Evaluating climate model using observations of tropical radiation and water budgets

1
Evaluating climate model using observations of
tropical radiation and water budgets

• Richard P. Allan, Mark A. Ringer
• Met Office, Hadley Centre for Climate Prediction
  and Research, U.K.
• Tony Slingo (ESSC, Reading University)

1) Are the unexpected observed changes in the
tropical radiation budget real? 2) Using cloud
forcing ratio to infer cloud properties 3)
Radiance simulations within the climate model
2
Are the large observed decadal changes in the
tropical radiation budget artefacts of the
instrumentation?
Wielicki et al. (2002) Science, 295, 841-844
3
Total Channel Offset ERBS orbit drift
Are the changes in semi-annual SW flux cycle
artefacts of orbit? Yes. Combination of 36-day
precession cycle AND drift in ERBS orbit
Offset increase expected. 1993 Jump in offset
due to power-off - also expected. Deep-space
checks in calibration over 15 years suggest
uncertainty lt 0.5 Wm-2
4
Decadal variability still appears robust and
significantly larger than climate models predict
Updated seasonal flux changes
Updated interannual time-series
5
Multi-spectral evaluation of simulated cloud
properties (!)
Ratio of cloud albedo effect to cloud greenhouse
effect. Removes to first order dependence on
cloud fraction Improves interpretation of cloud
properties from radiation budget Sub-sample
dynamic regime
6
Analysing cloud radiative effect in tropical
convective regime
(a) ERBS radiation ERA40 vertical velocity
ISCCP cloud (f) New dynamics model (atmos) (h)
New dynamics, coupled
Not enough cloud cover, too bright?
7
(No Transcript)
8
Conclusions

• Unexpected changes in semi-annual cycle of top of
  atmosphere albedo is an artefact combination of,
• Drift in orbit of ERBS satellite
• Mismatch between 72 day precession cycle and
  monthly means
• Still believe decadal variation in tropical
  radiation budget which climate models cannot
  reproduce
• Cloud forcing ratio and dynamic regime analysis -
  techniques to evaluate the climate model
• 20-year radiance simulations completed - how
  reliable is the Bates et al. HIRS data?