Observed and Simulated Decadal Variability in Clouds and Water Vapour Richard Allan, Tony Slingo Environmental Systems Science Centre, University of Reading Mark Ringer Hadley Centre, Met Office
INTRODUCTION Climate Sensitivity and Feedback How does cloud and water vapour respond to changes in surface temperature? How well are interannual variations in cloud and water vapour simulated by a climate model? HadAM3 Simulations (1979-1998) Explicitly calculate “UTH” radiances Satellite-like clear-sky diagnostics Evaluation of HadAM3 using satellite data
EXPERIMENTS [Allan et al. 2003, accepted QJ] Ensemble of AMIP-type HadAM3 runs Standard res, 19 levels, 1978-1999. HadISST SST/sea ice forcing Radiance code [see Ringer et al., 2002, QJ, 129, 1169-1190] Modified clear-sky diagnostics: OLRc & water vapour radiance Additional “all-forcings” run SATELLITE DATA - column water vapour, CWV [SMMR 1979-84, SSM/I 1987-99] - clear-sky OLR [ERBS 1985-89, ScaRaB 1994/5, CERES 1998] - Water Vapour channel brightness temperature, T6.7 [HIRS 1979-1998]
Interannual variability of Column Water vapour 1980 1985 1990 1995 See also Soden (2000) J.Clim 13
Can we use reanalysis CWV? ERA-40
Upper tropospheric moisture Clear-sky OLR sensitive to Ts and RH 6.7 mm cloud cleared radiance sensitive to upper tropospheric Relative Humidity Explicitly simulate 6.7 mm radiance Modified “satellite-like” clear-sky diagnostics
Model-obs differences & Clear-sky Sampling Type II HadAM3-OBS Type-I DT 6.7 DOLRc
EOF analysis of spatio-temporal variability in water vapour radiance
Interannual monthly anomalies: tropical oceans ga=1-(OLRc/sTs4)
(+additional forcings)
Clear-sky sampling: interannual variability Light blue: Type I (weighted by clear-sky fraction) Dark Blue: Type II (unweighted mean)
Large changes in OLR from 7 independent satellite instruments (Wielicki et al, 2002)
Allan & Slingo 2002, GRL, 29(7) OLR Clear-sky OLR RSW
EITHER: (1) satellite data wrong: cannot measure decadal changes in tropical radiation budget (and hence cloudiness variations) using most well-calibrated, stable radiometers OR: (2) Our understanding of what determines decadal changes in clouds and radiation is overly simplistic REGARDLESS: measurements of tropical radiation budget provide vital information on tropical cloudiness changes
Summary Climate model simulates low-level water vapour changes Clear-sky sampling important for infrared channel climatologies but not interannual variability Simulations of satellite brightness temperatures: Consistent decadal variability suggests small DRH realistic Climate models cannot capture the large decadal changes in the tropical radiation budget (1979-1998) More wide-field of view instruments (simple yet well-calibrated & stable) Note of caution: can multiple satellite intercalibration artificially remove decadal trends in the UTH radiances? Changes in atmospheric T also influences T6.7 decadal fluctuations
Climatological mean over 60oS-60oN oceans
Info on upper tropospheric water vapour 500mb omega Clear-sky OLR
Water vapour feedback: recent advances (1) Insensitive to resolution (Ingram 2002, J Climate, 15, 917-921) (2) Consistent with observations following post-Pinatubo cooling (Soden et al 2002, Science, 296, 727)
Is water vapour feedback really consistent between models? dOLRc/dTs ~ 2 Wm-2K-1 dOLR/dTs uncertain (Cess et al. 1990, JGR, 95, 16601) Allan et al. 2002, JGR, 107(D17), 4329. doi: 10.1029/2001JD001131. - Temperature lapse rate (Gaffen et al 2000, Science, 287, 1242) - Tropical Cloudiness (Wielicki et al, 2002, Science, 295, 841)