Zeng-Zhen Hu (1)Svetlana I. Kuzmina (2)Lennart Bengtsson (3)David M. Holland (4)

(1) Center for Ocean-Land-Atmosphere Studies4041 Powder Mill Road, Suite 302, Calverton, MD 20705

E-amil: hu@cola.iges.org

(2) Nansen International Environmental and Remote Sensing CenterBolshaya Monetnaya Street, 26/28, St. Petersburg, 197101, Russia

(3) Max-Planck-Institute for Meteorology Bundesstrasse 55, Hamburg, D-20146, Germany; and Environmental Systems Science Centre,

Reading, UK

(4) Center for Atmosphere-Ocean Science Courant Institute of Mathematical Sciences, New York University

251 Mercer Street, New York, NY 10012 USA

Arctic Sea-Ice Change and Its Connection with Arctic Climate Change in CMIP2 Simulations

From: http://www.giss.nasa.gov/research/observe/surftemp/2001fig1.gif

More and More People Believes that Global Warming Is True

Possible Global Warming Signal in Observed Sea-Ice Change (Decreasing is dominated)

However, Uncertainty in CGCM Simulating the Current Climate Is Still Large, Particularly in High Latitudes.

IPCC2001, Figure 8.2: December-January-February climatological surface air temperature in K simulated by the CMIP1 model control runs.

How Can People Believe the Simulated Future Climate Change ?!

IPCC 2001, Figure 9.6: (a) The time evolution of the globally averaged temperature change relative to the years (1961 to 1990) of the SRES simulations A2 (top) and B2 (bottom) (Unit: °C). See Table 9.1 for more information on the individual models used here. (b) The time evolution of the globally averaged precipitation change relative to the years (1961 to 1990) of the SRES simulations A2 (top) and B2 (bottom) (Unit: %).

What Cause the Uncertainty ?

• It is the model differences, including:

(1) Model dynamic frame

(2) Model physical processes cloud – radiation atmosphere chemistry land surface-hydrology sea – ice so on

Among these processes, sea-ice is one of very important sources resulting in the uncertainty!

Objectives1. The two-dimensional distribution of

the mean and uncertainty of Arctic sea ice, and climate changes at the time of CO2 doubling and their interconnection

2. The sensitivity of Arctic surface air temperature (SAT) change to sea-ice area change in different models and in various periods of forced integrations of the CMIP models

3. The possible influence of CO2 doubling on the north-south SLP gradient and the mean westerly winds

4. The differences resulting from model-dependent physics

Observational & CMIP2 Data

1. Observations

1x1 sea-ice concentration in the Arctic, 1953-1995 (Chapman and Walsh, 1991); seasonal 1x1 sea-ice thickness in 1960-1982 (Bourker and Garrett, 1987)

2. Models

14 CMIP2 CGCMs (http://www-pcmdi.llnl.gov/cmip):

(see next page)

3. Experiments

CMIP2 (http://wwwpcmdi.llnl.gov/cmip):

Control run: 1-80 years

Scenario run: Standard gradual increase (1 % per year compound) in CO2, 1-80 years; CO2 doubling at 60-80 years.

CMIP2 ModelsModel Output Flux Adjustment Sea-ice Component

BMRC SIT heat, freshwater thermodynamics

CCCma SIT heat, freshwater thermodynamics

CCSR SIT heat, freshwater thermodynamics

CERFACS SIT, SIC none Thermodynamics, statistical subgrid-scale SIT distribution

CSIRO SIT, SIC heat, freshwater, momentum

thermodynamics with a cavitating fluid rheology

ECHAM3 SIT heat, freshwater,

momentum

thermodynamics

GFDL SIT heat, freshwater thermodynamics with free-drift approximation

GISS SIT, SIC none thermodynamics

LMD SIT none diagnostic

MRI SIT, SIC heat, freshwater thermodynamics with free-drift approximation

NCAR SIT, SIC none thermodynamics with a cavitating fluid rheology

NRL SIT, SIC Annual mean heat, freshwater

diagnostic

HadCM3 SIT, SIC none Thermodynamics with free-drift approximation

HadCM2 SIT, SIC heat, freshwater thermodynamics with free-drift approximation

Surface Temperature (SAT) & Sea-Ice Thickness (SIT) Are Cooperated:

Larger (smaller) warming/uncertainty of SAT is tied up with larger (smaller) reduction/uncertainty of SIT

Surface Temperature (SAT) & Sea-Ice Concentration (SIC) Are Also Cooperated:

Larger (smaller) warming/uncertainty of SAT is associated with larger (smaller) reduction/uncertainty of SIC

NH Sea-Ice Area & Arctic SAT:(1) Sensitivity (-2.0 to 0.5 C/M km**2) is varied from model to model

(2) Sensitivity is different even in different period of a transient integration;

(3) Colder (warmer) Arctic climate may favor higher (lower) sensitivity;(4) Sensitivity seems not related to initial sea-ice area.

Sensitivity & the changes in poleward ocean heat transport (Holland and Bitz, 2003, Clim. Dyn.):

MRI model: lowest sensitivity corresponds to strongest poleward ocean heat transport

NCAR model: largest sensitivity corresponds to weakest poleward ocean heat transport

Projected SLP & Sea-Ice Thickness Changes: (1) Larger (smaller) decrease /uncertainty of SLP is associated with larger

(smaller) reduction/uncertainty of SIT;(2) The north-south SLP gradient and the mean westerly winds are enhanced.

Differences resulting from model physics:Both the mean and intermodel spread patternsshow considerable differences in some regions

between models with and without flux adjustment.

Main Results• At the time of CO2 doubling, Arctic SAT increases 1 to 5 C, SIT

decreases 0.3-1.8m, SIC reduces more than 10%.

• Values of the sensitivity of Arctic surface air temperature change with respect to sea-ice area change vary from -2.0 to -0.5 C/10^6 km^2 for most CMIP2 models. The sensitivity is model dependent. For some models, the sensitivity is different even in different period of a transient integration.

• Colder (warmer) Arctic climate may favor higher (lower) sensitivity. That may be associated with the intensity of poleward ocean heat transport.

• The north-south SLP gradient and the mean westerly winds are enhanced at the time of CO2 doubling.

• There are considerable differences between models with and without flux adjustment in some regions.

• There are NO significant and consistent sensitivity differences between the mean simulations with and without sea-ice dynamics.

• Both SIT & SIC are sensitive to the increase in greenhouse gas concentrations and connected with SAT & SLP changes in the Arctic.

• Simulated mean and intermodel spread patterns of SAT change are similar to those of SIT, SIC, and SLP changes, implying that the mean and uncertainty of projected Arctic climate change may be largely affected by the interaction between sea-ice and the atmosphere.

Further Information and Acknowledgements

Further Information :Web page: ftp://grads.iges.org/pub/hu/paper/2004HUetal_JGR.pdf (JGR-Atmosphere, 109, D10106, 2004)E-mail: hu@cola.iges.org;

Acknowledgements :The authors thank E. Schneider, R. Stouffer, B. Huang, and D. Straus for their discussion and suggestions, and also B. Wu and J. Adams for their assistance in processing the observed sea-ice data.

This work was supported by grant from the U. S. Department of Energy (De-FG02-01ER63256). DMH acknowledges support from the Office of Polar Programs of the National Science Foundation grants OPP-9901039 and OPP-0084286.

All CMIP2 modeling groups are acknowledged for making the simulations available. CMIP2 is supported and the model data are distributed by the Program for Climate Model Diagnosis and Intercomparison (PCMDI) at the Lawrence Livermore National Laboratory (LLNL).

END

•THANKS

CMIP2 models simulate the mean sea-ice concentrations reasonably well

CMIP2 models simulate the mean sea-ice thickness NOT very well