Observational constraints on mixed-phase clouds imply higher climate sensitivity

Transcription

1 Observational constraints on mixed-phase clouds imply higher climate sensitivity TRUDE STORELVMO AND I. TAN (YALE UNIVERSITY) COLLABORATORS: M. KOMURCU (U. OF NEW HAMPSHIRE), M. ZELINKA (PNNL)

2 TAKE-HOME MESSAGES CALIOP CLOUD TOP PHASE RETRIEVALS SHOW THAT GLOBAL CLIMATE MODELS (GCMS) UNDERESTIMATE THE RELATIVE AMOUNTS OF LIQUID IN MIXED-PHASE CLOUDS THIS HAS IMPLICATIONS FOR THEIR ABILITY TO CORRECTLY SIMULATE AN IMPORTANT CLOUD-CLIMATE FEEDBACK INVOLVING MIXED-PHASE CLOUDS

3 GCM UNDERESTIMATION OF SUPERCOOLED LIQUID CALIOP = Cloud and Aerosol Lidar with Orthorgonal Polarization FROM KOMURCU ET AL. (2014) CESANA ET AL. (2015) AND MCCOY ET AL. (2016) HAVE SINCE CONFIRMED THE GENERAL GCM UNDERESTIMATION OF SUPERCOOLED LIQUID

4 WHAT CONTROLS SUPERCOOLED LIQUID IN GCMS? Contribution to cloud phase variability in CAM5 The Wegener-Bergeron- Findeisen (WBF) process 54.9% Heterogeneous ice nucleation 9.4% 2.1% 4.9% 12.2% 16.5% Ice crystal fall speed 10ºC isotherm Tan and Storelvmo (JAS, 2015)

5 MIXED-PHASE CLOUD SUB-GRIDSCALE STRUCTURE DISCOVERED ~25 YEARS AGO BY MITCHELL ET AL. (1989) AND LI & LE TREUT (1992) 714 J O U R N A L O F T H E A T M O S P H E R I C S C I E N C E S VOLUME 73 THE STANDARD ASSUMPTION IN CLIMATE MODELS IS THAT LIQUID AND ICE ARE UNIFORMLY MIXED THROUGHOUT EACH ENTIRE MODEL GRID BOX IN REALITY, FIELD MEASUREMENTS SHOW THAT MIXED-PHASE CLOUDS MORE TYPICALLY CONSIST OF POCKETS CONSISTING SOLELY OF LIQUID OR ICE THIS HAS CONSEQUENCES FOR HOW THE WBF PROCESS SHOULD BE PARAMETERIZED IN LARGE- SCALE MODELS FIG. 2. Schematic diagrams contrasting (a) the idealized homogeneous mixture of ice/snow and liquid within a GCM grid box, which typicallytan spansand on thestorelvmo order of 100 km(jas, in the horizontal 2015) and 1 km in the vertical, with (b) the more realistic heterogeneous mixture of ice/snow and liquid that usually exists in separate pockets of liquid and ice on the order of tens of meters to 20 km according to satellite and field observations. The gray-shaded regions represent the mixing zones, where liquid droplets and ice crystals interact via the WBF process. In (a), the entire grid box is the mixing zone. In (b), the mixing zone is reduced to include only the regions outside the outlined pockets.

6 WHAT PARTICLES ARE RELEVANT AS ICE NUCLEI IN THE ATMOSPHERE? Murray et al. (2012)

7 ICE NUCLEATION SEEN FROM SPACE THE AMOUNT OF SUPERCOOLED LIQUID IS NEGATIVELY CORRELATED WITH (IN ORDER OF STATISTICAL SIGNIFICANCE): 1. MINERAL DUST 2. MINERAL DUST MIXED WITH POLLUTION 3. SMOKE Aerosol frequency of occurrence and SLF from CALIOP ( ) Tan, Storelvmo & Choi (JGR, 2014)

8 THE CLOUD PHASE FEEDBACK FOR COMPARABLE CLOUD WATER CONTENTS, LIQUID CLOUDS ARE OPTICALLY MUCH THICKER THAN ICE CLOUDS AS THE TROPOSPHERE WARMS DUE TO INCREASING ATMOSPHERIC CO 2, ICE CLOUDS ARE REPLACED BY LIQUID CLOUDS, AND THE OVERALL CLOUD OPTICAL THICKNESS INCREASES. THIS AFFECTS BOTH LW AND SW RADIATION, BUT THE SW EFFECT DOMINATES. THE RESULTING CLOUD-CLIMATE FEEDBACK IS NEGATIVE, AND MOST IMPORTANT AT MID/HIGH LATITUDES. Storelvmo, Tan and Korolev (2015)

9 IMPACT OF SUPERCOOLED LIQUID ON EQUILIBRIUM CLIMATE SENSITIVITY (ECS) a SLF Temperature ( C) SLFb ECS ( C) Low SLF Control in situ CALIOP CALIOP SLF1 CALIOP SLF2 High SLF Low SLF Control CALIOP SLF1 CALIOP SLF2 High SLF 5 atmosphere+ocean simulations with very different amounts of super-cooled liquid were run to equilibrium with both present-day and doubled atmospheric CO 2. Two of them (CALIOP-SLF1 and CALIOP-SLF2) were designed to have SLFs similar to CALIOP (achieved by reducing IN concentration and retarding WBF process). c d Modeling tool: The Community Earth System Model (CESM) Tan, Storelvmo & Zelinka (2016)

10 RELATIONSHIP BETWEEN SLF AND ECS ECS ( C) R 2 = p = Low SLF Control CALIOP SLF1 CALIOP SLF2 High SLF SLF

11 THE CLOUD PHASE FEEDBACK IN ACTION a Low-SLF b High-SLF c d Tan, Storelvmo & Zelinka (2016)

12 DIFFERENCES IN ECS CAUSED BY DIFFERENCES IN THE CLOUD OPTICAL DEPTH FEEDBACK b Low SLF c Control Wm 2K S e CALIOP SLF2 d CALIOP SLF1 30 S 1 0 f High SLF 1 30 N 30 N 30 N 2 30 S 30 S 30 S 3 Tan, Storelvmo & Zelinka (2016) 1 30 N 30 S LF SL F2 LIO P SL F1 2 CA P CA Lo w S LF 0.8 ntr ol 30 N τlw CTP+Amt Hig h S τ τsw LIO N Co λ (Wm 2K 1) a

13 CLOUD VS. NON-CLOUD FEEDBACKS Cloud feedbacks Non-cloud feedbacks λ (Wm 2 K 1 ) τ (R 2 = 0.99, p= ) Amt (R 2 = 0.098, p=0.61) CTP (R 2 = 0.16, p=0.50) λ (Wm 2 K 1 ) Water vapour (R 2 = 0.17, p=0.48) Lapse rate (R 2 =0.0010, p=0.96) Planck (R 2 = , p=0.93) Albedo (R 2 =0.21, p=0.44) ECS ( C) ECS ( C) Tan, Storelvmo & Zelinka (2016)

14 CONCLUSION CLOUD PHASE EXERTS A DOMINANT INFLUENCE ON THE OVERALL CLOUD-CLIMATE FEEDBACK, AND THEREFORE ON CLIMATE SENSITIVITY CLOUD PHASE IS ONE OF ONLY A HANDFUL OF KNOWN EMERGENT CONSTRAINTS ON MODEL PERFORMANCE GLOBAL HIGH-QUALITY CLOUD PHASE OBSERVATIONS ARE CRITICALLY IMPORTANT FOR GCM VALIDATION AND ULTIMATELY FOR RELIABLE PROJECTIONS OF FUTURE CLIMATE