On sensitivity of climate model to radiative forcing

1
(No Transcript)
2
On sensitivity of climate model to radiative
forcing Kirill Bulgakov
3
OUTLINE

• Sensitivity of equilibrium climate to CO2
  doubling in simulation of AGCM coupled to mixed
  layer ocean
• Evaluation of feedbacks from radiative damping of
  annual variation of global mean surface
  temperature

4

• The MGO AGCM
• Principal approach
• spectral representation of the main prognostic
  variables ?, D, T, q, ln p
• Model configuration
• horizontal resolution T21 or T30
• vertical resolution 14 or 25 s-levels of unequal
  thickness
• Parameterization of physical processes
• spectral treatment of solar and infrared
  radiative transfer
• diurnal cycle included
• vertical turbulent heat, moisture, and momentum
  exchange
• Tiedtke convection
• cloud prediction and precipitation formation
• gravity wave drag forcing
• heat and water transfer in 4-layer soil of 3 m
  depth

5

• Main references
• Shneerov B.Ye., V.P.Meleshko, V.A.Matyugin,
  P.V.Sporyshev, T.V.Pavlova, S.V.Vavulin,
  I.M.Shkolnik, V.A.Zubov, V.M.Gavrilina,
  V.A.Govorkova, 2001 The up-to-date version of
  the MGO global model of general circulation of
  the atmosphere (version MGO-2). MGO Proceedings,
  No.550, 3-43.
• Shneerov,B.Ye., V.P.Meleshko, V.P.Sporyshev,
  V.A.Matyugin, T.V.Pavlova, V.M.Gavrilina and
  V.A.Govorkova, 1999 MGO Atmospheric Global
  Circulation Model Current state. MGO
  Proceedings, No.547, 15-36.
• Shneerov, B.E., V.P. Meleshko, A.P. Sokolov, D.A.
  Sheinin, V.A. Lyubanskaya, P.V. Sporyshev, V.A.
  Matyugin, V.M. Kattsov, V.A. Govorkova, and T.V.
  Pavlova, 1997 MGO Global Atmosphere General
  Circulation and Upper Layer Ocean Model. MGO
  Proceedings, No.544, 3-123.
• Detailed description
• http//www-pcmdi.llnl.gov/modeldoc/amip2/mgo-01a/m
  go-01a.html

6

• Simulation of equilibrium climate with CO2 using
  different paramiterization schemes in MGO GCM
• convection schemes Tidtke, Kuo,
  Arakawa-Schubert
• variation of different parameters in cloud
  optics
• variation of parameters in cloud description
• variation of sea ice concentration

Model runs with 1xCO2 and 2xCO2 for 40 years each
and and analysis for the last 20 years.
7
Differences of computed and observed air
temperature (deg C) in winter. Runs with three
convection parameterizations.
Tiedtke convection
Kuo convection
A-S convection
8
Relationships between air surface temperature and
cloud-radiation forcing, precipitable water and
precipitation
- Simulation with Tiedtke convection
?CRF vs ?Ts
?W vs ??s
?Pr vs ?Ts
9
(No Transcript)
10
Surface air temperature change derived from
simulation of equilibrium climate with CO2
doubling in 14 AOGCMs
Mean sensitivity of equilibrium climate for 14
AOGCMs
?T143.2 0.7
11
Sensitivity of climate may be determined by
feedback parameter
- feedback for annual variation of global mean
surface temperature
1)
-outgoing annually normalized flux of
reflected solar radiation at the TOA
2)
-outgoing flux of longwave radiation
3)
- surface temperature
4)
- global average operator
5)
- contribution from feedback.
-coefficient of planetary emission
12
Evaluation algorithm   COMPUTEfeedback parameter
and gain factor USINGoutgoing terrestrial,
reflected solar radiation at TOA and surface
temperature FROM Earth Radiation Balance
Experiment and model simulated fluxes COMPARE
observed values with model simulations
13
Gain factors from the ERBE observation and the 13
AOGCMs
fs gain factor of solar component fl gain
factor of longwave component fsl gain factor of
overall feedback
14
Temporal change of annually averaged global
anomaly of surface air temperature (deg C),
computed in 16 AOGCMs and taken from observation
(CRU data) in relation to basic period 1951-1970.
Simulation with observed GHG concentration A2
scenario
Observation
Temperature anomaly, deg C
15
Gain factors from the ERBE observation and the
MGO AGCM with Tiedke convection
fsc gain factor of solar component flc gain
factor of longwave component fslc gain factor
of overall feedback
16
CONCLUSION

• Equilibrium climate shows large sensitivity to
  convection parameterization used in the model. In
  its turn, strength of cloud radiative forcing
  depends on strength of convection scheme.
• Current AOGCMs show wide spread in overall, water
  vapor and albedo feedbacks as compared with those
  derived from observation. Some improvements in
  simulation of annual cycle is further required.
• There is a need for more reliable data of
  vertical cloud distribution