Climate Change: Demystifying the Application of Earth Systems Models for

Climate Science Richard B. Rood

Cell: 301-526-85722525 Space Research Building (North Campus)

rbrood@umich.eduhttp://clasp.engin.umich.edu/people/rbrood

June 21, 2017

Some Resources

• Gettelman and Rood: Demystifying Climate Models: A User’s Guide to Earth Systems Models

– Springer, Open Source, (It is free.)

• Introductory Material OpenClimate Mini-page

• Model Introduction OpenClimate Mini-page

• Rood’s Class MediaWiki Site– http://climateknowledge.org/classes/index.php/Climate_Change:_The_Move_to_Action

Outline

• Models– Definition

• Models and Scientific Investigation• Models in Climate Science• Establishing Trust: Numerical

Experimentation • Looking Towards the Future• Summary

Assumption

• I am talking to an audience that knows what “model” means in the context of weather and climate science.

• Knows the jargon of meteorology

From: http://www.halfhull.com/main.jpg

What is a Model?

• Model (Dictionary)– A schematic description of a system, theory, or

phenomenon that accounts for its known or inferred properties and may be used for further studies of its characteristics

We live lives full of models

• Models are everywhere in our lives and work– Architecture– Epidemiology– Aerospace– Computer assisted design– Games– The bridge over the Missouri River– Landing things on Mars– Investing my retirement account– How much rent can I afford– My digital thermometer

What is a Model?

• Model (Dictionary)– A schematic description of a system, theory, or

phenomenon that accounts for its known or inferred properties and may be used for further studies of its characteristics

• Weather and Climate– Provide numerical approximations of the equations that

describe the atmosphere, land, ocean, ice, biology of the Earth –

– process definition, diagnostics, predictions, and projections– Solves conservation equations:

– energy, momentum, mass

Models and Scientific Investigation

OBSERVATIONS THEORY

EXPERIMENT

Models and Scientific Investigation

OBSERVATIONS THEORY

PREDICTION

Models and Scientific Investigation

OBSERVATIONS PROCESSES

SIMULATION

Computational Science (Post and Votta, PhysToday, 2005)

• Computational Science & Numerical Simulation– Given what we know, can we predict what will

happen, and evaluate (validate) that what we predicted would happen, happened?

– Validation: Comparison with observations– Philosophy: Do we ever know if we get the right

answer for the right reason?– Computational and natural science: Establish the

credentials of a model to help inform us about the application for which the model was designed.

Models in Climate Science

Models and Model Infrastructure

Models & Model Simulations

Infrastructure

Solves the conservation equations- Mass- Momentum (~ weather)- Energy (~climate)

Split into “processes”- Fluid dynamics- Radiation- Moist physics- Turbulence

Connects it all together. Critical for- Scientific credibility- Collaboration- Development- Efficiency- Analysis- End user

Observations and Models: Processes

Observations Models & Model Simulations

Define & test model “physics”

Infrastructure

PROCESSES DIAG. & TEST

Diagnostic applications

Observations and Models: Weather Forecasts

Observations Models & Model Simulations

Start Forecasts

Infrastructure

INITIAL COND.

Validation

FORECASTS

Prognostic applications

Observations and Models: Assimilation

Observations Models & Model Simulations

Assimilation & Reanalysis

Infrastructure

MELD

Initial ConditionsValidationScientific InvestigationData System Monitoring

Observations and Models: Predictions and Projections

Observations Models & Model Simulations

Assimilation & Reanalysis

Infrastructure

PROCESSES

INITIAL STATE

PREDICTIONS

PROJECTIONS

Diagnostic applications

Prognostic applications

Define & test model “physics”

Start Forecasts Validation

Complexity / Types of Models(Rood, Perspective)

• Conceptual / Heuristic Models– Integrated, theory based (ex. Geostrophic

balance) • Statistical models

– Past behavior and correlated information used to make predictions

• Physical models: First principle tenets of physics (chemistry, biology)– Mechanistic: some aspects prescribed– Comprehensive: coupled interactions, self-

determining

(State Earth will Warm)

(Details of Warming, Feedbacks)

BIOLOGY

CLOUD-WORLD

The Earth System Model: Climate Models

ATMOSPHERE

LAND

ICE(cryosphere)

SUN

BIOLOGY

OCEAN

Establishing Trust: Numerical Experimentation

• Hindcasting• Historical simulation

Let’s look at observations from the last 1000 years

Surface temperature and CO2 data from the past 1000 years. Temperature is a northern hemisphere average. Temperature from several types of measurements are consistent in temporal behavior.

Medieval warm period

“Little ice age”

Temperature starts to follow CO2 as CO2increases beyond approximately 300 ppm, the value seen in the previous graph as the upper range of variability in the past 350,000 years.

Let’s look at just the last 1000 years

Surface temperature and CO2 data from the past 1000 years. Temperature is a northern hemisphere average. Temperature from several types of measurements are consistent in temporal behavior.

Note that on this scale, with more time resolution, that the fluctuations in temperature and the fluctuations in CO2do not match one-to-one.

What is the cause of the temperature variability? Can we identify mechanisms, cause and effect? How?

{

What do we do?

• We develop models based on the conservation of energy and mass and momentum, the fundamental ideas of classical physics. (Budget equations)

• We determine the characteristics of production and loss (forcing) from theory and observations of, for instance, the eruption of a major volcano and the temperature response as measured by the global observing system.

• We simulate the temperature (“Energy”) response.• We evaluate (validate) how well we did, characterize the

quality of the prediction relative to the observations, and determine, sometimes with liberal interpretation, whether or not we can establish cause and effect.

Schematic of a model experiment.

T

T Start model prediction

Model prediction withoutforcing

Model prediction withforcing

Model prediction with forcing and source of internal variability, for example, El Nino, Pacific Decadal Oscillation

Observations

Statistical representation – not deterministic

What do we know from model experiments and evaluation (validation) with observations

• With consideration of solar variability and volcanic activity, the variability in the temperature record prior to 1800 can be approximated.

• After 1800 need to consider the impact of man– Deforestation of North America– Fossil fuel emission– Change from coal to oil economy– Clean Air Act

• Only with consideration of CO2, increase in the greenhouse effect, can the temperature increase of the last 100 years be modeled.

Let’s look at the “modern” record.

• Modern ~ Industrial Revolution~ Last half of 1800s

• When we have direct temperature measures

Figure TS.2320th Century Simulations

Example of Attribution

20th Century Simulations

Meehl et al., J. Climate (2004)

Look towards the future.

• Surface temperature anomaly• Intergovernmental Panel on Climate Change

(IPCC, every ~ 5 years)– IPCC assesses, does not “do” research

• Coupled Model Intercomparison Project (CMIP)– Scientist community designs protocol to evaluate

and establish trustworthiness of climate models • CMIP is not the same as IPCC, but are often

conflated.

IPCC (2007) projections for the next 100 years.

Summary: Models

• Basic scientific principle or law used in climate science is conservation of energy

• Models are an accounting, or calculating the budget, of – Energy– Mass– Momentum

• Credibility established by representation of the past, and, when possible, evaluating predictions and prejections

Summary: Energy Balance of Planet

• Earth’s energy balance– Energy from Sun– Energy sent back to space

• Things that absorb• Things that reflect• Moving energy around• Storing energy at the surface of the Earth

– Greenhouse gases hold the energy a while– Oceans pick it up and hold it longer– Ice takes it up and melts balances change

A fundamental conclusion

• Based on the scientific foundation of our understanding of the Earth’s climate, we know with virtual certainty– The average global temperature of the Earth’s surface

has risen and will continue to rise due to the addition of gases (esp, carbon dioxide) into the atmosphere that hold heat close to the surface. The increase in greenhouse gases is due to human activities, especially, burning fossil fuels.

– Historically stable masses of ice on land have melted and will continue to melt.

– Sea level has risen and will rise.– The weather has changed and will change.

Outline

• Models– Definition

• Models and Scientific Investigation• Models in Climate Science• Establishing Trust: Numerical

Experimentation • Looking Towards the Future• Summary

Some Resources

• Gettelman and Rood: Demystifying Climate Models: A User’s Guide to Earth Systems Models

– Springer, Open Source, (It is free.)

• Introductory Material OpenClimate Mini-page

• Model Introduction OpenClimate Mini-page

• Rood’s Class MediaWiki Site– http://climateknowledge.org/classes/index.php/Climate_Change:_The_Move_to_Action

Poll questions:

• I was formally introduced to weather or climate models in school.

• Our knowledge of climate change is adequate for us to take action to intervene to reduce carbon dioxide emissions.

• Climate models provide adequate information to inform decisions about adaptation.

• Climate models are trustworthy.• Provide any comments, qualifications inspired by the

questions above.• Write any questions or comments about climate

models and climate change you would like to make.• What do you want to get from this presentation?

Background Materials

Roles of Uncertainty / Variability at Different TimesHawkins and Sutton, 2009

Conservation principle

• There are many other things in the world that we can think of as “conserved.” For example, money.– We have the money that we have.

• If we don’t spend money or earn money, then the money we have today is the same as the money we had yesterday.

Mtoday = Myesterday

That’s not very interesting, or realistic

Conservation principle(with income and expense)

Mtoday = Myesterday + I - E

Let’s get some money and buy stuff.

Income

Expense

Conservation principle(with the notion of time)

Mtoday = Myesterday + N(I – E)

Salary

Income per month = IRent

Expense per month = EN = number of months

I = NxI and E= NxE

Income

Expense

Some algebra and some thinking

Mtoday = Myesterday + N(I – E)

Rewrite the equation to represent the difference in money

(Mtoday - Myesterday ) = N(I – E)This difference will get more positive or more negative as time goes on.

Saving money or going into debt.

Divide both sides by N, to get some notion of how difference changes with time.

(Mtoday - Myesterday )/N = I – E

Introduce a concept

• The amount of money that you spend is proportional to the amount of money you have:

• How do you write this arithmetically?

E = e*M

Some algebra and some thinking

If difference does NOT change with time, then

M = I/e

Amount of money stabilizes

Can change what you have by either changing

income or spending rate

(Mtoday - Myesterday )/N = I – eM

All of these ideas lead to the concept of a budget:

What you have = what you had plus what you earned minus what you spent

Conservation principle

Mtoday = Myesterday + I - E

Let’s get some money and buy stuff.

Income

Expense

Energy from the Sun

Energy emitted by Earth

(proportional to T)

Earth at a certain temperature, T

Some jargon, language

• Income is “production” is “source”• Expense is “loss” is “sink”• Exchange, transfer, transport all suggest

that our “stuff” is moving around.

Equilibrium and balance

• We often say that a system is in equilibrium if when we look at everything production = loss. There might be “exchanges” or “transfers” or “transport,” but that is like changing money between a savings and a checking account.– We are used to the climate, the economy, our cash

flow being in some sort of “balance.”– As such, when we look for how things might change,

we look at what might change the balance.– Small changes might cause large changes in a

balance

Conservation of Energy

• Conceptual model of Earth’s temperature from space

H = Heating = Production = Loss

lT = Cooling = Loss

D means the change in something, a difference

T is Temperature and t is time

DT

Dt= H -lT

Earth: How Change T?

Energy from the Sun

Energy emitted by Earth

(proportional to T)

Earth at a certain temperature, T

Stable Temperature of Earth could change from how much energy (production) comes from the sun, or by changing how we emit energy.

The first place that we apply the conservation principle is energy

• We reach a new equilibrium

HT

THt

T

LossProduction

-0

Changes in orbit or solar

energy changes this

The first place that we apply the conservation principle is energy

• We reach a new equilibrium

HT

THt

T

LossProduction

-0

Changing a greenhouse gas

changes this

Balancing the Budget

• Today’s Money = Yesterday’s Money + Money I Get – Money I Spend

• Today’s CO2 = Yesterday’s CO2 + CO2 I Get – CO2 I Spend

• Today’s Energy = Yesterday’s Energy + Energy I Get – Energy I Spend

Or Tomorrow?

Conservation principle• Conserved Quantities:

– mass (air, ozone, water)– momentum, – Energy

• Need to Define System– Need to count what crosses the boundary of

the system– System depends on your point of view

Point of View

SUN EARTH

EARTH: EMITS ENERGY TO SPACE BALANCE

PLACE AN INSULATING

BLANKET AROUND EARTH

FOCUS ON WHAT IS

HAPPENING AT THE

SURFACE

Simple earth 1

Models and Modeling

Models

• Blogs on Model Tutorial (Start with #3)• Models are everywhere• Ledgers, Graphics and Carvings• Balancing the budget• Point of View• Cloak of Complexity

What is a Model?

• Model– A work or construction used in testing or perfecting a

final product.– A schematic description of a system, theory, or

phenomenon that accounts for its known or inferred properties and may be used for further studies of its characteristics.

• Numerical Experimentation– Given what we know, can we predict what will

happen, and verify that what we predicted would happen, happened?

Models are everywhere

http://www.halfhull.com/main.jpg

How Many Use Spread Sheets?

Ledgers, Graphics and Carvings

• Ledgers • Spreadsheets Computers

Radiation Balance Figure

Let’s build up this picture

• Follow the energy through the Earth’s climate.

• As we go into the climate we will see that energy is transferred around.– From out in space we could reduce it to just

some effective temperature, but on Earth we have to worry about transfer of energy between thermal energy and motion of wind and water.

Building the Radiative Balance

What happens to the energy coming from the Sun?

Energy is coming from the sun.Two things can happen at the surface. In can be:

Reflected

Top of Atmosphere / Edge of Space

Or Absorbed

Building the Radiative BalanceWhat happens to the energy coming from the Sun?

We also have the atmosphere.Like the surface, the atmosphere can:

Top of Atmosphere / Edge of Space

Reflect

or Absorb

Building the Radiative BalanceWhat happens to the energy coming from the Sun?

In the atmosphere, there are clouds which :

Top of Atmosphere / Edge of Space

Reflect a lot

Absorb some

Building the Radiative BalanceWhat happens to the energy coming from the Sun?

For convenience “hide” the sunbeam and reflected solar over in “RS”

Top of Atmosphere / Edge of SpaceRS

Building the Radiative BalanceWhat happens to the energy coming from the Sun?

Consider only the energy that has been absorbed.

What happens to it?

Top of Atmosphere / Edge of SpaceRS

Building the Radiative BalanceConversion to terrestrial thermal energy.

1) It is converted from solar radiative energy to terrestrial

thermal energy.(Like a transfer between accounts)

Top of Atmosphere / Edge of SpaceRS

Building the Radiative BalanceRedistribution by atmosphere, ocean, etc.

2) It is redistributed by the atmosphere, ocean, land, ice, life.(Another transfer between accounts)

Top of Atmosphere / Edge of SpaceRS

Building the Radiative BalanceTerrestrial energy is converted/partitioned into three sorts

SURFACE

3) Terrestrial energy ends up in three reservoirs

(Yet another transfer )

Top of Atmosphere / Edge of Space

ATMOSPHERECLOUD

RS

WARM AIR(THERMALS)

PHASE

TRANSITION

OF WATER(LATENT HEAT)

RADIATIVE

ENERGY(infrared or thermal)

It takes heat to• Turn ice to water• And water to “steam;”

that is, vapor

Building the Radiative BalanceWhich is transmitted from surface to atmosphere

SURFACE

3) Terrestrial energy ends up in three reservoirs

Top of Atmosphere / Edge of Space

ATMOSPHERECLOUD

RS

(THERMALS)(LATENT HEAT)(infrared or thermal)

CLOUD

Building the Radiative BalanceAnd then the infrared radiation gets complicated

SURFACE

Top of Atmosphere / Edge of Space

ATMOSPHERECLOUD

RS

(THERMALS)(LATENT HEAT)(infrared or thermal)

CLOUD

1) Some goes straight to space

2) Some is absorbed by atmosphere and re-emitted downwards3) Some is absorbed by clouds and re-emitted downwards

4) Some is absorbed by clouds and atmosphere and re-emitted upwards

Want to consider one more detail

• What happens if I make the blanket thicker?

Thinking about the greenhouseA thought experiment of a simple system.

SURFACE

Top of Atmosphere / Edge of Space

ATMOSPHERE

(infrared or thermal)

1) Let’s think JUST about the infrared radiation• Forget about clouds for a while

2) More energy is held down here because of the atmosphere

• It is “warmer”

3) Less energy is up here because it is being held near the surface.

• It is “cooler”

Thinking about the greenhouseWhy does it get cooler up high?

SURFACE

Top of Atmosphere / Edge of Space

ATMOSPHERE

(infrared or thermal)

1) If we add more atmosphere, make it thicker, then

2) The part coming down gets a little larger.• It gets warmer still.

3) The part going to space gets a little smaller• It gets cooler still.

The real problem is complicated by clouds, ozone, ….

Think about that warmer-cooler thing.

• Addition of greenhouse gas to the atmosphere causes it to get warmer near the surface and colder in the upper atmosphere.

• This is part of a “fingerprint” of greenhouse gas warming.

• Compare to other sources of warming, for example, more energy from the Sun.

Think about a couple of details of emission.

• There is an atmospheric window, through which infrared or thermal radiation goes straight to space.– Water vapor window

• Carbon dioxide window is saturated– This does not mean that CO2 is no longer able to absorb.– It means that it takes longer to make it to space.

Thinking about the greenhouseWhy does it get cooler up high?

SURFACE

Top of Atmosphere / Edge of Space

ATMOSPHERE

(infrared or thermal)

3) Additional CO2 makes the insulation around the window tighter.

1) Atmospheric Window 2) New greenhouse gases like N20, CFCs, Methane CH4 close windows

The real problem is complicated by clouds, ozone, ….

So what matters?

Things that change

reflectionThings that

change absorption

Changes in the sun

If something can transport energy DOWN from the surface.

THIS IS WHAT WE ARE DOING

Think about the link to models

• energy reflected = (fraction of total energy reflected) X (total energy)

• energy absorbed = total energy - energy reflected = (1-fraction of total energy reflected) X (total energy)

• fraction of total energy reflected – Clouds– Ice– Ocean– Trees– Etc.

Radiation Balance FigureIn this figure out = in

Energy in Earth System: Basics

Can we measure the

imbalance when the Earth is

not in equilibrium?

Science Observations Evaluation Measurement

Can we do the counting to balance the budget?

Radiative Balance (Trenberth et al. 2009)In this figure out does not = in

IPCC (2001) projections for

the next century

IPCC (2007) projections for the next century

IPCC (2013) projections for the next three centuries

Radiative Forcing Changes

Interesting History of This Plot at RealClimate

• Questionnaire: Kansas City, AMS, Broadcast Meteorologists, Model Short Course