Review High Resolution Modeling of Steric Sealevel Rise

Review High Resolution Modeling of Steric Sea-level Rise Tatsuo Suzuki (FRCGC, JAMSTEC) Understanding Sea-level Rise and Variability 6 -9 June, 2006 Paris, France

Introduction Sea level change under global warming in medium and low resolution AOGCMs in the IPCC TAR (Resolution of the ocean component < 1º 1º) Ø The model’s projections show a considerable spread. (Different models predict different distributions of sea level change. )

Some simulations show common features of sea level change under global warming Ø Sea level rise in the Arctic Ocean Ø Dipole pattern in the North Atlantic Ø Sea level rise along the northern boundary of the ACC Ø Reduced sea level rise in the Southern Ocean Ø Sea level rise in the Kuroshio extension Sea level change in high resolution model

Model for Interdisciplinary Research on Climate (MIROC) version 3. 2 • MIROC_hi – Atmosphere: T 106, 56 vertical levels – Ocean: 0. 28125º 0. 1875º, 48 vertical levels • MIROC_med – Atmosphere: T 42, 20 vertical levels – Ocean: 1. 4º 0. 56º(near equator), 44 vertical levels In both models, the same physics have been used.

Sea surface height (SSH) averaged between 1981 -2000 (20 C 3 M) MIROC_hi MIROC_med Sea level patterns corresponding to major ocean gyres are well reproduced.

Changes in SSH under global warming (A 1 B – 20 C 3 M: difference between the last two decades of the 21 st century and 20 th century ) MIROC_hi MIROC_med The regions with large sea level changes are more restricted to specific areas and the magnitudes are more pronounced in MIROC_hi than MIROC_med. These differences are caused by representation of the detailed ocean structure.

SSH averaged between 1981 -2000 (20 C 3 M) MIROC_hi MIROC_med Sea level distribution associated with the subpolar Gyre is realistically represented in MIROC_hi

SSH averaged between 1981 -2000 (20 C 3 M) MIROC_hi MIROC_med Sea level distribution associated with the subpolar Gyre is realistically represented in MIROC_hi represents the sea level distribution associated with the current system east of the Philippines.

Changes in SSH under global warming (A 1 B – 20 C 3 M) MIROC_hi MIROC_med Color: difference of sea surface height between last two decades of 21 st century and 20 th century (cm) Contour: sea surface height during last two decades of 20 th century (cm) MIROC_hi exhibits a region of reduced sea level rise in the North Pacific Gyre.

Changes in SSH under global warming (A 1 B – 20 C 3 M) MIROC_hi MIROC_med MIROC_hi exhibits a region of reduced sea level rise in the North Pacific Gyre. In MIROC_hi, there is a reduced sea level rise east of the Philippines.

SSH averaged between 1981 -2000 (20 C 3 M) MIROC_hi MIROC_med

SSH and Surface current vector averaged between 1981 -2000 (20 c 3 m) MIROC_hi MIROC_med The Kuroshio and the Kuroshio extension in MIROC_hi is realistically represented.

Changes in SSH under global warming (A 1 B – 20 c 3 m) MIROC_hi MIROC_med There was a reduced sea level rise north of the Kuroshio Current and an enhanced sea level rise to the south in MIROC_hi. Kuroshio acceleration is caused by changes in wind stress and the consequential spin-up of the Kuroshio recirculation (Sakamoto et al. , 2005).

Extreme sea level rise caused by warm ocean eddies • Warm ocean eddies increase the risk of flooding in the coastal areas. Okinawa Island had flood-damage on the 22 nd of July 2001 without the passage of atmospheric low. The temporal high sea level is responsible for the warm eddy above 15 cm (Tokeshi and Yanagi, 2003; Nozaki et al 2003, etc. ).

RMS of sea level variability (eddy component) during the 20 th century (cm) MIROC_hi (20 C 3 M) TOPEX/POSEIDON 1991 -2001 The regional distribution of sea level variability is well represented in MIROC_hi.

Changes in RMS of sea level variability (eddy component) under global warming (A 1 B -20 C 3 M) Globally averaged RMS of sea level variability (cm) 0. 3 cm -12 -3 3 12 cm The enhanced eddy activities are confined to specific areas. Those area are overlapped with the enhanced sea level rise around some coastal regions and Islands. Flood risk increases.

Summary ü Sea level changes in MIROC_hi are similar to that in MIROC_med on a large scale. ü High resolution model presents more detailed ocean structure changes under global warming. ü High resolution model can estimate changes in extreme sea level variability associated with ocean eddies.

Experiments Ø Spin-up Ø 109 -years integration forced by fixed external condition for 1900 for MIROC_hi Ø 560 -years integration forced by fixed external condition for 1850 for MIROC_med Ø Control-run Ø 100 -years integration for MIROC_hi Ø 400 -years integration for MIROC_med Ø 20 C 3 M-run Ø Forced by historical data during the 20 th century Ø A 1 B-run Ø IPCC SRES A 1 B scenario Ø B 1 -run Ø IPCC SRES B 1 scenario Some trends are shown in control runs. So, we subtract these control trends from 20 C 3 M run and scenario runs.

1. Steric contribution (thermal expansion and haline contraction) to Globally averaged sea level rise estimated indirectly from density changes as the equivalent volume change under mass conservation as the Boussinesq approximation was adapted in the ocean model : 　 H : globally averaged sea level rise, S: surface area of the ocean, Z: ocean depth, : in situ density, : difference from the reference state 2. Contributions of ice-sheet melt estimated using the methods of Wild et al. [2003]

Can 20 th Century Sea Level changes be Explained? Global averaged heat content Model (20 C 3 M-run) vs. observation (Levitus-data)

Can 20 th Century Sea Level changes be Explained? Decadal variation of Sea Level (mm/year) during 1950 -1998 Regression (Sea level vs. NAO index) Model (20 C 3 M-run) Observation (Ishii-data) North Atlantic Oscillation Index • Distribution is reproduced. • Phase is different.

Can 20 th Century Sea Level changes be Explained? Decadal variation of Sea Level (mm/year) during 1950 -1998 Regression (Sea level vs. NP index) Model (20 C 3 M-run) Observation (Ishii-data) North Pacific Index • Distribution is reproduced. • Phase is different.

Can 20 th Century Sea Level changes be Explained? Sea level trend (mm/year) during 1993 -2003 Model vs. Observation (TOPEX/POSEIDON) • Sea level rise along North boundary ACC and Kuroshio-extension is reproduced. • Decadal pattern of sea level is overlapped. (NAO etc. )

Heat content Levitus vs. MIROC(Hi and Mid resolution)

Globally averaged sea level rise Ø The steric contribution during 21 st century in MIROC_hi is similar to that in MIROC_med. Ø The contributions of the Greenland Antarctic ice-sheet exhibited opposite tendencies in both models, as in previous estimations [Church et al. , 2001] Ø The amplitude of ice-sheet melt in MIROC_hi was larger than that in MIROC_med Different sensitivity A 1 B run induced global warming of about 4. 0 C in MIROC_hi and 3. 4 C in MIROC_med at the end of the 21 st century, respectively.

Why steric contribution in MIROC_hi is similar to that in MIROC_med in spite of the different sensitivity? • Total heat flux into the ocean for the 21 st century is similar in the both models • The upper ocean in MIROC_hi warms up more than that in MIROC_med (different heat uptake? ). The reasons for these differences are current problems.

difference of sea level distribution between 2080 -2100 in A 1 B run and 1980 -2000 in 20 C 3 M run Spatial standard deviation of the decadal-mean field of sea level change with respect to the control run

Sea level rise trend (thermo-steric contribution) CO 2 1% increase run total upper 450 m lower 450 m

Sea level rise trend (halo-steric contribution) CO 2 1% increase run total upper 450 m lower 450 m

Future projection of the sea level change Zonal averaged sea level rise (A 1 B-run) 2081~2100 cm 2061~2080 2041~2060 2021~2040 2001~2020 Latitude