Liu, J. et al., PNAS, 2012 World Weather Open Science Conference, Montreal, Canada, August 17, 2014 Jiping Liu University at Albany, State University of.

Liu, J. et al., PNAS, 2012

World Weather Open Science Conference, Montreal, Canada, August 17, 2014

Jiping Liu

University at Albany, State University of New York

Contributors: Mirong Song, Radley Horton, Yongyun Hu, and Chaoyuan Yang

Reducing Spread in Climate Model Projections of a September Ice-Free Arctic

• IPCC AR4 “… late-summer sea ice is projected to disappear almost completely towards the end of the 21st century under the SRES A2 scenario in some models” (2007)

• Arctic might be ice-free in 2015-2016 Arctic Marine Shipping Assessment report (2009) Wadhams, Duarte, Maslowski, et al.

Large uncertainty in the projected timing of the ice-free Arctic in a warming environment!

Land physicsand hydrology

Ocean ecology,biogeochemistry

Atmospheric circulation and radiation

Atmospheric chemistry, aerosols

Ocean circulation

Plant ecology, land use

Climate/Earth System Models Sea Ice

CMIP5 vs. CMIP3

• A more diverse set of model types• A number of improvements in physics, numerical algorithms, and configurations• A new set of scenarios called Representative Concentration Pathways (RCPs)

Interactive CO2

CMIP5 - Coupled Model Intercomparison Project Phase 5 • A coordinated project by climate modeling community • An important resource for IPCC AR5, and beyond

~650 ppm CO2

~1370 ppm CO2

RCP4.5: medium-mitigation emission scenarioRCP8.5: high-emission scenario

Time-series of simulated September sea ice extent from 1979 to 2100 for 30 CMIP5 models (thick black: observation; thick red: multi-model ensemble mean)

Projected timing of September ice-free Arctic (defined as less than 1 million km2) for 30 CMIP5 models under RCP4.5 (blue bars) and RCP8.5 (red bars)

RCP8.5: 2011-2098!

Two different approaches:

• Model selection (based on the ability to reproduce the observed sea ice climatology and variability since 1979)

• Constrained estimation (based on the strong and persistent relationship between present and future sea ice conditions)

(a) Climatology and(b) linear trend of September sea ice extent for observations (black bar on left) and each CMIP5 model (gray bars) during 1979-2011

20% of observed climatology

30% of observed trend

ACCESS1.0, ACCESS1.3, BNU-ESM, CESM1-BGC, CESM1-CAM5, HADGEM2-CC, MIROC5, MIRO-ESM.CHEM, MPI-ESM.MR

Time-series of simulated (color lines) September sea ice extent from 1979 to 2100 for 9 selected models

Timing of ice-free: 2054

2007-2011ensemble mean

(9 selected models)2007-2011

Obs.

when September ice extent is below 1.7 million km2 (ensemble mean)

Sea ice concentration

Sea ice thickness

Two different approaches:

• Model selection (based on the ability to reproduce the observed sea ice climatology and variability since 1979)

• Constrained estimation (based on strong and persistent relationship between present and future sea ice conditions)

Scatter plot of September sea ice extent in present mean state (2007-2011) versus September sea ice extent in the projected future state (2018-2022) simulated by 30 CMIP5 models (small dots) (big dot: observed present mean state)

r=0.93

r=0.95

present state

pro

ject

ed

sta

te

present statepro

ject

ed

sta

te

2013-17 centered on 2015

2095-99 centered on 2097

Evolution of correlation (blue lines) and constrained estimation of September sea ice extent (red lines) based on the relationship between simulated September ice extent averaged for 2007-2011 and projected September ice extent averaged for 5-year sliding windows

Timing of ice-free: 2056-2060 (centered on 2058)

Projected timing for the September ice-free Arctic in CMIP5 models (RCP8.5)

Large spread: 2011-2098 (original)

Reduced spread: 2054-2058!

Increased maritime activities in the Arctic Ocean and substantial climate impacts have been emerging in the Arctic in advance of the ice-free state!

CMIP5: decadal prediction experiments

10-year simulations initialized every 5-year

1960 1970 1980 1990 2000 2010 2020

10 yrs

10 yrs

10 yrs

10 yrs

10 yrs

10 yrs

10 yrs

10 yrs

10 yrs

10 yrs

10 yrs

September sea ice extent (black: observations; thick red: ensemble mean of decadal hindcasts)

NCEP CFSv2 NASA GEOS-5

Temporal correlation of September sea ice extent anomaly between observation and decadal hindcasts (lead time of 1-year)

NCEP CFSv2 NASA GEOS-5

NCEP CFSv2 NASA GEOS-5

Temporal correlation of September sea ice extent anomaly between observation and decadal hindcasts (lead time of 2-5 year average)

Thanks for your attention!