By: Matthew Couldrey
Figure 1a: Multi-model mean projection of dynamic and steric (i.e. due to thermal and/or haline expansion/contraction) sea level rise averaged over 2081-2100 relative to 1986-2005 forced with a moderate emissions scenario (RCP4.5), including 0.18 m +/- 0.05 m of global mean steric sea level change. b: Root-mean square spread (deviation) of projections from the 21 model ensemble. (From Church et al 2013, their Figure 13.16)
Greenhouse gas forced climate change is expected to cause the global mean sea level to rise over the coming century, which will affect millions of people (Brown et al 2018) and cost trillions of US dollars (Jevrejeva et al. 2018). However, local factors are important in determining how much sea level change any particular place will experience, and these regional effects can double or entirely counteract the global mean change (Figure 1a). Furthermore, regional patterns of sea level change are challenging to predict, and climate models differ in their projections of this spatial pattern (Figure 1b). My research as part of the FAFMIP project (Flux Anomaly Forced Model Intercomparison Project, http://fafmip.org) aims to better understand why models disagree on the distribution of future sea level change.
Dynamic sea level (ζ) is the local sea surface height (above a geopotential surface) deviation from its global mean. Dynamic sea level is zero when averaged over the whole ocean surface, and its change over time (Δζ) shows the local change relative to the global mean. Therefore, positive values of Δζ indicate locations where sea level rise is larger than the global mean. Note that negative values of Δζ can correspond to locations of sea level rise (where the local change is smaller than the global mean, but still a rise) as well as sea level fall.
The hotspots in Figure 1b show locations where models from the previous generation of coupled climate (CMIP5) models disagree on the spatial pattern of sea level rise. The Southern Ocean is one of the regions where the pattern is uncertain, owing to a mixture of inter-model spread in 1) the ocean response to wind forcing, 2) changes in circulation, and 3) the redistribution of heat and freshwater. In an attempt to disentangle these causal processes, my research makes use of simulations where the oceans of several different models are forced with exactly the same changes in air-sea fluxes of heat, momentum (wind) and freshwater.
Figure 2: Thermal and haline contributions to dynamic sea level change across five Atmosphere-Ocean models, rows correspond to different models (named in left hand legends). Left panels: Zonally integrated change in ocean heat content per degree of latitude. Right panels: Zonal mean dynamic sea level change (Δζ, solid lines), and contributions from thermal expansion alone (dotted lines) and thermal plus haline effects (dashed lines).
The Southern Ocean dynamic sea level response is characterised by a strong north-south gradient, with relatively little change near the Antarctic continent and a northward-increasing rise (Figure 2, solid lines of right panels). This change arises partly because more heat gets added to lower latitudes of the Southern Ocean, peaking around 40 ˚S: note the ‘hump’ in the zonal ocean heat content change (left panels of Figure 2). However, the zonal dynamic sea level change (Δζ) shows a gradient then plateau (Figure 2, solid lines of right panels), rather than a ‘hump’, unlike the zonal heat content change. This is because of two reasons: the tendency of seawater to expand or contract changes markedly as you move from 70 ˚S to 45 ˚S. This means that the same heat input causes more dynamic sea level change at lower latitudes (where seawater is warmer) than at higher latitudes (where the temperature is lower). This ‘thermosteric’ or thermal expansion effect alone (Figure 2, dotted lines of right panels) would act to emphasise the ‘hump’ in sea level change suggested by the heat content change. In fact, the ‘haline contraction’ effect is what works against the thermal effects and flattens the hump into the gradient-plateau feature that we observe (Figure 2, dashed lines of right panels closely match the solid lines).
This work highlights that while ocean heat uptake sets the broad patterns of sea level change in the Southern Ocean, it’s the salinity changes that set the details. Furthermore, all the different models shown in Figure 2 were forced with the same pattern and magnitude of air-sea heat flux change. This means that the diversity in patterns of dynamic sea level change across different models largely arises due to differing ocean responses to climate change, rather than each model’s climate sensitivity (i.e. how much a particular model warms per unit of greenhouse gas emitted).
References
Brown, S., R. J. Nicholls, P. Goodwin, I. D. Haigh, D. Lincke, A. T. Vafeidis, & J. Hinkel, 2018, Quantifying Land and People Exposed to Sea-Level Rise with no Mitigation and 1.5∘C and 2.0∘C Rise in Global Temperatures to Year 2300, Earth’s Future, 6, 583-600, DOI 10.1002/2017EF000738
Church, J. A. , Clark, P. U., Cazenave, A., Gregory, J. M., Jevrejeva, S., Levermann, A., Merrifield, M. A., Milne, G. A., Nerem, R. S., Nunn, P. D., Payne, A. J., Pfeffer, W. T., Stammer, D., and Unnikrishnan, A. S.: Sea Level Change, in: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, edited by: Stocker, T. F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S. K., Boschung, J., Nauels, A., Xia, Y., Bex, V., and Midgley, P. M., Cambridge University Press, 2013. DOI 10.1017/CBO9781107415324.026
Jevrejeva, S., L. P. Jackson, A. Grinsted, D. Lincke, and B. Marzeion, 2018: Flood damage costs under the sea level rise with warming of 1.5 ∘C and 2 ∘C, Environ. Res. Lett., 13, DOI 10.1088/1748-9326/aacc76