Climate%20Models

1
Climate Models
Primary Source IPCC WG-I Chapter 8 - Climate
Models and Their Evaluation
2
Part 1 Model Structure
3
The Climate System
How do we simulate this?
4
Starting Point Fundamental Laws of Physics
1. Conservation of Mass
But - these are complex differential
equations! How can we use them?
2. First Law of Thermodynamics
By solving them on a grid.
3. Newtons Second Law
Plus conservation of water vapor, chemical
species,
5
Global Climate Models Structure
(Bradley, 1999)
6
Resolution Increases over Time
Computing demand increases inversely with cube of
horizontal resolution.
Increased computing power has allowed increased
resolution
7
Development of Global Climate Models (GCMs)
and increasing complexity.
Which should be favored?
8
Global Climate Models Land-Atmosphere Link
Differing scales distributed surface properties
9
Global Climate Models Development of Ocean
Models
(Bradley, 1999)
10
Global Climate Models Parameterization
Important processes smaller than a grid
box e.g., thunderstorms (atmospheric convection)
few km
(www.physicalgeography.net)
Whats a model to do?
Parameterization Represent the effects of the
unresolved processes on the grid. Assume that
unresolved processes are at least partly driven
by the resolved climate.
(www.physicalgeography.net)
11
Higher Resolution Can Help
Part of a Global Climate Model 2.5 (lat) x
3.75 (lon)
Regional (limited-area) Climate Model 0.5
(lat) x 0.5 (lon)
12
Higher Resolution Can Help
Part of a Global Climate Model 2.5 (lat) x
3.75 (lon)
Regional (limited-area) Climate Model 0.5
(lat) x 0.5 (lon)
13
EndPart 1 Model Structure
14
Part 2 Model Evaluation
15
How Are Models Evaluated?

• Testing against observations (present and past)
• Comparison with other models
• Metrics of reliability
• Comparison with numerical weather prediction

16
What Limits Evaluation?

• Unforced (internal) variability
• Availability of Observations
• Accuracy of Observations
• Accuracy of Boundary Conditions (Forcing)

These help determine what is good simulation.
17
GCM Simulations of Global T
58 simulations, 14 GCMs
5-95 confidence limits (obs)
Ensemble Average
18
Time Average Surface Temperature (1980-1999)
C
Mean Model Average of 23 GCMs
19
Errors in Simulated Surface Temperature
(1980-1999)
Lines Observed mean Colors (top) Ensemble
mean - obs.
Spatial pattern correlation 98 (individual
models)
Colors (bottom) RMS differences in
simulated-observed time series (i.e., typical
error)
20
Annual Variability (Seasons)
Lines Observed Standard Deviation (of monthly
means) Colors Ensemble mean - observations
21
Diurnal Temperature Range (1980-1999)
C
Mean Model Average of 23 GCMs
Tendency to be smaller than observed Problems
with clouds? Boundary layer?
22
Atmospheric Zonal Average (1980-1999)
K
Mean Model Average of 20 GCMs Vertical Axes
Left - Pressure (millibars) Right - Elevation
(kilometers)
Tendency for cool polar tropopause. Persistent
feature of GCMs, though now smaller
23
Mean Reflected Solar Radiation
(1985-1989)
Average of 23 GCMs (dashed) Colors Individual
Models
Satellite Observations (solid)
24
Mean Emitted Infrared Radiation
(1985-1989)
Satellite Observations (solid)
Average of 23 GCMs (dashed) Colors Individual
Models
25
Zonal Average Precipitation
(1980-1999)
Observations (solid)
Average of 23 GCMs (dashed) Colors Individual
Models
26
Annual Mean Precipitation(1980-1999)
Observations
Average of 23 GCMs
27
Atmospheric Specific Humidity (1980-1999)
g/kg
Mean Model Average of 20 GCMs
Vertical Axes Left - Pressure
(millibars) Right - Elevation (km)
(bias)
Moist bias in tropical troposphere
- 40
40
28
Ocean (Potential) Temperature (1957-1990)
Mean Model Average of 18 GCMs
29
Ocean Salinity (1957-1990)
PSU
Vertical Axes Depth (m)
Mean Model Average of 18 GCMs

• PSU practical salinity units
• based on conductivity of electricity in water
• PSU 35 ? water is 3.5 salt

30
Ocean Heat Transport
(Feb 85 - Apr 89)
Models 1980-1999
31
Sea Ice Simulation
March
September
14 GCMs (1980-1999)
Number of Models models with ice cover gt 15
in the 2.5 x 2.5 region. Red lines Observed
15 concentration boundaries
32
El Niño - Southern Oscillation (ENSO)
Recent GCMs ( 2000-2005)
Power amount of variability occurring for a
cycle length (period)
Previous generation GCMs ( 1995-2000)
33
Are Models Improving? - 1
Normalized RMS error / observed space-time
variability
34
Are Models Improving? - 2
(Reichler and Kim, 2007)
Performance Index combines error estimates of
Sea level pressure Temperature Winds Humidity P
recipitation Snow/Ice Ocean salinity Heat flux
35
End Part 2 Model Evaluation
36
Part 3 Model Feedbacks
37
Positive Feedback Example
How does Earths temperature get established and
maintained?
38
Greenhouse Effect - 1
IR radiation absorbed re-emitted, partially
toward surface
Solar radiation penetrates
39
Greenhouse Effect - 2
IR radiation absorbed re-emitted, partially
toward surface
Net IR 25-100 W-m
Emitted IR 200-500 W-m
40
Greenhouse Effect - 3
Cooler atmosphere - Less water vapor - Less
IR radiation absorbed re-emitted
Solar radiation penetrates
41
Greenhouse Effect - 4
Cooler atmosphere - thus less surface
warming - cooler surface temperature
Solar radiation penetrates
42
Positive Feedback

1. Perturb climate system
2. Positive feedback moves climate away from
   starting point
3. A destabilizing factor

Other examples - ice-albedo feedback -
CO2-ocean temperature feedback
43
Negative Feedback

1. Perturb climate system
2. Negative feedback moves climate back toward
   starting point
3. A stabilizing factor

• Example
• Decrease Earths temperature
• Cooler Earth emits less radiation (energy)
• Outgoing radiation lt solar input
• Net positive energy input
• Earth warms up from net energy input

44
Key Feedbacks - 1

• Water Vapor
• Warmer atmosphere can contain more water vapor
• Increased water vapor increases greenhouse effect
• Atmosphere warms further

• 2. Clouds
• Clouds cool the climate (reflect sunlight) and
  warm the climate (block outgoing infrared
  radiation)
• Changes in cloud distribution can thus amplify or
  reduce the warming

45
Key Feedbacks - 2

• Snow-ice albedo
• Warmer climate has reduced snow and ice
• Surface reflects less and absorbs more solar
  radiation
• Climate warms further

• 2. Lapse rate (decrease of T with height)
• In warmer climate, especially tropics,
  temperature decreases less with height
• Upper troposphere warms more than surface
• Upper troposphere emits energy to space (infrared
  radiation) more effectively than surface,
  countering the greenhouse effect.

46
Feedback Strengths
47
End Part 3 Model Feedbacks
48
END
Climate Models