sponsored links
TED2014

Gavin Schmidt: The emergent patterns of climate change

March 12, 2014

You can't understand climate change in pieces, says climate scientist Gavin Schmidt. It's the whole, or it's nothing. In this illuminating talk, he explains how he studies the big picture of climate change with mesmerizing models that illustrate the endlessly complex interactions of small-scale environmental events.

Gavin Schmidt - Climate scientist
What goes into a climate model? Gavin Schmidt looks at how we use past and present data to model potential futures. Full bio

sponsored links
Double-click the English subtitles below to play the video.
We live in a very complex environment:
00:12
complexity and dynamism
00:15
and patterns of evidence
00:16
from satellite photographs, from videos.
00:18
You can even see it outside your window.
00:21
It's endlessly complex, but somehow familiar,
00:24
but the patterns kind of repeat,
00:28
but they never repeat exactly.
00:30
It's a huge challenge to understand.
00:33
The patterns that you see
00:37
are there at all of the different scales,
00:39
but you can't chop it into one little bit and say,
00:43
"Oh, well let me just make a smaller climate."
00:46
I can't use the normal products of reductionism
00:48
to get a smaller and smaller thing that I can study
00:52
in a laboratory and say, "Oh,
00:55
now that's something I now understand."
00:57
It's the whole or it's nothing.
01:00
The different scales that give you
01:03
these kinds of patterns
01:06
range over an enormous range of magnitude,
01:08
roughly 14 orders of magnitude,
01:11
from the small microscopic particles
01:14
that seed clouds
01:16
to the size of the planet itself,
01:19
from 10 to the minus six
01:21
to 10 to the eight,
01:22
14 orders of spatial magnitude.
01:24
In time, from milliseconds to millennia,
01:26
again around 14 orders of magnitude.
01:29
What does that mean?
01:32
Okay, well if you think about how
01:34
you can calculate these things,
01:36
you can take what you can see,
01:38
okay, I'm going to chop it up
01:40
into lots of little boxes,
01:41
and that's the result of physics, right?
01:43
And if I think about a weather model,
01:45
that spans about five orders of magnitude,
01:47
from the planet to a few kilometers,
01:49
and the time scale
01:52
from a few minutes to 10 days, maybe a month.
01:54
We're interested in more than that.
01:58
We're interested in the climate.
02:00
That's years, that's millennia,
02:01
and we need to go to even smaller scales.
02:03
The stuff that we can't resolve,
02:06
the sub-scale processes,
02:07
we need to approximate in some way.
02:09
That is a huge challenge.
02:11
Climate models in the 1990s
02:13
took an even smaller chunk of that,
02:15
only about three orders of magnitude.
02:17
Climate models in the 2010s,
02:19
kind of what we're working with now,
02:21
four orders of magnitude.
02:23
We have 14 to go,
02:26
and we're increasing our capability
02:28
of simulating those at about
02:31
one extra order of magnitude every decade.
02:32
One extra order of magnitude in space
02:36
is 10,000 times more calculations.
02:38
And we keep adding more things,
02:41
more questions to these different models.
02:43
So what does a climate model look like?
02:46
This is an old climate model, admittedly,
02:49
a punch card, a single line of Fortran code.
02:51
We no longer use punch cards.
02:55
We do still use Fortran.
02:57
New-fangled ideas like C
02:59
really haven't had a big impact
03:01
on the climate modeling community.
03:04
But how do we go about doing it?
03:07
How do we go from that complexity that you saw
03:08
to a line of code?
03:13
We do it one piece at a time.
03:15
This is a picture of sea ice
03:17
taken flying over the Arctic.
03:19
We can look at all of the different equations
03:21
that go into making the ice grow
03:23
or melt or change shape.
03:26
We can look at the fluxes.
03:28
We can look at the rate at which
03:29
snow turns to ice, and we can code that.
03:31
We can encapsulate that in code.
03:34
These models are around
03:36
a million lines of code at this point,
03:38
and growing by tens of thousands of lines of code
03:40
every year.
03:43
So you can look at that piece,
03:44
but you can look at the other pieces too.
03:46
What happens when you have clouds?
03:48
What happens when clouds form,
03:50
when they dissipate, when they rain out?
03:52
That's another piece.
03:54
What happens when we have radiation
03:56
coming from the sun, going through the atmosphere,
03:58
being absorbed and reflected?
04:00
We can code each of those
very small pieces as well.
04:02
There are other pieces:
04:06
the winds changing the ocean currents.
04:07
We can talk about the role of vegetation
04:11
in transporting water from the soils
04:15
back into the atmosphere.
04:17
And each of these different elements
04:19
we can encapsulate and put into a system.
04:22
Each of those pieces ends up adding to the whole.
04:25
And you get something like this.
04:31
You get a beautiful representation
04:33
of what's going on in the climate system,
04:36
where each and every one of those
04:38
emergent patterns that you can see,
04:42
the swirls in the Southern Ocean,
04:45
the tropical cyclone in the Gulf of Mexico,
04:47
and there's two more that are going to pop up
04:49
in the Pacific at any point now,
04:51
those rivers of atmospheric water,
04:53
all of those are emergent properties
04:56
that come from the interactions
04:59
of all of those small-scale processes I mentioned.
05:01
There's no code that says,
05:04
"Do a wiggle in the Southern Ocean."
05:06
There's no code that says, "Have two
05:08
tropical cyclones that spin around each other."
05:11
All of those things are emergent properties.
05:14
This is all very good. This is all great.
05:18
But what we really want to know
05:20
is what happens to these emergent properties
05:21
when we kick the system?
05:23
When something changes, what
happens to those properties?
05:25
And there's lots of different ways to kick the system.
05:28
There are wobbles in the Earth's orbit
05:31
over hundreds of thousands of years
05:33
that change the climate.
05:35
There are changes in the solar cycles,
05:37
every 11 years and longer, that change the climate.
05:39
Big volcanoes go off and change the climate.
05:42
Changes in biomass burning, in smoke,
05:46
in aerosol particles, all of those things
05:49
change the climate.
05:51
The ozone hole changed the climate.
05:53
Deforestation changes the climate
05:57
by changing the surface properties
05:59
and how water is evaporated
06:01
and moved around in the system.
06:03
Contrails change the climate
06:06
by creating clouds where there were none before,
06:08
and of course greenhouse gases change the system.
06:11
Each of these different kicks
06:15
provides us with a target
06:18
to evaluate whether we understand
06:20
something about this system.
06:23
So we can go to look at
06:25
what model skill is.
06:28
Now I use the word "skill" advisedly:
06:31
Models are not right or wrong; they're always wrong.
06:33
They're always approximations.
06:35
The question you have to ask
06:37
is whether a model tells you more information
06:39
than you would have had otherwise.
06:42
If it does, it's skillful.
06:44
This is the impact of the ozone hole
06:47
on sea level pressure, so
low pressure, high pressures,
06:49
around the southern oceans, around Antarctica.
06:52
This is observed data.
06:55
This is modeled data.
06:57
There's a good match
06:59
because we understand the physics
07:00
that controls the temperatures in the stratosphere
07:02
and what that does to the winds
07:06
around the southern oceans.
07:07
We can look at other examples.
07:10
The eruption of Mount Pinatubo in 1991
07:11
put an enormous amount of aerosols, small particles,
07:14
into the stratosphere.
07:17
That changed the radiation
balance of the whole planet.
07:18
There was less energy coming
in than there was before,
07:21
so that cooled the planet,
07:24
and those red lines and those green lines,
07:26
those are the differences between what we expected
07:28
and what actually happened.
07:30
The models are skillful,
07:32
not just in the global mean,
07:34
but also in the regional patterns.
07:36
I could go through a dozen more examples:
07:39
the skill associated with solar cycles,
07:41
changing the ozone in the stratosphere;
07:44
the skill associated with orbital changes
07:46
over 6,000 years.
07:49
We can look at that too, and the models are skillful.
07:51
The models are skillful in response to the ice sheets
07:53
20,000 years ago.
07:56
The models are skillful
07:58
when it comes to the 20th-century trends
07:59
over the decades.
08:02
Models are successful at modeling
08:04
lake outbursts into the North Atlantic
08:06
8,000 years ago.
08:09
And we can get a good match to the data.
08:11
Each of these different targets,
08:15
each of these different evaluations,
08:17
leads us to add more scope
08:19
to these models,
08:22
and leads us to more and more
08:23
complex situations that we can ask
08:26
more and more interesting questions,
08:30
like, how does dust from the Sahara,
08:32
that you can see in the orange,
08:35
interact with tropical cyclones in the Atlantic?
08:37
How do organic aerosols from biomass burning,
08:40
which you can see in the red dots,
08:44
intersect with clouds and rainfall patterns?
08:46
How does pollution, which you can see
08:49
in the white wisps of sulfate pollution in Europe,
08:51
how does that affect the
temperatures at the surface
08:55
and the sunlight that you get at the surface?
08:58
We can look at this across the world.
09:02
We can look at the pollution from China.
09:05
We can look at the impacts of storms
09:09
on sea salt particles in the atmosphere.
09:12
We can see the combination
09:16
of all of these different things
09:18
happening all at once,
09:21
and we can ask much more interesting questions.
09:22
How do air pollution and climate coexist?
09:24
Can we change things
09:29
that affect air pollution and
climate at the same time?
09:31
The answer is yes.
09:33
So this is a history of the 20th century.
09:36
The first one is the model.
09:39
The weather is a little bit different
09:41
to what actually happened.
09:42
The second one are the observations.
09:44
And we're going through the 1930s.
09:46
There's variability, there are things going on,
09:48
but it's all kind of in the noise.
09:51
As you get towards the 1970s,
09:53
things are going to start to change.
09:56
They're going to start to look more similar,
09:58
and by the time you get to the 2000s,
10:00
you're already seeing the
patterns of global warming,
10:02
both in the observations and in the model.
10:05
We know what happened over the 20th century.
10:08
Right? We know that it's gotten warmer.
10:10
We know where it's gotten warmer.
10:12
And if you ask the models why did that happen,
10:13
and you say, okay, well, yes,
10:16
basically it's because of the carbon dioxide
10:18
we put into the atmosphere.
10:20
We have a very good match
10:22
up until the present day.
10:24
But there's one key reason why we look at models,
10:26
and that's because of this phrase here.
10:30
Because if we had observations of the future,
10:32
we obviously would trust them more than models,
10:34
But unfortunately,
10:38
observations of the future
are not available at this time.
10:40
So when we go out into the
future, there's a difference.
10:45
The future is unknown, the future is uncertain,
10:48
and there are choices.
10:51
Here are the choices that we have.
10:53
We can do some work to mitigate
10:55
the emissions of carbon dioxide into the atmosphere.
10:57
That's the top one.
11:00
We can do more work
11:02
to really bring it down
11:04
so that by the end of the century,
11:06
it's not much more than there is now.
11:08
Or we can just leave it to fate
11:10
and continue on
11:14
with a business-as-usual type of attitude.
11:16
The differences between these choices
11:19
can't be answered by looking at models.
11:23
There's a great phrase
11:28
that Sherwood Rowland,
11:29
who won the Nobel Prize for the chemistry
11:31
that led to ozone depletion,
11:35
when he was accepting his Nobel Prize,
11:37
he asked this question:
11:40
"What is the use of having developed a science
11:41
well enough to make predictions if, in the end,
11:43
all we're willing to do is stand around
11:47
and wait for them to come true?"
11:49
The models are skillful,
11:52
but what we do with the
information from those models
11:55
is totally up to you.
11:58
Thank you.
12:00
(Applause)
12:02

sponsored links

Gavin Schmidt - Climate scientist
What goes into a climate model? Gavin Schmidt looks at how we use past and present data to model potential futures.

Why you should listen
Gavin Schmidt is a climate scientist at Columbia University's Earth Institute and is Deputy Chief at the NASA Goddard Institute for Space Studies. He works on understanding past, present and future climate change, using ever-more refined models and data sets to explore how the planet's climate behaves over time.
 
Schmidt is also deeply committed to communicating science to the general public. As a contributing editor at RealClimate.org, he helps make sure general readers have access to the basics of climate science, and works to bring the newest data and models into the public discussion around one of the most pressing issues of our time. He has worked with the American Museum of Natural History and the New York Academy of Sciences on education and public outreach, and he is the author of Climate Change: Picturing the Science, with Josh Wolfe.
sponsored links

If you need translations, you can install "Google Translate" extension into your Chrome Browser.
Furthermore, you can change playback rate by installing "Video Speed Controller" extension.

Data provided by TED.

This website is owned and operated by Tokyo English Network.
The developer's blog is here.