III' Science Questions: Climate Prediction and Climate Model Testing

1
III. Science Questions Climate Prediction and
Climate Model Testing
2
III. Science Questions Climate Prediction and
Climate Model Testing

• Peter Pilewskie CLARREO Visible and
  Near-Infrared Studies
• Stephen Leroy Testing Climate Models with
  CLARREO Feedbacks and Equilibrium
  Sensitivity
• V. Ramaswamy Radiation spectra at TOA
  and climate diagnoses
• Mike Mishchenko Constraining climate models
  with visible polarized radiances
• Kevin Bowman Observational constraints on
  climate feedbacks A pan-spectral approach

3
Climate Model Observing System Simulation
Experiments

• Bill Collins
• UC Berkeley and LBL
• with A. Lacis (GISS) and V. Ramaswamy (GFDL)
• and J. Chowdhary, D. Feldman, S. Friedenrich, L.
  Liu, V. Oinas, D. Schwarzkopf

4
Simulation and the CLARREO questions

• Societal objective of the development of an
  operational climate forecast The critical
  need for climate forecasts that are tested and
  trusted through state-of-the-art observations.
• Objectives of OSSE
• Use models as perfect worlds to understand
  utility of CLARREO for detection and attribution
  vs. models.
• Prepare climate modeling communityfor direct
  application of all-sky radiances for evaluation
  and assimilation.

5
Climate prediction and its components
IPCC AR4, 2007
6
Historical radiative forcing
IPCC AR4, 2007

• Probability that historical forcing gt 0 is very
  likely (90).
• However, confidence in short-lived agents is
  still low at best.

7
Forcing scenarios for 21st century
IPCC AR4, 2007
Longwave The 5 to 95 percentile range of at
2100 is 50 of the mean. Shortwave The models
do not agree on sign or magnitude of forcing.
8
Projection of regional temperatures

• Roughly 2/3 of warming by 2030 is from historical
  changes.
• Uncertainties at 2100 are from physics and
  emissions.

IPCC AR4, 2007
9
Uncertain cloud radiative response
Change in cloud radiative effects in 21st
century A1B Scenario
Change from 1980-1999 to 2080-2099
IPCC AR4, 2007

• Models do not converge on sign of change in cloud
  radiative effects.
• Trends in cloud radiative effects have magnitude
  lt 0.2 Wm-2 decade-1.

10
Goals of the OSSEs

• Test the detection and attribution of radiative
  forcings and feedbacks from the CLARREO data
• Determine feasibility of separating changes in
  clouds from changes in the rest of the climate
  system
• In solar wavelengths, examine feasibility of
  isolating forcings and feedbacks
• Quantify the improvement in detection and
  attribution skill relative to existing
  instruments

11
Role of climate models in OSSEs

• Goals of OSSEs require projections of climate
  change.
• Sole source of these projections climate models
• Advantages of climate models for this
  application
• Identification of forcings for each radiatively
  active species
• Separation of feedbacks associated with water
  vapor, lapse rate, clouds
• Tests of CLARREO concept with climate models
• To what extent can forcings and feedbacks can be
  separated and quantified using simulated CLARREO
  data?
• What are the time scales for unambiguous
  detection and attribution?

12
Application of CLARREO to Models
ForcingProjection
AttributedForcing
Climate System
CLARREO
Climate Models
Forcing
FeedbackProjection
AttributedFeedback
Climate System
CLARREO
Climate Models
Forcing
13
Schematic of Tests
Forcing Projection
SimulatedForcing
Climate Models
CLARREO Emulator
Forcing
Compare
FeedbackProjection
SimulatedFeedback
Climate Models
CLARREO Emulator
Forcing
Model Feedback
Compare
14
Individual forcings in Climate Models
IPCC AR4, 2007
MIROCSPRINTARS
15
Individual feedbacks in Climate Models
IPCC AR4, 2007
16
Major steps in Climate OSSEs

• Conduct OSSEs with 3 models analyzed in the IPCC
  AR4
• Add adding two new components to these models
• Emulators for the shortwave and infrared CLARREO
• More advanced spectrally resolved treatments of
  surface spectral albedos
• Results from emulators serve as surrogate CLARREO
  data
• Estimate the forcings and feedbacks from
  emulators
• Compare to forcings / feedbacks calculated
  directly from model physics

17
Models for Climate OSSEs

• Three models for OSSEs
• NASA Goddard Institute for Space Studies (GISS)
  modelE (Schmidt et al, 2006)
• NOAA Geophysical Fluid Dynamics Laboratory
  (GFDL) Coupled Model CM-2 and CM-2.1 (Delworth et
  al, 2006)
• NCAR Community Climate System Model CCSM3
  (Collins et al, 2006).

18
Model Simulations for Climate OSSEs

• Three classes of simulations for OSSEs
• Pre-industrial conditions with constant
  atmospheric composition
• 21st century with the IPCC emissions scenarios
• 20th and/or 21st centuries with single forcings,
  e.g., just CO2(t)

IPCC AR4, 2007
19
Candidate CLARREO Emulators

• MODerate spectral resolution atmospheric
  TRANSmittance (Modtran4) version 3 (Berk et al,
  1999)
• Spectral resolution of Modtran4
• 0 to 50,000 cm-1 1 cm-1
• Blue and UV 15 cm-1
• Relationship to CLARREO
• Infrared 1X
• UV/Blue/NIR 10-100X
• Alternate emulators
• AER, GISS, GFDL, and NCAR LBL codes

Berk et al, 1999
20
TOA shortwave spectrum

• Profile AFGL mid-latitude summer with 2000 AD
  long-lived greenhouse gases.
• Sun-satellite geometry solar zenith angle
  53o, satellite zenith 0o.
• Spectral parameters 15 cm-1 resolution with no
  instrumental convolution.
• Radiative transfer code Modtran 4,

21
Shortwave spectral forcings
Absolute Forcing
Relative Forcing

• Forcing calculations
• CO2 287 to 574 ppmv (2CO2-1870)
• N2O 275 to 316 ppbv (2000-1870)
• CH4 806 to 1760 ppbv (2000-1870)
• N2O 100 to 120 PW (2CO2 feedback)

22
Primary steps in the OSSE

• Phases for the study
• Linking the CLARREO emulator with the climate
  models
• Adoption of spectral surface emissivity and BDRF
  models
• Simulations for a constant composition to
  determine the natural variability
• Simulations of CLARREO measurements for
  transient climate change

Model Archive
CLARREO Emulator
Emulation Validation
23
Natural variability in the spectra
25-day Variability, Central Pacific
25-day Variability, Western Pacific
Huang et al, 2002

• Goal quantify signal-to-noise ratios for
  forcings and feedbacks (cf Leroy et al, 2007)
• .Calculations pre-industrial conditions for
  background radiance field

24
Issues for the Emulation

• For speed and expediency, we recommend using
  using the existing IPCC archives for
  emulation.
• The reason? Centennial length simulations are
  very expensive.
• The trade-offs
• Highest temporal sampling daily means of model
  state
• Nominal temporal sampling monthly means of
  model state
• This precludes reproducing the space-time track
  of CLARREOs orbit
• For solar, we can reproduce monthly-mean solar
  zenith (latitude)
• Result Our results are an upper bound on
  detection/attribution skill
• Our results would reflect perfect diurnal
  sampling at each grid point.
• Alternate, but remote, possibility time-slice
  experiments
• Advantage interactive coupling and capture
  space-time sampling

TimeSlice
TimeSlice
TimeSlice
TimeSlice
25
Additional Issues for the Emulation

• Atmospheric conditions
• All-sky predominant condition for 100-km pixels
• Clear-sky sets upper bound for
  detection-attribution skill for non-cloud
  forcings and feedbacks
• Detection and attribution projection onto
  spectral basis functions for single forcings
  and feedbacks

Anderson et al, 2007
26
First stage of the OSSE

• Objective Configuration and initiation of the
  OSSEs
•  
• Simulation of CLARREO measurements from IPCC
  model results, including
• Calculations for pre-industrial conditions
• Calculations for transient climate change with
  all forcings
• Perform parallel calculations for all-sky and
  clear-sky conditions
• Estimation of natural (unforced) variability in
  the simulated CLARREO data

27
Second stage of the OSSE

• Objective Detection and estimation of radiative
  forcings
•  
• Simulation of CLARREO measurements from IPCC
  model results, including
• Calculations for transient climate change from
  single forcings
• Calculation of spectral signatures of shortwave
  and longwave forcings from reference
  radiative transfer calculations
• Estimation of radiative climate forcing from
  simulated clear-sky CLARREO data
• Projection global CLARREO simulations onto
  single-forcing spectral signatures to isolate
  time-dependent forcings
• Repeat forcing estimation for all-sky fluxes
• Quantify degradation in forcing estimates and
  time-to-detection from the substitution of
  all-sky for clear-sky observations

28
Conclusion of the OSSE

• Objective Detection and estimation of radiative
  feedbacks
•  
• Estimation of radiative climate feedbacks from
  the simulated CLARREO data
• Estimation of surface-albedo feedbacks for clear
  and all-sky data
• Estimation of water-vapor/lapse-rate feedbacks
  for clear and all-sky data
• Estimation of cloud feedbacks from all-sky data
  only
• Comparison of estimates with feedback estimates
  derived independently
• Characterize improvements in estimates and
  time-to-detection relative to existing
  satellite instruments

29
Key questions for Climate OSSEs

• Can clear-sky shortwave forcings and feedbacks be
  detected and quantified using CLARREO data?
• Can all-sky shortwave forcings and feedbacks be
  detected and quantified using CLARREO data?
• Can all-sky longwave forcings and feedbacks be
  detected and quantified using CLARREO data?
• To what extent is it possible to isolate forcings
  and feedbacks associated with changes in
  specific species and processes in the CLARREO
  measurements?