ice dam, when it failed it had burst catastrophically. How catastrophically? The modern Willamette Valley where you find Portland, Oregon, was once a canyon terrain, but is now filled in almost flat with sediment. How much sediment? Calculations suggest that the first of what may have been up to 72 sequential Missoula Floods carried with it 5,000 cubic kilometers of rock and debris down the Columbia River gorge! This was moved all the way from northeastern Washington State, down the Columbia River to Astoria, Oregon, and roared on out into the Pacific Ocean where it can now be mapped using sonar bathymetry systems.

As one practical example of applied geomorphology, there is a gold deposit in southern Venezuela called Chiricayen. The gold was first discovered when a Pemon Indian women went to a waterfall to bathe. When she returned, her family noticed bright flecks of gold in her black hair. The crucial issue was where did this gold come FROM?  If it all just collected at Chiricayen, then this was just a minor gold deposit. If it came from a particular sedimentary unit, then it implied a far larger and concentrated gold resource. The waterfall location was in a shelf along a cliff of a 1.7 billion-year-old sedimentary unit called the Roraima Formation. A quick glance at the site from a helicopter suggested that the shelf was a down-dropped block, fault-controlled. However, a geomorphologist gave careful examination to stereo-pair photos that I took from a helicopter passing in front of this apparent shelf. He pointed out that the sedimentary layer pattern (thick units and thin units in a particular order) was the same on the front of the shelf as in the neighboring walls. In other words, this was not a down-dropped catchment, Instead, one single layer above it probably hosted paleo-rivers, ancient meandering streams, that hosted gold from a distant (now weathered-away) source. This single observation provided a new understanding of how these gold deposits form – and redirected the gold exploration strategy of our host agency in Venezuela.

Geomorphology can have life-saving consequences, too. The rate of sediment movement off of a stratocone volcano like Mount Rainier has a huge impact on downstream communities. Recognizing lahar (water-and-volcanic debris flow) deposits from time past can indicate what can be expected in time future. In 1906 a “Pineapple Express” (a so-called “sky river” from the tropical western Pacific) dumped 20 inches or more of rainfall on the volcano, precipitating a huge flood. This led to the mistaken conclusion that making the Nisqually River straighter would make things safer. However, vast amounts of dredging and river-straightening made no one safer, but instead meant less salmon and more sediment movement to southern Puget Sound. 

It’s worse than that. Shortly after the 1980 eruption of Mount St Helens, a farmer asked Rocky Crandall, a USGS volcanologist, about boulders in his soil that made plowing difficult. Crandall recognized that this was part of a lahar – and traced the source to the BACK side of Mount Rainier. A lahar called the Osceola roared down here 5,600 years ago – and continued on to Puget sound, and then continued another 30 kilometers further under the water. Another lahar called the Electron did the same thing just 500 years ago. Lahar is an Indonesian word, and refers to a fast-moving wall of wet volcanic debris, not unlike wet concrete in consistency, traveling up to 60 kilometers per hour with car-sized boulders entrained in it. Nothing can stop these things, and one lahar killed 23,000 people in Colombia in 1982. It ripped the cathedral from its foundations and killed most of the occupants of Armero in just a few minutes. Ultimately, the recognition dawned on everyone that a crumbling volcano will unavoidably succumb to gravity. Geophysical studies have shown that the near (west) side of Mount Rainier is hydrothermally altered – old-time miners call rocks like this “punky” - and it is just waiting to collapse. The Nisqually Valley is literally filled flat with layer after layer of lahar deposits – up to 22 of them. Sadly, it is now also covered with human development - up to 500,000 people are exposed to the utter inevitability of a lahar. 

A careful geomorphology study has thus shown that the altered and friable west side of Mount Rainier will someday inevitably fall off and flow all the way to Puget Sound – sweeping absolutely everything before it. To try to protect the human population, the USGS has installed Acoustic Flow Monitors in the drainages leading down from the volcano. An artificial intelligence system monitors multiple sets of paired geophones 24 hours a day, and the entire system is coupled to a siren warning system. Calculations suggest that the inhabitants will have 45 to 60 minutes warning of an oncoming lahar from Mount Rainier, once a moderate-sized earthquake (or another eruption) sets free what is now called the Sunset Amphitheater. 

~~~~~