i'm going to give you a an overview of

what it is that that we do as a as a

group and what we do collectively as

scientists to try and understand what

we're dealing with with with climate and

the climate system and climate change

the world is extremely complex and

dynamic the patterns that you can see

here in these clouds which is not real

where this is a computer simulations is

not this is this isn't a this is in a on

a laptop not on a non a satellite but

the patterns that you can see a very

similar to the ones that you see in the

satellite you can see storms you can see

convection you can see waves going

through you can see things repeating but

not quite right

that's the essence of the system that

were talking about it's a it's a

complicated chaotic dynamic system it

never quite repeats and yet all of the

processes that are going on the

underlying all of the things that are

going on with these clouds there

understandable there are computable we

can go and measure all of those

different individual things how do we

deal with that the challenge is is one

of scale the scales of the climate

system range from micrometers for small

aerosol particles that are from the

nuclei of clouds in the atmosphere all

the way out to the size of the planet

itself which is about 10 to the 8th

meters that's about 14 orders of

magnitude that's an enormous scale range

in which to encapsulate all the things

that are going on and even in time you

know you're going from things that are

happening on micro second scales to

things that are happening over millennia

again 14 orders of magnitude an enormous

range of scales to comprehend them play

everything together that's enormous

challenge what you have when you build a

model then is-is-is you have some

fundamental concerns we don't have

infinite computer power right so how

does that work well with weather models

right the ones that produce the forecast

that you see every day

the kind that the kind of result physics

the things that we can expressly

included in these models go from the

global scale down to about the

50-kilometer scale right that's only a

very small part of that huge range of

scales that that actually impact of

climate and the weather right the stuff

that we can resolve the stuff that we

can say okay we totally understand that

we can put in the exact equations is

just that small part all the rest of it

is what we call these subscale processes

now that stuff is is how a cloud

actually forms how rainfall actually

comes from a cloud system how

evaporation works at other very very

small scale and all of those things need

to be approximated in some way we need

to have some empirical phenomenological

approximation to that physics to allow

the rest of it to work and that's a real

challenge right there's this is this is

turbulence this is a small-scale

heterogeneity this is small mountains

and small hills and small lakes all of

those things have to be squeezed into

our approximation of those subscale

processes so there's no guarantee that

this works right there's no guarantee

that this is even a program that will

lead to successful predictions and so

the test all of this program is whether

it does actually produce successful

predictions

now you might not like the weather

forecast that you get you might think

that they're terrible but actually

they're much much better than they used

to be where the models are skillful of

producing forecasts three four five six

days almost 10 days out in ways that

were unforeseen able even 20 years ago

now climate models are slightly

different beast climate models have to

use longer timescales because time is a

longer time scale and so that compromise

they make is they don't go to quite such

fine detail

and so back in the nineteen nineties

climate models really only occupied that

very small part of that of that space

now we can go a little bit further so in

the 20,000 tens that so the models that

were working with now you know we're

coming out about the same spatial

resolution as the weather models and

we're taking up more times you know

we're going out for a thousand years and

working on the physics at the same

potential and yet we still have this

large large range of subjects scale

processes that we still have to include

this is a climate model I this is an old

tram or you'll be pleased to know we

don't use punch cards anymore it's a

single line of Fortran we still do you

use for transfer any of the old-timers

here who think that that their skills

are no longer valid

they asked a valid yes you can have a

job with each of these colored bands

with a single subroutine a single

calculation of some physical aspect of

the code arm and you would put them into

a machine one at a time and if you got

them in the wrong order you have to

start over again it would take months to

produce any output arm and it was a very

very clunky thing and so obviously if

you have to write out every line using

punch cards you don't include or of

comments so you know we inherited this

code and we're going

what does this go do we have no idea

anyway so it doesn't quite look like

that anymore but the the essence the the

idea of what we're doing is directly

related to what we did back in the

nineteen eighties and in the missus

before that 1970 ok so how do we

actually go around building a model we

do it one piece of the time and i use

the a jigsaw analogy by the beginning

and I'll kind of work with that a little

bit now it's very much like putting

together pieces of a jigsaw so this is a

piece of the jigsaw it's a picture of

arctic sea ice taken from a flight going

across the North Pole each of those ice

floes that you can barely see there are

a few tens of kilometers across and you

can see leads

the middle i was the summertime so the

ice is breaking up

and so there's a lot of physics is going

on there right there is the physics of

the Sun coming in reflecting off the ice

because it's white

there's the physics of the of the the

heat being absorbed in the ice melting

it from the top and is melting from the

bottom going on at the same time there's

dynamics how each of the ice floes

interacts with each other all of those

things can be encapsulated measured and

put into some kind of formula so here

are some of the formulas

yeah so physics you know I don't expect

you two to look at these your critique

these are but this is basically what we

do we write down what the boundary

conditions are we use calculus we use

basic conservation of energy that's

that's a big deal and for each of those

different things we then take some code

and you get lots and lots of lines of

code there are some comments now that

that's pleasing arm mr. the last bit

that there's some melting and we make

that one piece into like a component and

so okay well when we do see ice that's

the piece of code that we're going to

use and we can add in lots of different

processes so that was the sea-ice

process now there's one other processes

associated with clouds there's a process

associated with radiation and the solar

radiation coming through the atmosphere

and being absorbed at the different

layers of the atmosphere that's one more

piece there are other pieces the winds

and the way use and other pieces the

flow of water through plants the flow of

water and rivers back down to the ocean

so the point is that each of those bits

makes up part of the picture of the

climate system now I've left and pieces

out because our picture of the timing

system is not complete right we don't

know everything about the system and yet

when we put it all together we end up

with a simulation

that has all of the emergent properties

that we can see in the real world right

so these storm systems in the Southern

Ocean you can anybody has been to the

Southern Ocean knows exactly what they

look like or storms in the North

Atlantic Kevin can tell you about those

tropical cyclones in in the Pacific the

convective bands in in the was a couple

of tropical cyclones are going to cycle

around each other kind of meat

yeah now within the code there was no

code that says hey do an atmospheric

River that hits Oregon there's no code

that says do tropical cyclones and make

them dance around each other right

all of these things are emergent

properties that just arise from the fact

that the system that you're modeling is

chaotic is dynamic and all of the

different processes that we put together

intersect and interact in ways that you

can't predict a priori now just having

done that i think is a intellectual

endeavor worthy of of of Cato

and-and-and Chomsky right trying to

understand exactly what's going on in

such a complex system and being

successful at doing so is an enormous

challenge and i think one we can rightly

be proud of but we're not just doing it

to understand the system we're doing it

because that system can be kicked in

many many different ways and right now

humans are kicking this system in a very

particular way so let's look at some

ways in which the system can be kept

right so there are waffles in the

Earth's orbit these bubbles kind of make

a difference on ten thousand two hundred

thousand year time scales and a big

driver of the of the of the Ice Age

cycle that we've been seeing for the

past 2.5 million years right so Iceland

used to have a lot more ice

it's not quite as I see

as it used to be and and and it will get

messy as well with 10,000 years ago

20,000 years ago

Iceland was almost completely emaciated

right all right yeah and that's due to

these bubbles so that's that's a very

powerful kick to the system right that's

everything changed

you know can we understand that ok there

are changes because of the Sun the Sun

has cycles that has activity and it's

over its lifetime of the lifetime of the

earth has expanded and become more

active by about thirty percent over the

last four billion years that has an

impact on the climate the set the solar

cycles an 11-year cycle have an impact

on the climate we need to understand

those things volcanoes as have also well

aware have an impact on the climate not

just locally but they can have a global

impact on the climate by the injection

of sulfates into the stratosphere

changing the albedo of the palette

changing the amount of energy coming

into the system right so you're getting

the idea there are lots of different

ways in which we can change the climate

right or the system itself can be

changed whether its biomass burning

ozone depletion land-use change

contrails or of course greenhouse gases

all of these change the balances and the

flows of energy in the system and when

you change the balances and flows of

energy in the system you change the

climate so what we need to understand is

whether the simulations that i discussed

before where the weather models are the

climate models whether they have skill I

want to be very clear about this

these models are not complete right

there were holes in the jigsaw there's

whole bunches of things that we have to

and empirically fill-in write the word

that we can't predict from a priori

considerations

so why do we think that these things are

useful

we think that they're useful because

they have

demonstrated skill skill is I model

results skillful if it's better than

what you would have had otherwise

that's not to say it's perfect it's not

to say that there are uncertainties but

it's something that's better than what

you had before with the weather forecast

you could stick your finger in the air

and predict what the weather is going to

be tomorrow based on just persistence of

the weather today that probably doesn't

work very well here but it works quite

well in New York but we can do better

than that with models right weather

models a skillful despite the fact that

there was that whole range of things

that we had to empirically include and

so we need to demonstrate that climate

models are skillful at estimating the

changes to the climate when you kick

that system in all of the different

myriad ways that that system can be

kicked

okay so is that true is there are the

models skillful so well they know

volcanoes are a current topic of

interest in in Iceland I said well

actually not not anymore right we're

done with the volcanoes right so that's

it for the next week right

remember no more volcanoes ok

all models skillful with respect to big

volcanic forces so the large big volcano

that we had that had a global impact was

magnitude but went off in june nineteen

ninety one arm and this is a graph of

how much stuff basically there was in

the stratosphere are from the main

eruption and then it took like three or

four years for all that stuff to kind of

fall out of the atmosphere

ok so this is actually quite a large

amount of stuff right on the amount of

stuff that that is there that's the

technical term stuff and got some people

pay attention okay

the amount of stuff that was there cause

the radiation to be from the Sun to be

significantly less than it was right

because each of those little particles

is white and so when the Sun shines on a

white particle it reflects back to space

and that reduces the total amount of

energy coming in the amount of energy

going out was pretty much the same and

so you have less energy coming into the

system right so basic conservation of

energy suggest that that would lead to a

cooling of the planet but by how much

how much should the planet call with a 3

watt per meter squared energy imbalance

over a few years you have to do the

calculation and the climate models allow

you to do that calculation so what do

they suggest well what they suggested

are the the little thin lines that you

can see there's a bit of noise because

of the different weather and each of

those models and then the red line there

is what actually happened

so the models were able to estimate

quite precisely the rate at which this

the climate called because of this

forcing and we know that is better than

what we could just worked out by doing a

calculation on the back of the envelope

and in fact the first time that we did

that calculation we did it in October

1991 and made this prediction before any

of those changes had actually occurred

the model was skillful and it wasn't

just fearful estimating the global

cooling which you know kind of my energy

balance arguments you might argue was

the easy part but it was also skillful

at predicting that in the wintertime you

got a strange pattern in Europe where

you actually get winter warming after a

big volcano and that turns out to be a

dynamic response changes in the winds

that are affected by the temperature

changes our loft that actually push more

warm water over more warm air over over

Europe during the wintertime and that is

also captured by the model and the

models are skillful even with the

dynamics and the radiation and the

temperature there are lots of other

example

solar cycles you know we can measure and

calculate the change in ozone in the

stratosphere because of this change in

the sun with models are skillful because

of that in overall changes those wobbles

in the Earth's orbit like even 6,000

years ago they were relatively important

and we can look at that difference and

the models are skillful estimating how

the northern hemisphere temperatures got

better because our change because of

that ok

the response to the ice sheets 20,000

years ago the models are successful are

getting that right

the 20th century multi-decade all trends

were successful getting that right

models are skillful even bizarre things

that have happened to the climate about

8,000 years ago there was a huge great

lake that covered most of Manitoba and

Ontario much larger than all of the

great lakes that exists right now and it

was kept in place by the remnant by the

remnant ice sheet that was kind of

standing over hudson bay about 8,000

years ago that I sheet finally broke and

all of that water that was kept in that

Lake rushed into hudson bay and into the

labrador sea and into the North Atlantic

around here and at that same time there

is evidence of a cooling event that

happened in the Greenland ice cause in

Europe and in Newfoundland and all

around here and I'm sure that they would

a nice and have been affected as well

the models are skillful at responding to

that water going in by changing the

circulation in the in the ocean and by

changing the temperatures in the

rainfall patterns associated with that

the models are skillful across a whole

range of different ways that we can kick

that system

ok

and that system is complicated and

global right this is a simulation again

of tiny little atmospheric particles and

all the different kinds of particles

that you can have these orange swirls

are dust coming from the Sahara Desert

and you can see them affecting the

climate and affecting a deposition of

stuff all the way across the Atlantic

the white wispy parts over over Europe

you can see that's atmospheric pollution

sulfates right and you can see that

what's happens in Europe does not stay

in Europe the green and the Reds

indicate where there is our where there

are fires and whether its biomass

burning and where is organic carbon and

black carbon being put into the

atmosphere and again these things don't

just stay where they are stuck there

they started you can see the pollution

over China that again does not just stay

in China the blue dots associated with

the big tropical storms are associated

with sea salt particles so as the winds

whip up the waves small particles of sea

salt which are a key element in the

production of low-level clouds you can

see them again in the in the Southern

Ocean all of these things are connected

and important and we can answer now with

the click with the with the computer

models that we have we can answer

questions that associate that are

associated with all of those things

how does pollution change climate and

air quality all at the same time I love

this but this this animation that they

they did a really nice job here this is

my colleagues at NASA Goddard Space

Flight Center and now there's another

volcano in Madagascar that just went off

and so you can include all thoughts of

things in these models and answer all

sorts of actually quite subtle and

interesting nuance questions

how things work over the 20th century

when we've seen an amount of global

warming right so is this is a model

simulation this is the observations you

can see the patterns of whether you know

they're not correlated the weather isn't

predictable 50 years in advance but the

emerging patterns have changed that you

see towards the end of the 20th century

and that we end up where we are today

those patterns are predictable they are

predicted by the models the models are

skillful at getting these patterns right

this warming in the Arctic compared to

the Antarctic the warming over land

compared to the ocean the warming in the

North versus the woman in the south

these are skillful models but then let's

think about the real reason why we need

models right and then this this comes

from a rebuttal that the two scientists

wrote when somebody you know they talked

about modeling of of future tropical

cyclones and somebody wrote a response

to them they say well why are you

looking observations and they responded

i think quite finally what if we had

observations of the future we obviously

would trust them more than models but

unfortunately observations of the future

and not available at this time and

that's funny but it's actually the real

reason why we build models we want to

have predictive capability of situations

where we don't yet have any observations

so that we can make plans accordingly

whether it's a weather forecast whether

it's a climate forecasts whether it's an

economic projection which we could get

into is as well but Sir what we want our

is an ability to make informed decisions

about the future right so predictable

lity and predicting the future is the

absolute fundamental reason why we're

looking at models right that's what

science is all about making predictions

testing your theories adapting them

making them better making them more

predictive so what does the future hold

what we have

in the future of course our choices and

we don't have to fall down one

particular line the economic an economy

technology how we structure societies

these things are not written in stone

we have choices society has choices you

individuals have choices and these

aren't you know completely accurate

labels we don't know really what will

happen if we just continue as we go

along this is an estimate of that we

have a pretty good idea of what it would

take to bring everything kind of back

down to the level that it was you know

at the beginning of this century this

aggressive mitigation i am kevin will

tell you this is not going to happen

this is the under two-degree world that

people talk about sometimes business as

usual

that's a 56 degree world that's a

totally different planet

remember how different the ice age was

to today

well that is just as different but in

the opposite direction the ice age was a

different planet different organization

of ecosystems where people lived where

things live where things grew the same

will be true for the business-as-usual

planet where were more likely to end up

in my opinion is some effort serious

mitigation not today not tomorrow maybe

in the next 10 years 20 years 30 is

armed and now I'll still be a planet

that has significantly warmer

temperatures particularly on land

particularly in the North particularly

in the Arctic this will be a radically

different planet as well but exactly how

different is still to be determined but

the choices are really ours to make

now I'm a scientist and and I spend my

time working on those small scale

processes and the encapsulation of all

of that physics that you store in those

animations

and these projections these predictions

put us in a very odd position and it was

very clearly encapsulated by showed

Rowland who are was the one of the the

chemists who discovered the reactions

that led to ozone depletion and some of

you here will remember spray cans and

freon and the Montreal Protocol and how

that was negotiated and how that changed

so sure Roland understood quite clearly

the consequences of having made

scientific discoveries that have

real-world implications and he said in

an interview what's the use of having

developed a science well enough to make

predictions if in the end all willing to

do is stand around and wait for them to

come true and that's really the dilemma

that people like me fais do we just

continue to make predictions or do we go

out and tell people about what these

predictions mean do we go and help

people work out what they should do

about the energy system i'm not an

expert in energy systems you know should

we use nuclear power should be used

geothermal power should we do this

should we do that

those are decisions that are not up to

me though I would like to think that the

information that we produce informs the

decisions that you make individually and

that you make as a city or as a country

and as the world and I don't know what

our future will hold arm but i'm i'm

hopeful that whatever future we choose

it's an informed future so thank you

very much