Why are earthquakes so hard to predict? - Jean-Baptiste P. Koehl
- 721,341 Views
- 18,302 Questions Answered
- TEDEd Animation
Rocks in the crust generally tend to break when deformed and form large cracks called “faults”, while rocks in the mantle typically deform in a plastic manner forming shear zones. Three types of deformation exist: elastic (e.g., pulling a rubber band, extending it, and watching it come back to its original shape once the pulling stops), plastic (e.g., the Play-Doh experiment: once we stop applying forces to the dough, it keeps the shape it was given), and brittle (e.g., pulling too hard on a rubber band and break it). Earthquakes are typically the result of brittle deformation and are generated during movement of the tectonic plates along faults. Earthquakes correspond to a release of energy accumulated over a certain period. While some faults produce earthquakes, others move aseismically (without producing earthquakes), releasing accumulated energy progressively because they are “lubricated” by, e.g., fluids circulation along the fault or partial melting of adjacent rocks during fault-displacement. The Earth would perhaps be a more peaceful place if all faults were aseismic.
Rocks are composed of different minerals, each having a different pressure–temperature field of stability. For example, granite is mostly composed of silica (SiO2), most of which forms quartz, the most common mineral in granite and the most stable mineral at Earth’s surface temperature and pressure conditions. However, granite is not only composed of quartz and may contain some other minerals like hornblende, biotite, potassium feldspar, and plagioclase. Among these, hornblende is probably the one least stable at surface conditions, then biotite and plagioclase. Thus, during friction and heating of a granite along, for example the San Andreas fault, a small fraction of the hornblende minerals may melt together with a very small portion of biotite and plagioclase, and represent the “lubricating” fluid described. However, once crystallized, the fluid generally does not show any biotite, hornblende and plagioclase, as the fluid components reorganize to form minerals more stable at near-surface conditions.
One way geologists are trying to tackle the earthquake-prediction challenge is to closely track the behavior of very active faults which move several times a week, so we can identify any patterns and build up a picture of what’s to come. Highly active faults like the San Andreas in California have given us lots of helpful material: A combination of historical records, modern-day satellite monitoring, and geological and geophysical mapping has helped scientists estimate that it causes one major earthquake every 100 years, on average. Knowing this could help cities prepare their inhabitants for a possible threat. Depending on how deep and far away an earthquake occurs, smartphone sensors could offer from a few seconds to 1–2 minutes of advanced warning (e.g., NASA’s Quakesim software; more information here). This technique is already in use in places like Japan and California where earthquakes occur frequently (see the smartphone apps developed by the University of California, Berkeley, here).
To learn more about the the educator's organizations and what they do, check out their websites here:
UiT The Arctic University of Norway in Tromsø: uit.no/startsida
Research Centre for Arctic Petroleum Exploration: www.arcex.no/
Create and share a new lesson based on this one.
More from Awesome Nature
lesson duration 04:20