Thread by @celestelabedz

Hello friends, today I will be using my expertise as a seismologist to tell you how to make an earthquake early warning system out of cats.

Thread:

Myths about animals predicting earthquakes have been around for a long time. Scientific research shows NO evidence that anyone (human or animal) can reliably predict earthquakes BEFORE they happen. Once a quake has already started, though, animals have an edge.

The edge is that many animals can sense more subtle motions than humans. The first waves to arrive after a nearby quake (P-waves) show much weaker shaking than later waves (S-waves & surface waves). In many cases, P-waves are barely noticeable to humans, but tangible to animals.

This effect is where a lot of animal “predictions” come from. A pet feels P-waves and reacts, then a few seconds later, the human feels the S-waves and/or surface waves and thinks their furry friend was psychic.

Modern earthquake early warning systems look for P-waves, as well. Seismometers detect initial waves emanating from a quake that has already begun, a central computer recognizes the pattern, and alarms sound for people further away who will soon feel shaking.

So, if we have earthquake early warning systems that look for P-waves and cities full of pets that can feel them, let’s combine the two in a thoroughly ridiculous proposition.

First, we need the perfect subject. Cats and dogs are both widely-popular pets that are known to react to small earthquake waves, but cats (on average) spend a larger portion of their day sedentary, reducing the frequency of false paw-sitives.

Second, we need to monitor cat activity. Since many cats’ reaction to quakes is to freak out and bolt for cover, an accelerometer is key. Human activity trackers with accelerometers already exist, so Fitbits become Kittbits when affixed to a collar.

Third, we need to collect this crowdsourced (or rather: meowdsourced) data. Fitbits connect via Bluetooth, so cat owners can sync theirs with internet-enabled devices (laptop, phone, etc) that stream data to a central system like in traditional earthquake early warning.

Fourth, we need a great name. I propose the Pet-based Urban Rapid Response to Shaking: PURRS.

PURRS works like this: An earthquake occurs. Signals of cats going bananas start pouring in from near the epicenter, propagating outward with the seismic waves. When enough activity in the right distribution is detected, the system cat-egorizes it as a quake and sounds alarms.

Now let’s run some stats for Los Angeles. Wikipedia says that the Los Angeles-Long Beach-Anaheim Metropolitan Statistical Area contains 13,310,447 people. Census Reporter says that Los Angeles averages 2.8 people per household, so 4,753,731 households.

The American Veterinary Medical Association says that 25.4% of U.S. households own cats, with an average of 1.8 cats per cat-owning household. Therefore, we have a total of 2,173,406 cats to work with. (I’m not counting feral cats here; it would be harder to Fitbit them.)

Density matters more than numbers here, though. Wikipedia says that the LA-LB-Anaheim MSA covers 4,850.3 square miles, giving us an average of 448 cats per square mile.

The ShakeAlert web site says that LA’s current early warning system uses the CA Integrated Seismic Network. I searched its network code on the Incorporated Research Institutions for Seismology map tool, and found an average density of 0.059 per square mile in the LA area.

So, with over seven THOUSAND times the station density as the current system, data from PURRS could be way more detailed, even despite those false paw-sitives from cats randomly getting the zoomies.

Cost-wise, basic Fitbits are like $70, so outfitting every cat in the LA area would cost over $150 million. Unfortunately, this is significantly more expensive than the estimated $38.2 million that the LA Times says a west-coast-wide traditional early warning system will cost.

PURRS is a hard sell, not gonna lie. Upkeep would also be expensive and difficult; even just trying to cat-alog that much continuously-streaming data would be a monumental task. We would also have to turn it off during thunderstorms and on the 4th of July.

So, is PURRS a *practical* solution for avoiding earthquake cat-astrophes? Absolutely not.

Would it *technically* work? Purr-haps, and that’s what I’m here for.

Would my time have been better spent doing my *real* seismology research? Don’t tell me what to do with my weekend.

I'm glad so many people find my plan a-mew-sing! The PURRS seis-meow-meters might not be ready in time for the next quake, but YOU sure can be. Further your earthquake preparedness edu-cat-ion at earthquakecountry.org.

And if you're in California, you don't need to wait for PURRS to have a few seconds of pre-meow-nition; you can already get a whisker of warning to help you stay safe!

Los Angeles: earthquake.lacity.org/shakealertla

State-wide: myshake.berkeley.edu