Saturday, September 21, 2013

With Every Fault is there an Earthquake?



There is nothing that motivates a teacher more than a student asking a serious question. Perhaps even more motivating is a teacher who, like us, wants to help a student.

Q: My 5th grade student Awva asked this question. Thank you!

With Every Fault is there an Earthquake? 
 
Amy S

A: With every fault there once WAS an earthquake - that's how the rock broke in the first place. With most faults there are many earthquakes over time. 

However, there are active faults (like the San Andreas) and there are (usually temporarily) inactive faults. There are faults mapped in Precambrian times that seem to have been completely inactive for over 600 million years. However, a fault implies a zone of crustal weakness, and it is not at all uncommon for a fault to be re-mobilized by a later tectonic event. Because the zone is already weakened, it is preferentially broken open again - it's easier to break than the surrounding rocks. I've seen billion-year-old faults in Venezuela that were activated again as the Atlantic Ocean opened between North America and Europe the first time, again when it closed, and then yet again when the modern Atlantic reopened a second time.

To make things even more complicated, however, there are faults that have "slow creep" or "silent" earthquakes. The continental-margin subduction fault off the coast of Vancouver Island, Washington State, and Oregon is an example. These events are movements of the underlying oceanic slab beneath the continental crust. Shallower parts of these faults are fluid-lubricated, and motion on them is so slow that they are detected only by continuously-recording GPS instruments - they are not felt. These silent earthquakes can take more than a day to move two fault surfaces past each other, instead of the milliseconds that would be involved for a more typical earthquake. Thus the strain is slowwwwwly released. However, the strain on rocks deeper down is then increased - this is where the surfaces are locked because fluids no longer reach them.

When one of these huge subduction earthquakes happen, it's a real attention-getter. The Great Tohoku earthquake of 2011 was a magnitude 9 event. It caused massive damage, and the Fukushima nuclear reactor complex is just getting more dangerous every week as time goes on since the original tsunami destroyed the cooling system.

~~~~~

Sunday, September 15, 2013

Rock Ages and Distance from Volcanoes



Sometimes questions come in that start off all “rong” – It then becomes necessary for us to set the record straight. Sincere efforts, however, should not go unrewarded. The following question is an example.



Q: My name is Mike W.  I'm a HS science teacher from Sibley, Iowa.  I'm really trying to change my labs and activities so they're more realistic and I have an idea.

I've attached an image of the graph I want the kids to eventually make (image showed  a fake graph with dots lining up along a diagonal line, one axis being distance and the other age attached), where they plot age of igneous rock v. distance from active volcano.  Obviously, it's a generally linear graph, which is great b/c they then can get an equation.  But, I want to do more than just give them a data table.  

I was thinking...

Could I give them rock "samples" (15-20) and each rock sample that they get has an identification number and distance from volcano.  In addition, each "sample" would have a visual representation of the remaining radioactive/non-radioactive isotopes left....getting them to use radiometric decay to date their rock sample.  For instance, sample B would be 2.5 km from the volcano and it would have 5 blue circles (parent) and 35 green circles (daughter).  They could deduce that three half lives have passed and the rock is x number of millions of years old.

I want this to be accurate, yet doable in the classroom.  Do you have any thoughts or suggestions???

Thanks so much!!! 

--Mike W

A: Ummmm. I think you have started from a false premise - that rocks are older the farther away they are from a volcano. I can think of many exceptions to this "rule" -- in fact so many exceptions that it cannot possibly be true.

However, I certainly understand and support your hope to get your students actively involved. Real-world, really cool examples are what drew me from a physics PhD track to geosciences. 

I hope soon to have a publication on the web that your students might be able to make use of. It concerns magnetic and gravity and radiometric data acquired around Newberry Volcano in Oregon (Newberry's footprint is at least 20 times larger than Mount St Helens). If you look at the potassium-40 image, you can see that the youngest eruption (~1,300 YBP) blew tephra out eastward at least 100 km: there is a bright red trail extending from Paulina Peak to the east. Careful sampling has shown that it thins the farther east you go. The problem: there are older tephra underlying this blast that didn't reach as far away. 

Volcanoes are generally messy creatures - many of them may look like neat stratocones, but when you unravel their history with careful mapping, you find that they have very complex histories, and the symmetric cone represents just the latest eruptive episode with the first stage of erosion still well underway. Dating volcanic flows and tephra layers is very difficult - when I served a 5-year tour as chief scientist for volcano hazards, I poured must of the extra funding I could lay hands on into telemetry bottlenecks and a dating laboratory. When you want to know how dangerous a volcano is, you need to see how many times - and how violently - it has erupted over time. It turns out that radioelement dating is extraordinarily complex. First, you have different half-lives that limit their applicability: C14, argon-argon, rubidium-strontium, potassium-argon, uranium-lead, etc. Then you have different availabilities - there is almost no potassium or uranium in mafic and ultramafic rocks, for instance. You must painstakingly look and hope for sphene crystals to do any argon dating. THEN you have the contamination issue: when you are doing argon-argon dating, you have to deal with the fact that the air has more argon in it than the grains you are trying to date. The statistics involved are mind-boggling, but it turns out that where you sample is the most critical part affecting success. 

Where you CAN get a good age-vs-distance curve is with an ocean-floor spreading center. The symmetric magnetic banding across the Mid-Atlantic Ridge is a case in point. Deep-sea drilling has sampled the different magnetic-polarity bands and has verified that the farther you move away from the spreading center the (a) older the rocks are, and (b) the deeper the rocks are. You can get a VERY clear age-vs-distance pattern there, because this is a relatively steady-state, on-going process. 
~~~~~