Photo courtesy National Geographic

Transcription

1 Welcome to the Polar Climate WG! Quick update: 1. CSL proposal (~85% allocation awarded, 16 month POP starts October 1) 2. All NCAR CMIP5 data on ESG within next month 3. Observational needs/uses living document updated, more input welcome 4. Many CICE development and CMIP5 analysis activities ongoing. Let us know how we can help! Photo courtesy National Geographic

2 Arctic 21st century climate projections and feedbacks using CESM Jennifer E. Kay National Center for Atmospheric Research (NCAR) Dave Bailey, Marika Holland, Andrew Gettelman (NCAR) Sandrine Bony (Laboratoire de Météorologie Dynamique) Johannes Karlsson, Gunilla Svensson (Stockholm University)

3 Equilibrium Arctic response to 2xCO 2 What explains the greater Arctic warming in CAM5? Larger 2xCO 2 forcing (no tropospheric response) Less negative shortwave cloud feedbacks More positive surface albedo feedbacks Kay, Holland, Bitz, Blanchard-Wrigglesworth, Gettelman, Conley, and Bailey 2012, J. Climate CESM Special Issue

4 What about the transient 21 st century climate evolution in CCSM4 and CESM-CAM5?

5 CESM 21 st century surface warming (RCP8.5) CESM-CAM5 warms more than CCSM4 by the midlate 21 st century, both globally and in the Arctic. (RCP8.5 similar to 2xCO 2 )

6 What about Arctic sea ice loss? More 21 st century Arctic sea ice loss in CESM- CAM5 than in CCSM4 (RCP8.5 similar to 2xCO 2 )

7 Negative shortwave cloud and positive surface albedo feedback differences consistent with 21 st century Arctic response differences (again RCP8.5 similar to 2xCO 2 ) Understanding the 2xCO 2 equilibrium response (Kay et al. 2012) is relevant for RCP8.5 transient 21 st century projections.

8 Do we trust CESM projections of future Arctic climate? Which CESM CMIP5 contribution is a better match to the best available observations (especially sea ice and cloud observations)?

9 Late 20 th century Arctic sea ice extent/thickness a similar match to observations Neale et al. J. Climate (in prep)

10 COSP-enabled Arctic cloud fraction comparisons show improvement from CAM4 to CAM5 Kay, Hillman, Klein, Zhang, Medeiros, Gettelman, Pincus, Eaton, Boyle, Marchand and Ackerman, J. Climate CESM Special Issue (2012) Barton et al., JGR (under review)

11 How do CCSM4 and CESM-CAM5 compare to the CMIP5 ensemble of opportunity in RCP st century projections?

12 Transient 21 st century (RCP8.5) SURFACE TEMPERATURE

13 Shortwave Arctic feedbacks Unlike CESM, shortwave feedback differences do not clearly explain the 21 st century Arctic inter-model spread in the CMIP5 ensemble of opportunity.

14 Summary We found more Arctic warming in the CAM version with relatively large 2xCO 2 forcing, weak negative shortwave cloud feedbacks, and strong positive surface albedo feedbacks (CAM5). Over the 21 st century (RCP8.5), the Arctic warms by 9 K in CESM- CAM5 and 7 K in CCSM4. The 2xCO 2 Arctic response differences (Kay et al. 2012) helps explain the RCP8.5 response differences. While shortwave feedback differences associated with clouds are important for explaining CESM Arctic response differences, they do not clearly explain the larger Arctic inter-model spread in the CMIP5 ensemble of opportunity.

15 Transient 20 th century simulations: CCSM4 warms more than CESM-CAM5 Aerosol and greenhouse gas responses both important for explaining 20 th century warming amounts. What about 21 st century projections?

16 Models project a cloudier Arctic as the climate warms 1) Negative shortwave cloud feedback (reduces Arctic amplification and warming) 2) Positive longwave cloud feedback (enhances Arctic amplification and warming) Arctic clouds affect non-cloud feedbacks (e.g., positive surface albedo feedback). July 2, 2007 Qu and Hall 2005, Vavrus et al. 2009, Kay et al. 2011, Kay et al. 2012

17 21 st century Arctic cloud increases in CMIP5 (RCP8.5)

18 Weak relationship between Arctic and global shortwave cloud feedbacks. Why?

19 CAM5 clouds better than CAM4 clouds, both globally and in Arctic e.g., global (left) and Arctic (right) evaluation of CAM clouds using satellite observations and instrument simulators (COSP) Figures from Kay et al. 2012b, JClim

20 No correlation between positive surface albedo feedbacks and negative shortwave cloud feedbacks

21 Arctic cloud response to 2xCO 2 by surface type in two CMIP5 models

22 Weak relationship between Arctic total cloud fraction and the positive shortwave surface albedo feedback

23 Evidence that Arctic cloud properties affect albedo feedbacks CESM-CAM5 has optically thinner clouds and stronger positive surface albedo feedbacks than CCSM4.

24 Evidence that negative Arctic shortwave cloud feedbacks affect Arctic warming CESM-CAM5 has smaller Arctic cloud amount and cloud liquid water path increases and less negative shortwave cloud feedbacks than CCSM4.

25 Feedback parameter primer Feedback parameter (λ) = top-of-atmosphere flux change per degree surface air temperature warming Gregory and Mitchell 1997, Taylor et al. 2007, Gettelman et al. 2012, Kay et al. 2012

26 Which processes enhance GHG-induced Arctic amplification? DEFINITE Surface albedo feedbacks (Arctic more positive) Planck feedback (Arctic less negative) NO GHG forcing (Arctic less positive) Water vapor feedback (Arctic less positive) DEBATED Atmospheric heat transport Clouds Lapse rate feedback (Arctic positive, negative globally) Ocean heat transport (increases with increasing GHG) What is most important?

27 Do forcing differences contribute to more warming in CAM5 than in CAM4? Yes. CAM4 CAM5 Global 2xCO 2 forcing 3.5 Wm Wm -2 Arctic 2xCO 2 forcing 2.6 Wm Wm -2 Note: IPCC AR4 says global 2xCO 2 forcing is 3.7 Wm -2 with 10% uncertainty. These values are within that range.

28 Transient 21 st century (RCP8.5) TOTAL CLOUD FRACTION