EPS20 Set 5 Risk Prepared Readiness

Hazard

Likelihood of an earthquake occurring in an area

Exposure

People, buildings, infrastructure, and economic activity in areas that could shake

Vulnerability

How susceptible exposed people and assets are to harm

Risk

Expected consequences; Risk = Hazard x Exposure x Vulnerability

Probability vs. prediction

A statistical likelihood, not a forecast of when and where a specific quake strikes

USGS Bay Area forecast

72% chance of at least one M6.7+ earthquake in the Bay Region, 2014-2043

Major Bay Area faults

San Andreas, San Gregorio, Hayward, Rodgers Creek, Calaveras, Concord-Green Valley, Greenville

US average annualized earthquake loss (AEL)

~$14.7 billion; California ~$9.6 billion

AEL meaning

Long-term average of expected building-related losses per year, not actual yearly losses

Seven Steps to Earthquake Safety (in order)

1 Identify hazards, 2 Make a plan, 3 Prepare kits, 4 Identify building weaknesses, 5 Protect yourself during shaking, 6 Check injuries and damage, 7 Follow your plan

Water to store

1 gallon per person per day, for 3-5 days

Safe action during shaking

Drop, Cover, Hold On (not the doorway)

Myth: quake injuries are all from collapsing buildings

False; 55% of Northridge injuries came from falling objects

Myth: good building codes mean safe buildings

False; codes don't retroactively fix older buildings

Prediction vs. forecasting

Prediction specifies when, where, and how big (can't be done reliably); forecasting gives probabilities over time windows

What must a valid prediction specify?

Time interval, location, magnitude range, probability of chance occurrence

Iben Browning prediction

1990 prediction of a New Madrid earthquake around Dec 3 based on planetary alignments (179-year cycle); not scientifically supported; nothing happened; caused panic

Why Browning's idea failed

No theoretical basis; 1982 alignment and earlier ones produced nothing; he claimed unsupported success with Loma Prieta

Candidate precursors (none reliable)

P-wave velocity changes, ground uplift/tilt (Palmdale Bulge), radon, resistivity changes, swarms, animal behavior

Haicheng 1975

Successful prediction; evacuation; few deaths

Kwangtung 1976

False alarm; people in tents for ~2 months

Tangshan 1976

Missed alarm; no precursors; >250,000 deaths

Earthquake early warning (EEW)

Detects the first P-waves and sends alerts before damaging S and surface waves arrive; not prediction

Why EEW works

P-waves are fastest but less damaging than the waves that follow

EEW limitations

Little or no warning near the fault; data latency; speed vs. certainty tradeoff (bigger quakes take longer to confirm)

EEW uses

Personal safety, automated control, situation awareness

BART and ShakeAlert

In use since 2012; if shaking exceeds MMI IV, trains slow from ~70 to ~27 mph; up to ~50 sec of warning

Japan EEW example

Shinkansen: 24 trains stopped without derailment during Tohoku; chip manufacturer cut losses from $15M to $200K

Foreshocks and aftershocks

Smaller quakes before and after a mainshock; aftershock rate decays by Omori's Law

SBA and FEMA aid

SBA loans up to $40,000 personal property and $200,000 real estate; FEMA grants average $5,000-$15,000, max combined $41,000

Cumulative effect of earthquakes on infrastructure

Cascading failure (e.g., Kobe), where infrastructure failures compound