Thank you.

In putting this together clipped off the
other Natural Hazards I'm going to talk

a little bit about other natural hazards
in the framework that I lay out here

really would apply to floods earthquakes
hurricanes tornadoes the whole gamut.

But I thought would be easier to work
through the risk analysis example for

one and
only allude to other natural hazards.

Occasionally.

I thought I'd try and get your attention
first as to why we might care about

earthquakes this is actually a building
in California May second of eighty three

I remember because it was exactly a year
before my son was born and a couple of

a student I went over took pictures about
twenty four hours after this happened.

You can actually see.

This is a unreinforced masonry
building that are more of

our architecture colleagues were here but
none reinforced masonry building is when

you have brick walls where they
are actually providing the support and

the bricks don't do well in an earthquake
they all move in different directions and

you get what I call this
doll house effect here.

If you have a brick like
we have in the southeast.

Usually it's just a veneer it's
usually wood frame house and

just kind of got brick stacked up for
looks but

this is actually masonry walls that
are actually holding up the roof and

much of the structure and
apologized to be in black and white.

I couldn't find the original so

I copied them back out of some articles
that are pretty old at this point.

Here's an interesting single
family house from the twenty's.

You'll notice that it's
not quite plumb anymore.

But what's happened is it's shifted
off its foundation by about a foot and

it was not bolted it was required.

Now under code to the foundation so.

The whole house is shifted about a foot
but that's actually really tough on water

connections natural gas connections and
you don't sort of go back and

just push the House back into place and
live happily ever after.

So these are both are from Coalinga
this is actually kind of a signature

picture from the.

Well I'm afraid earthquake
that took place.

I think was the second game of the World
Series in eighty nine something called

the Cypress Structure a double deck
freeway and this is these are probably

close to three feet concrete pillars
this is the top level of the freeway.

Here's I took about such eight students
there who got in about thirty six hours

after the earthquake and
spent three days there.

In San Francisco area
after the earthquake.

So here's maybe the luckiest
guy in San Francisco that day.

He's on the lower level of the Cypress
Structure and you can see the rebar and

the like coming loose here.

And if he'd been about
two hundred yards ahead.

He would have been here.

And this is not a waffle maker.

But it is but the physics are similar to
a waffle maker and sixty seven people died

in the collapse of the Cypress Structure
which is over just across the Bay Bridge

on the Oakland side needs to kind of
a lot of other mess around as well.

Now this is more recent this is
actually a residential building and.

Dodging yang in sets one province from
the May twelfth earthquake of this here.

And you can reinforce ministry doesn't
do so well I was there in late July.

Here's actually another
residential building and

you can see it's kind of large pieces of
it has gone down and you can actually see

behind the fence big piles of masonry
rubble here some temples or another.

It is another residence structure and you
see sort of the Dollhouse effect where you

can see into the structure and this is
about two months after the earthquake

there's been some clean up
obviously a lot of people displaced

a lot of people living in tents and this
was not right at the epicenter of this

was probably a good forty miles from
the epicenter of the earthquake.

Here's a temple up in
the mountains a little bit all but

closer to the earthquake this is
not my poor photography skills but

you see the posts here are a little bit
leaning This is a wood frame structure so

the thing holds together pretty well but
in fact it's kind of

over at about a five degree
angle probably not salvageable

just down from that you actually see sort
of a whole bunch of damage there some

historic sites and some old temples and
the like up here and

you see a bunch of it has come down and
a little further down the hill.

You can see quite a bit of this whole
temple complex has been pretty wiped

out and then one this one I did not take
this is after than the squalor earthquake.

This is something called non structural
damage which I have a five year project

we're better a year into it where you look
at the damage not to the structure itself

but to things like drops the wings like
these pendant lights here like this.

H.V. a C. system sort of
the contents of the building and

even if the building itself stands
up you see a lot of times the.

The sort of guts of the building
don't do so well and the building.

If this happens is.

Safe.

And the structural engineers
acquired a success but

in fact it's not very
functional as a hospital as a.

As an office building as a school or
whatever.

So this nonstructural damage is kind of
a new frontier in terms of research that.

We're working on.

So.

Risk versus hazard just sort of
a definitional thing to start with

hazardous some kind of a drill extreme
geophysical bent like an earthquake.

Hydrologic event a forehead meteorological
event tornado or hurricane and we

distinguish that from risk risk is kind of
what we care about more that's the threat

to human life safety property or social
systems caused by that hazard of that.

So it's kind of like the tree falls in.

In.

The past and the ones who go to
hear it does it make a sound.

Well if a hazard happens in the middle
of the Mojave Desert if the earthquake

happens a little Mojave Desert.

There's nothing there in
terms of people roads.

Buildings.

Do we really care to have
a saddle seismologist care but

we as planners as social
scientists as the general

public really don't care it's just
kind of a scientific curiosity.

We care when it hits the built environment
and that's why people in architecture

civil engineering city planning
care because it creates a risk to

human life safety and property and I would
argue to a larger social systems as well.

So we talk about risk we talk about
when things we care about are possibly

damaged or damaged or.

Are potentially damaged.

So what's risk assessment is that analytic
process that combines ground shaking.

I'm going to talk about seismic
risk assessment we could

talk about a similar process as I said for
floods or hurricanes.

Sometimes we even use the same
software to deal with them but

it's an analytic process that
combines ground shaking and

other geological facts for when the
earthquake advance with characteristics of

the bitten Grom up at the estimate
physical economic and

social consequences that
are going to accrue from that so.

Why would we want to do that.

Why would we.
I mean it's an interesting kind

of intellectual problem but
why would we want to assess risk.

First we might.

If we understand the risk we might be able
to as we say designed mitigation measures

we might be able to think about
how to build differently.

How to go back and patch things that we've
already built to make them more secure or

more likely to sustain whatever
kind of event that would happen so

both new construction and
retrofit of existing construction.

We took we call that mitigation
measures you could even have other

kind of mitigation measures
like flood insurance where

you don't really save the building but
you kind of put some financial.

Safeguards in place so if the building
gets washed away you have a way to

rebuild but mitigation measures so
we want to know what the risk is so

we can really think
about how to respond to.

Plan emergency response as a plan or

I'd like to think about what we
can we do to divert the risk.

Let's don't build in the floodplain
let's don't build on the barrier island.

If we build in a hazardous
area let's build

a building strong enough that can risk.

Withstand a likely earthquake and there's
a whole other branch of the hazards

business and there's a little bit of
competition sibling rivalry and that's

emergency response which is really all
about how do you pick it up afterwards.

How do I clean up.

How do I actually take care of
once Galveston has been hit.

How do I get the people out and shelter
and how do I provide them with food

water medical care and
maybe financial resources.

So that's called emergency response and
and the two are somewhat.

Separate but risk analysis really supports
but so you really need to understand

the risk before you can do that
figure out how to address it or

figure out how to pick up after
risk assessment the best risk

assessment SAP where today belongs
to the insurance companies.

It's proprietary you can't get it.

I've seen some tests and some fly offs but

the best stuff belongs to the insurance
companies because they use it to set

premiums to figure out how much to try
to do for flood insurance or earthquake

insurance or actually with hurricane they
just double it every year so they don't

really have to do the risk assessment
is just going to take it up every year.

But insurance companies Penn.

Friends and others of the masters of
the universe who are doing such a good job

on Wall Street also own large
portfolios buildings Metlife

owns a lot of buildings the Ohio teachers'
retirement plan owns a lot of of

of real estate office buildings and
the like and

they often want to do an analysis of their
portfolio to find out if we have a seven.

In L.A. What does that do to the Ohio
teachers' retirement system.

What does that do to metlife funding
inability to pay off on life insurance so

oftentimes you see this badly
wrong owners of real estate

trying to assess Well we're a little
heavy on real estate in the L.A. basin.

Maybe we should buy some at risk for
a hurricane or Chicago or somewhere else.

That's not subject to so that they're
not overexposed to a particular kind of

risk and the other thing risk analysis is
very useful to create public awareness so

people understand that they're
living in earthquake country or

hurricane country and can get a better
sense of what that risk is like suppose of

some reasons we might want to engage
in some kind of risk analysis.

So here's a little a graphic
that sort of goes through.

Say from from source to site.

We've got some kind of earthquake that
happens at a place not a two dimensional

place it's usually at depth twenty
thirty forty kilometers down often and

that that slippage of a fan or
other kind of event creates.

Waves that then travel over
a path to you at your house.

And so we'd like to know you know
what's the probability what's

the likelihood of an earthquake
of a particular magnitude and

I guess we had a three two here a couple
years ago and they called my friend Barry.

Good night and he said.

In feel what.

I remember and California you always
have the little Foros that kind of

keep you away or
kind of get your attention to earthquakes.

And I've done a lot of work at MIT America
they don't have those little reminders.

They just are going to get the eight zero
at one point without all the pap being

of those the smaller of that but
what's the probability of an earthquake of

a specific magnitude you might think
of them as a probability distribution

as to how likely ones of
different sizes are then.

We talk about attenuation How does
that add like Luckily that energy

from the earthquake does not say constant
distance sort of like a sound if I yell

hear that sound kind of degrades over
space the same thing with the energy for

the earthquake that actually degrades
over space the story that I still

don't believe but everybody says it's true
that the great middle American earthquake.

Eight hundred eleven eight hundred twelve
reversed before the Mississippi River.

OK I'm willing to believe that but
supposedly it's at the south

east corner of Missouri rang
the church bells in Boston.

Now I know the attenuation
at MIT America is much

more spread out than it is in California
but it's about twelve hundred miles from

almost Memphis and
I am not sure I believe that that event

although it was a seventy nine rang
the church bells in Boston but

ten your way should is how fast does
that energy die out over space.

How is the ground going to fail the big
hazard with earthquakes is not the fault

opening up all of you saw Charlton Heston
in the movie earthquake and

you see these big fissures coming up.

That's really not the biggest.

You can get that in if you're the one poor
fool that has this house built on top

of fault that's it's bad but there
are a lot of other kinds of ground failure

landslide liquefaction We'll talk
about those in a bit but how.

Was the ground going to fail as
well as shake and then we need

to know once we have that ground shaking
at a location and it could be different at

different locations along that attenuation
How is the structure going to buy.

How is the building.

How is this lovely reinforced
concrete building going to behave

hasn't been reinforced masonry building
going to behave like the ones we saw at

the beginning has a wood frame structure
going to behave how are things like

highway overpasses and bridges going to
behave and then finally has that going

to affect social and economic systems
if we take out the bridges in St Louis.

What does that do for commerce.

If we take out the it's winter and

we take out the gas pipelines that go
from the Gulf up to the northeast.

What does that do to heating oil as
we're experiencing that which is

due to gas supplies.

What are the social economic consequences
of those physical damage and

I actually have been arguing the last
few years to extend the notion of risk

assessment to not just include physical
damage which is kind of the here.

Which is where my structural
engineering friends usually go but

to really push at the next level and

think about how does it affect
social systems how does that affect.

The functioning of the economy and

various social aspects that
we'll talk about later.

OK so
here's a overall view of risk analysis.

And the first thing is let's kind
of understand the hazard and

this is probabilistic in
kind of a lot of fun.

You know how likely are earthquakes of
different sizes how does a ten year.

You know we don't have what's called a
recurrence curve that tells us how likely

is it to come back kind of like one
hundred year flood is a five hundred year

earthquake.

What's the problem with stick shaking
maps we can actually make a map that

shows us how likely are we to get certain
levels of shaking over a large area.

We probably want to do some.

Geologic investigations map secondary
hazards like Lancelot it's got a lot of

that in such a long earthquake that
the ground motion there was so

strong it caused a number of earthquakes
parts of the Bay Area actually

had this thing called liquefaction where
you shake the ground it becomes kind of

a Cold War jelly and doesn't support
foundation so the buildings actually took

over got a lot of that on
the margins of San Francisco Bay But

that's sort of the you know
the the world of geologist and

seismologists and they hopefully can
summarize that into a probabilistic hazard

map then the next step is
actually a little trickier.

We want to understand how different
kinds of structures behave and

different types of structures have
different response to a given level

of ground motion as I said
unreinforced masonry doesn't do very

well reinforced concrete it's
not the greatest thing to be and

would frames pretty good mobile home
that's so good so different types

types of structures behave differently
we want to understand how they behave.

Then we want to kind of where the planners
come in and map existing land use

an individual structure types and
estimate and map the damage to those

existing uses And actually if
we were really planners we might

want to forecast future land uses and
look at what the damage would be there.

Think about if we head
through a cast what.

Well we don't have the technology in
one thousand nine hundred but we could

a forecast what the damage might look like
if we had had come into Galveston have

a better idea of what we're going to
get rather than just experience it and

then hopefully all that drives some
kind of policy decisions land use owning

building codes investment in
infrastructure investment in

emergency response building shelters
putting away supplies like water food

medicine that we can actually access
building a better Fema that was.

Bonds a little better at cetera.

OK Well interesting enough Lucky for
me that all of this really.

All of the features the buildings
the hazard the soil types all

of the stuff is distributed geographically
so it's kind of hard to do this

on the back of a three by five card you
really need to have a system that lets you

capture those on a framework where you can
actually deal with risk assessment I guess

I wrote my first article looking kowtowed
G. I asked as a risk assessment platform

in eighty four and I reckon we the field
that's come around to this notion and

all the risk assessment software and
how it's built on top of G.I.S. but

sort of fusing G.I.S. it's the natural
platform for doing risk assessment.

But you on the one or
two building scale and you can do this for

one building the architecture building and
it's risky but typically as a planner I'd

like to look at a city or a metropolitan
region or some other large area and

G.I.S. is the tool of choice to do that
and that's pretty widely recognized now.

So here's just a kind of a look at here
some essential facilities and expected

ground motion essential facilities or
police fire schools hospitals in the new

Madrigal area of America a lovely place
it's about ten percent of seven states

it's everybody's backyard so it's nobody's
front and center thing to think about you

know it's a little bit of Tennessee a
little bit of Kentucky the down state part

of Illinois Chicago's appear it's kind
of not anybody's primary responsibility.

Step A lot of people live there.

It's got Memphis and saying well listen
that all the gas lines go to the northeast

from the Gulf through here a lot of
interstate traffic goes east west through

there and it's had it's the site
of the biggest earthquake in U.S.

history in eighteen eleven eighteen
twelve and actually we're just winding.

Ten years of work on looking
at hazards in this area.

One thing we did was build
a risk analysis software tool.

That is useful in other places but

we actually focused on this and kind of
the next generation of web based G.I.S.

software that's really a step ahead of any
of the other products that are out there.

We would argue that might be true.

Key components of risk assessment.

We talked about this and that flow chart
is just sort of summarizes that you'd like

to have some geologic information and
if you're in Wisconsin

probably earthquakes
are not your first concern.

You've got to have a good
inventory a property at risk and

we've become fairly expert at collecting
building data sets a property at risk

with the kinds of information that we need
to have to actual model actually modeling

the damage that's the broad ability of
those those buildings and structures.

Basically the structural engineers
spend their careers doing this.

We'll talk about and see a little bit
about how they do that and then we kind of

tied to the world economists social
scientists planners emergency

response teams where the social and
economic consequences of that because of

Bam's jots don't tell me that there's two
thousand buildings going to be damaged.

How many people are going to need shelter
how many people are going to die.

How many casualties are we going
to have give me some social and

economic metrics and I argue that that's
important because the decision makers

city council state officials
those kinds of metrics they

can understand probabilistic
distributions of of spectral acceleration

are probably not their forte but
if you can tell them how many people

are likely to be without homes how
many jobs are likely to be lost.

What's it going to cost in
terms of tax revenue those

kinds of consequences are more salient as
the political scientists would say and.

More meaningful to officials local and
state and regional.

So let's look at each one of those kind of
briefly what do we know about the geology.

We'd like to maybe have a probabilistic
distribution of likely earthquakes you

know how how likely are we to have
a seven zero in San Francisco L.A.

Taiwan on sites on St Louis wherever.

Or another way to do.

That's that.
Simple or if you could have size and

location of scenario let's just
assume we're going to have a seventy.

In Memphis and now we don't want to
assume something that's completely.

Unlikely but basically we can rather
than sort of look across the whole

distribution we can choose
a particular earthquake and

understand its consequences so
we maybe choose a six to a seven

when we see what each of those
will do to a particular location.

We got to understand the ten you
asian the sort of degrading of the of

the energy from source to cite
how does that energy decline and

sort of decrease over space usually
put those together and we can have

a ground shaking map so we know different
places different locations what's

the level of ground motion we're going to
have and then secondary hazards landslides

liquefaction couple of others that are
really induced hazards Spiers very common.

You know you break the world over lines
you break the gas lines mean what's likely

to happen that somebody
is going to drop a match.

Now I don't have the water
to fight the fire.

So happen second Cisco in ninety six.

Landslide what compaction fire and
maybe chemical spills if you have

some nice kind of not so nice chemical
sitting around and you shake it up and

break the containers you can have
a lot of chemical spills as well.

Those would be really induced hazards
addition to the secondary hazards.

So those are the kinds of things you'd
like to know in sort of the geology domain

and here is the ground motion map.

These are expressed in decimals.

G. of a the force of gravity.

So we go from about a point three G.
to point eight G.

This is actually Shelby County
Tennessee Memphis it's really

about here and this is the Mississippi
River and we've actually assumed a seven.

And we put it in a particular place and
we put it over this way.

So the the the attenuation is
happening from west to east so

the the if we had assumed it was here
that would go a different way but

the attenuation is a function
the level of ground shaking is

a function both of how far and
where the the earthquake is located but

also some characteristics of the soil and
we've superimposed it here on

census tracks just because the software
uses that for the inventory.

This is a kind of more this is our.

May visit them which actually looks at.

An earthquake off in this direction and
kind of spreads it and

you can see the soil types coming into
play here where it's actually softer

looser soils up these creek beds
in the actual getting kind of war.

More ground motion stronger ground
motion in those particular areas and

of course there are lots of different
attenuation relationships and

sort of different ways to estimate this
and the science is not all in the can.

Here's a ground motion map.

This is actually using not G.'s but a kind
of older Modified Mercalli scale from one

knot felt we don't have much of that
up to this blue that strongly felt up

to these darker areas which happen to be
right on the soft soils around the same

Cisco Bay where we get extreme damage and
it's on the Hayward Fault which is over on

this side of San Andreas actually does
this pretty much action comes right down

through here through the Stanford
campus and on down like that but

the Hayward Fault is another fault and
has a fairly high probability of going up.

I think two percent chance and
fifty years.

It goes right to the Berkeley campus.

So any of you who are thinking
about moving to Berkeley.

Pirates like insurance.

Here's just another ground motion
map this is made America this is

two G.'s one point five
in the civil ten area.

So basically ground motion maps over
different scales and of different types.

Here's a liquefaction
map here you see this

kind of tan color our soils that
are likely to liquify interestingly

enough they look they're a little soils
they don't exactly fellow flood plains but

they often are pretty close to flood
plains this is an area there's a fairly

detailed map produced by U.S.G.S.
for an area up around Paducah Kentucky.

But you can see that the bottoms of the
valleys tend to have these liquefy a bull

unconsolidated soils that if
you shake them turn into a very

loose jelly as if you kept your foot
on the wet sand of the beach and

you'll see it turns into just kind of
a soft and not good for foundations.

I don't have a picture of it but I've seen
some of liquefaction in Japan where you

have buildings laid over about forty
five degree angle from a faction.

OK that's so that's what the geologist
can tell us but we don't care if it.

If nobody's there and
it happens out in the woods.

We really don't care and little bit
is that we care because of the bit.

We care about what's on top of it.

That means something to us.

So we'd like to have the buildings by
structure type I talked about unreinforced

masonry and wood frame.

We've got to have an inventory of all
those you know how many by where probably

energy I asked database infrastructure
systems water sewer in particular

I break those a little bit separately
from the transportation network but

but that's another infrastructure system
but when we usually treat a little

bit differently than water and
sewer systems and

then essential facilities we care about
because they are key to the seventy

two hour spots there what you need
to pick up the pieces you know when.

Have an earthquake it be nice to have
your hospitals know where they are and

also know that they are functional for
that post event.

Period.

Police and start fire stations these
are the emergency responders and

schools and treat to a gymnasium are often
used as shelters but other kinds of

shelters sometimes churches other kinds
of facilities are used to shelter so

essential facilities usually are
particular buildings that get a special.

Kind of look and
a special level with tension because they

are needed in the emergency
response process and

we might even have a mitigation strategy
that says let's reinforce the hospitals

the emergency response centers the police
stations of the fire stations so that

we could actually have fire protection
if we had water after the earthquake

rather than having all the doors jammed
on the fire station so we can't get

the fire trucks out to get to the hydrants
that may or may not have water.

So here's a Memphis building infant or a
couple of us who we are and moved to have

worked on this for a number of years we
have a building by building inventory to

an eighty seven thousand building database
where we actually have taken data that

we can get and modeled what the structure
type is there is no list of buildings by

structure type be nice if you could wiggle
your nose and go get that not at that.

So we've actually built software tools
that take data that you can get like

a number of stories use and

then make inferences which we've
calibrated with field work to estimate.

Are they concrete moment resisting
frame concrete shear wall concrete tilt

up a lot of us in warehouses and
sometimes research buildings

wood frame actually has the best for Jodi
Kerr unreinforced masonry is the worst for

Wired or buildings typically
steel frames are the best and

he's light metal frames are kind of
like Butler buildings kind of real.

Gage they don't do so
well reinforced Masonry is better than and

reinforced masonry but
it's a little brittle.

So it doesn't do all that well in
an earthquake so anyway we'd like to have

you know we would like to maybe guess at
what the buildings are or have a detail

building by building inventory noticed
that about ninety three percent

in this would probably not be true in
California but over ninety percent

typically are wood frame because of the
vast number of single family houses and

some other buildings as well or
wood frame but and they do relatively well

although little damage to each of them
can really stack up to a lot of damage.

So we'd like to have structure type we'd
like to maybe know the footprint shape

matters massing if it's if it's
built with too much mass above that

doesn't do well in earthquake
height building type age.

Location all those are sort of
characteristics of the building.

We'd like to know because we want
to measure it's its response.

We like to have of the things more what
I'm here to sort of make that bridge

to social and economic consequences.

What are people doing in the building.

Engineers call it occupancy planners
call it land use and was it a school or

is it a warehouse is it a a T.
shirt shop is it.

A police station.

What's happening in that building is very
important for how the how important it's

going to be and how it's going to play
into the social and economic consequences.

We always carry about care about value.

Both of the structure but
also of the contents.

Contents of a hospital or high tech.

Computer manufacturing might be

a lot higher than the contents of just
a parking garage or something like that.

So they also care about
not just buildings but

here's a water network
these are all they I think.

Six inch and

larger water lines in the Memphis
area also got the pump stations and

tanks were actually belittle software tool
that looked at if you broke parts of it.

What would happen to the overall
response of the system.

We'd like we did have for
that pipe material in

the land we have wood pipes still
which is interesting but concrete cast

starts using canned cast iron pipes
as a lot of that still in Memphis.

It doesn't do as well it's not
as ductile as other materials.

More modern materials concrete cylinder
pipe and others diameter for for

flow length longer is more likely to get
broken than shorter age location and

we also need to have no understand how
those are connected together which pipes

hook to which ones if we're going
to model damage to the system.

Here's just a transportation network looks
like a map that you might have a little

bit of difference though all
the dots are bridges and

we have a string of data about each of the
bridges the bridges are the most fragile

port of the transportation network and
when you take out the bridges you

typically get some problems with
throughput and look at this one's great.

All these little blue that's actually
a red line but there are a lot of bridges

and you think about bridges any
kind of overpass covert All

those are types of bridges that you would
see in the transportation network and

they're the most vulnerable part
you might get some landslide and

that sort of thing but really the bridges
or what's going to most likely fail.

Here is just essential facilities we've
got the hospitals on the left and

schools on the right and we have the same
kind of building information for

them as well as capacity a number
of beds those sorts of things for

the hospital square footage.

For the schools for storage.

All right so let's say we know
something about the geology.

We know something about the the building
stock the infrastructure stop

the transportation network.

Now we have to actually understand how
those things behave when they don't

behave the same.

And so we need something called a ball or
ability function or more often called

a fertility curve that will take that
structure and say how is it going to

what's going to happen to it at different
levels of shaking if I should get one.

J What's going to happen
as opposed to happen.

Versus point two G.'s.

And they're typically based
on laboratory experiments.

We have a lot of civil engineer colleagues
that shake buildings till they break and

kind of instrument them and
take pictures of them and really try and

learn about that analytic models to do
it all in kind of a virtual world or

expert judgment which means I've been
to three earthquakes I've seen some so

I'm going to draw for
Jodie Kerr based on that.

Here is a privilege occur.

This is actually.

For inner story drift
that's one way of measuring

it of one percent that says basically for
each story's going to be one percent

offset from the other that's not too bad
you know story drift of two percent that

would probably be moderate damage or in
a story drift of eight percent each story

is eight percent shifted from the other
ones the buildings that you guys design

usually don't do well
at eight percent shift.

You can take a slice through
this thing if you want and

at any place you could think of that as
the probability distribution of being in

no damage moderate damage or severe damage
and I can take that slice the different

kinds of levels of ground motion and
basically when I cut through that way.

I'm just looking at a probability
distribution being in different

damage States.

When I first started learning about risk
assessment I was kind of taken with all

the probability in it and I guess have
continued to enjoy that over the years

developing fragility curves as I said Rex
expert judgment go to some earthquakes and

look next and call it good.

But I would.

We're testing tons of
money spent by N.S.F.

on building facilities and
doing laboratory tests.

Let's see if I can make
this thing work here.

Here's a test that baffle I have
some people I'm working with.

This reporter size I think you might see
a person in here that's a chair so you can

see the size of this but they actually
do more or less per size structures and

then shake them around N.S.F.
badge it supports the this equipment and

that's why there's no money for
the social science work

right as Andre says this is
very impressive enjoy that.

Here's another one looking at some
nonstructural kinds of stuff a.

Pipe Here's a rubber heater you
can see the water spilling out of

it as it shakes through
the water coming out and

we have a project where we're looking at
piping damage and nonstructural Bam edge.

And the engineers on the project
have all this of Quitman and cost

a fortune to test the stuff up before they
had a photon they broke the equipment.

About six months ago and
that was really kind of a bad thing but

I do different ways of sort of
understanding behavior at different

different levels of ground motion the best
is laboratory testing and it models.

Kind of the new cheaper way to do it.

You've got to calibrate
them against something and

hopefully it's a laboratory testing.

So here we have some essential for

on campus over in
the North Avenue research area.

We have a.

Testing Lab like that where we
actually built a full sized.

To bed a firehouse unreinforced masonry
firehouse typical of what you find in

middle America and the Roberta we shook it
apart and instrumented and took videos to

see how much it took to break it and how
it would break and all that sort of thing.

So that's what's going on right here on
campus as well as those other places but

here are some essential facilities and
we're looking at sort of the probability

of moderate damage to schools and
here the different school sites.

So we're actually applying those
vulnerability curves based

on the structure type based on the ground
motion so we're bringing all those things

together to start talking about what
kind of damage we're going to get Here's

basically building cabinets
by that are moderate damage.

Or at least moderate some moderate or

extensive or complete collapse damage so
moderate or greater and

by census track this is the software is
actually counting the green ones are under

sixty nine the red ones are ten times as
much over three hundred sixty two and

this is the earthquake scenario earthquake
five hundred year return period set up

here probably marked tree and the
earthquake motion coming down this way but

notice it's not just a function
of the number of buildings.

How many buildings you have is of course
important but also what kinds of buildings

you have to estimate where we're going
to have how many buildings of what

kind damaged and this is just buildings
in general but we could do it for

office for housing for any particular
type of building that we might like.

So this is actually a software
program called has us.

I was on the team that helps back that and

it's now used pretty
widely around the country.

You probably saw some of
the has the estimates.

They showed on T.V. for

Galveston as to how much damage was going
to occur in terms of dollars and they

actually showed a pattern of damage in the
low lying areas kind of from Galveston up

to Houston and was basically has a side
put so it doesn't just do quakes.

But as I said I think we need to go
further than give me a count of how

many buildings get broken.

We want to know.

One of the social and economic
consequences just the physical damage

produces consequences like need for
temporary shelter

long term housing displacement there
are still people not back in the world and

Suppose part shortly people who
are not back in Oakland closely.

While income housing tended to be
vulnerable and when it got knocked down.

It wasn't rebuilt so
the low income housing

in Oakland down by the Cypress Structure
is not there anymore.

Direct economic losses
business interruption

one of the masses not just because of cost
so much the pics the building but the but

if a business is not operating for
six months or a year.

What are the losses from that.

How does that affect employment and

one that I particularly like because
I think local officials like it.

How does that affect tax revenue My
experience has always been that while for

officials always care about tax revenue
as the state government at this point.

And if you can tell them that they're
going to lose thirty percent of their

tax revenue for two years.

They tend to get interested in
risk analysis in a hurry so

understanding those consequences I
would argue is key to planning response

recovery and mitigation of the hazards.

So some links between fragility
physical damage and social impacts.

Here's kind of a fertility curve of sorts
where we get hazard intensity more hazard

intensity more damage some kind of a curve
like that we kind of take this physical

damage and lay that over here as
they sort of X. coordinate and

then the more physical damage you get the
more social consequence you get all the at

some point when you get enough damage the
social system is pretty much broken and

you don't get more social damage
because it's not functioning and

so I've been arguing to try and

link social science models of various
kinds to this physical damage to extend.

The notion of risk assessment to
social and economic consequences.

So here's a little graph that.

A lot of folks like.

And it kind of breaks the world up into
hazards earthquakes we talked about

tornadoes hurricanes floods.

I don't know that we can put all hazards
in there but basically there's a different

science involved in each of these
maybe both meteorologists deal

with both of these but the tornado folks
are probably different from the hurricane.

So that's kind of the physical event
the hazard that we talked about.

Here's where we we have we.

We have in the built environment.

Our intellectual interests are about
building stock How do we build it.

What kind of how many stories
transportation systems planners and

engineers in georgia
tech work some with the.

Infrastructure systems planners
certainly the engineers

are involved with this critical facilities
of those special facilities for recovery.

So we really kind of live in this
domain with the engineers and

sort of the bridge between these two are
the vulnerability or fragility curves and

say how that turns into that and

then we care about a lot of social
scientists economists and sociologists or

maybe more expert than we
are on consequences and

I divide them in the short term and
long term housing.

We have a number of things like emergency
shelter in the short term are still people

being emergency shelters and such on
right now but longer term relocation and

displacement people move that.

Move to Houston actually from
New Orleans and got hit by a negative.

Right.
Economic loss price changes direct damage

while longer term fiscal impact
business failures health.

Certainly the short term casualties.

Fatalities but longer term psychological
distress chronic injury and

social disruption where we have the kind
of emergency supplies food water and then.

Kind of long.

Term family stress
neighborhood disruption.

So those are all the consequences and
I skipped over this little guy here that I

call the the lens of social
vulnerability and the lens of social

vulnerability focuses this physical Bamaga
and everybody doesn't get the same share.

Yes What the poor folks get an extra
helping of all these consequences because

the rich folks don't go for
emergency shelter down to the school.

They go take the American Express and
check into the health.

But the folks who don't have those
resources are stuck in the temporary

shelter and sometimes for very long term
same things with all these consequences.

And there's a twenty years
of social science research.

So this.

Lens of social vulnerability tries to
incorporate that historic research to show

that these physical damages
the consequences are not uniformly spread

through the population but accrue to
those with less resources in over.

Overly large way and
the minorities the poor or the old

the children that really take the biggest
hit in terms of these consequences.

It's not the more well off more employed.

Types of folks.

So here's a actually daytime
casualties by census track for Memphis

it's actually gives us an estimate of
kind of casualties Green's one up to over

six by census track looks a lot like the
physical damage but not exactly the same.

Here's where you get six people in
this large census track including some

wiggle fireball area out
of here on Mud Island.

What we named for
a place to live Mud Island

get to get a clue about
liquefaction right and Mud Island.

OK but there's a lot of stuff that Memphis
has put out there on Mud Island but

you actually get a lot of damage
in kind of an reinforced masonry.

So it's not just following the ground
motion it's a function of what kind.

Structures you have and
the density of population as well.

Here's actually a system
we built a software tool.

Based on our G.I.'s and

it actually you can barely see
there's about four or five red.

Water lines that are broken here but
they cut off the water first flow

to all the black lines so actually
what number of undamaged winds close

of conic to Vittie two pipes and
tanks are also cut off and our system

hides that population in this case we say
that ten thousand one hundred eighty.

Thank God we didn't say point four
two households will be without water.

If these parts of
the system are damaged so

we actually take physical damage to
a water network and then say how many

people and we can actually say do you want
to know just people that are low income or

just kids or just seniors we can actually
forecast what types of people are going

to be turned out knocked out of water
supply and we can also look at.

I don't have here employment type so

you can look at particular
parts of the economy.

If you were in Memphis or
soon if you were in Miami.

You might care about the tourist industry.

OK so understanding risk risk
assessment can do a few things for

us it can identify the sources and
types of respect be nice.

Help us protect our Tad's
mitigation help us prior

ties how we're going to respond and
try to fix the built environment or

build it differently in the future and
identify emergency response needs.

It also our plan is to test alternative
plans of mitigation investments to see

which are going to be more effective some
maybe if we have a plan that preserves

the floodplain we'd have a lower level
risk than one that put people in there.

If we were in the Bay area maybe we
were the density along the Hayward fault

although there are a lot of economic
pressures that push a lot of

development towards that.

And so conclusions risk assessments
useful in planning and hazardous error.

As I would argue this is a technique
that's come of age still complicated but

it's actually at there and available now.

Risk assessment is required if
you're going to do a welcome

mitigation plan that's a required by Fema

if you don't have a local mitigation
plan they say they won't give you.

Emergency funds after an event.

That's not always true but that's
certainly been the party line a risk

assessment should be instead extended to
include social economic consequences.

And if you're going to do it.

Build it on top of a G.I.S. framework and

you get a lot of side benefits from
having all that data in there as well.

The future of the future risk assessment
a couple years ago we had a workshop

in Boulder we commissioned six papers and
eighteen reviews and

we have an edited book that really talks
about where does this go from here.

And so if you're really interested
in risk assessment this is a quick

book that really looks at.

Where are we today but more importantly
where is this going to be in five or

ten years because of a future
of risk assessment software

methodology and application.

So questions that.

Social Capital One that a lot
of people have looked at in fact

I just but certainly a lot of kind
of demographic characteristics like

income level family structure.

Minority status political power female
headed households all those have been

shown to be correlated with
with sort of increased helping

the consequences and I just kind of
summarize of Matt that graphic there but

there's a whole literature about
what factors are related to

social vulnerability and
there's been some look at social capital.

I don't know that the Government's
political factors have been looked at as

carefully as like a pretty
it looks like a seven.

Well this is this is expert judgment so
we didn't do a test they looked like

it looks like a seven nine to
me feels like a seven minute.

Yes there were certainly
some large tremors and it's

not as many as you'd like to I
mean we keep praying for more and

more advanced to validate the models but
and

what's interesting about this is the
building construction types in the U.S.

and Japan are very different from most
of the rest of the world even we get.

I think was to seventy five earthquake
in Mexico City a lot of problems with

the construction not enough free bar
rebar to not enough inspection so

not enough rebar all rusty rebar and
so it's a little hard to generalise.

Across places the best validation at this
point probably comes at the individual

building level or into the lab testing so
we kind of know that building is probably

going to behave that way because we shook
it apart over here in North Avenue and

then as we scale that up.

Yeah we miss some stuff actually
I've actually become very I love

dealing with my structural engineering
friends in this arena because I've become

planning you know we worry about a little
bit about well how is our population

forecast going to be and
validation issues but

you know they say hey you know
hundred percent or no problem.

You know if it's only twice as much
as what we said or half as much.

We're good.

So I've gotten a lot more
comfortable with uncertainty

because there is a lot of uncertainty
a lot of uncertainty in the geology and

the seismology there's uncertainty and
have the buildings respond.

There's even more uncertainty about
the social and economic consequences.

But I've gotten very comfortable
with uncertainty and

I used to think the engineers really had
very sharp pencils and they were at their

five and six decimal places when they
tell me that hey hundred percent plus or

minus were good.

Gotten to throw that back out and
when they tell me that the economic

consequence models of the social models
are not very accurate I say Yeah you know.

So it kind of like you're
building models and so

I took me a long time to learn that.

Yes we need to get better.

Yes we need on the ground validation.

We get it a lot more with hurricanes or
floods.

I don't know the tornado world as
well I think there's more of that but

earthquakes are so rare it is hard to get

good validation that we were in some
things from the developing world but but

it says it's a different building stock
and different building technology.

Yeah.

And I don't know I mean
I've read some of it I

don't have an opinion as to whether
the animals can anticipate or not.

And there's also been some some studies
that show that there's some release of

are gone or some other kind of bizarre
gases if you actually sample that

you can actually because they're so

like some movement that happens beforehand
some gases come off if you pick that up

with a sensor you'll know what has been
done in terms of anticipating earthquakes

is one thing that's very interesting is
because you could have an earthquake in

the Mojave Desert and they you'd still
get a lot of energy back in L.A.

you can actually beat the speed of
those waves with a radio wave and

you can actually send a message back to
the gas distribution system that said you

know a five just happened at this lat long
out in the desert fracture the sign back

closed the system off so you basically
rock down the system something's

doing with with act with buildings but
particularly networks of say natural gas.

You're basing close the balance on that
and so by the time the energy gets there.

The system has actually stopped
the flow of of natural gas and

that's just not anticipating it
until Bates the ground motion for

it gets there but it actually is
driven by the actual event and

the fact that the radio waves can
travel faster than the the gram ocean.

City.

Really I would say three places in depth
there are some in the northeast and

there's some hazard some rower level
mapping in South Carolina a big earthquake

happened in Charleston Of course
eighteen eighty five I think.

But new Madras has gotten
a lot of attention

to the Wasatch Front Salt Lake City has
actually been mapped fairly carefully and

all of the West Coast particularly
Puget Sound area was and

is that all of California and

up into the Northwest has got fairly
sophisticated mapping and risk.

Alice's been done for this.

Big big tsunami there some years
back I think in the sixty's

that basically caused some big
shifts around around Anchorage but

less has been done there because there's
not as much property as rich at risk.

So it has not gotten the attention
that we have seen even Salt Lake or

Memphis and a lesser order of
magnitude less in New York actually

New York has some seismic exposure
not just to Wall Street but

but the course of so
much property at risk there.

Dad real damage there would be and

it's not built the building stock is not
built like California or Japan's Yeah

yeah

yeah there's a lot of that actually and
in fact that Memphis went through a whole

big go around that they didn't want

to to admit that they had
an earthquake hazard and and

institute a stronger building code that
the international building code there

in a particular zone we have to build
to to deal with stronger ground motion.

They would not adopt that in Memphis and
Shelby County because they said if we

do that all the development will go
across the river to Arkansas and

so they basically didn't do that
in saying what's the best but

where I used to live county commissioners
when they did a seismic safety element

that is required part of a comprehensive
plan there back in the seventy's probably

there was a thing called the Rinconada
fault and they said well we're going

to declare that they legislated that to
not be a fault to say that it was in.

Active Christmas about four years ago
the Rinconada fault went off and knocked

out downtown Paso Robles California
which is a little town on WANT TO one.

And so yes a lot of people like
to have their head in the sand.

I think if you can show the economic
consequences the problem is the actions

you take today have real and
kind of measurable costs and consequences

now as opposed to the probability of being
safe for some time in the future and I

can't tell you with risk analysis that yes
you're going to get one in sixteen months.

It might be ten years and
so the real sure war costs

accrue now but the benefits
are somewhere out in the future and

uncertain So yeah it's a tough sell.