Modelling Present-Day and Future Caribbean Sea Level Changes

By H.A. Dijkstra, IMAU, Utrecht University, The Netherlands

The Caribbean  Sea is connected  to the North  Atlantic  Ocean  and exchange  transport values (in Sv, 1 Sv = 106 m3/s) are fairly well known (Fig. 1). It plays an important role in the transport  of mass, heat and salt in the Atlantic Ocean circulation  system.   The upper ocean circulation  in the Caribbean  Sea is characterized  by a warm through flow

known as the Caribbean Current. The path and intensity of this current are strongly modulated by the passage of large anticyclonic ocean eddies. These eddies penetrate the Caribbean Sea from the Atlantic and can be traced upstream to originate from the North Brazil Current retroflection area. As they have a clear dynamics related expression in the sea surface height, they may exert a large influence on regional sea level.

Sea level is changing over the North Atlantic Ocean and in particular over the Caribbean Sea.  The linear trend over the Caribbean Sea is estimated to be about 1.7 mm/yr over the period 1993-2010 with strong regional inhomogeneities (Torres and Tsimplis, 2013). Several processes were identified that could be responsible for the variations in sea level change in the Caribbean Sea (local wind changes, temperature and salinity changes, connections to El Niño/Southern Oscillation and vertical land movements) but no precise attribution of these changes has been made. As the flow conditions  in  the  Caribbean  Sea  strongly  depend  on  the  inflow  from  the  Atlantic Ocean, regional sea level changes are, however, expected to be strongly dependent on changes in the large-scale Atlantic flow, in particular changes in the North Brazil Current.

For the projection of the future regional sea level changes under the increase of greenhouse  gases, currently  Global Climate Models (GCMs) are used. In the IPCC- CMIP5  model  simulations  several  Representative  Concentration  Pathways  (RCPs) were used to study the effect of different radiative forcing scenarios on global climate and sea level. Under the RCP4.5 scenario, mean sea level increases (over the years 2000- 2100) along all coastlines is projected to range from 20 to 50 cm (peak at 40 cm) and under an RCP8.5 scenario from about 40 cm to 80 cm with a peak at 75 cm. However, for the projection of the future regional sea level changes in the Caribbean Sea,  an  adequate  modelling  of the  regional  changes  in  the  North  Atlantic  Ocean circulation are crucial. These are not well captured by the GCMs as used in the CMIP5 because the ocean models do not have enough spatial resolution (typically 100 km is used).   Once eddies are resolved in the ocean models, the connection between the North Brazil Current and the Caribbean Sea, and hence much more detail in regional sea level changes,  can be captured.

What is also not considered in the CMIP5 simulations (Church et al., 2013) is the extreme  scenario  that  the  Atlantic  Meridional  Overturning  Circulation  (MOC)  will undergo rapid changes in the near future because of its sensitivity to freshwater perturbations in the northern North Atlantic. Observational evidence from the RAPID mooring  array  indicates  that  the  zonally  averaged  volume  transport  at  26oN  has decreased over the last decade.   The upper branch of the MOC passes through the North Brazil Current. The frequency and intensity of the North Brazil Current eddies varies with the variability of the MOC. Although MOC collapses have not been found in projections of the present-day GCMs, there are indications that these models fail to adequately represent the essential processes leading to such rapid changes (Weijer et al., 2012).

Substantial decreases in the MOC due to freshwater input from the Greenland Ice Sheet have been simulated in global high-resolution models. For example, in the POP model,  the MOC decreases  by about 15 Sv in 50 years due to a strong  (0.5 Sv) freshwater perturbation. Over a period of 50 years also the North Brazil Current and consequently the circulation in the Caribbean Sea changes drastically (Fig. 2a) with substantial consequences for the annual mean sea level (Fig. 2b). This is not only due to changes in the mean path of the North Brazil Current, but also due to changes in eddy formation regions and the propagation corridors.

My presentation will address (past and possible future) connections between changes in the large-scale Atlantic Ocean circulation and sea level changes in the Caribbean Sea, with focus on the modelling requirements to determine: (i) how much of the Caribbean  regional  sea level  changes  over  the past  decades  can be attributed  to dynamic sea level changes caused by variations in the Atlantic large-scale flow, and (ii) whether future changes in ocean currents in the North Atlantic will have a major influence  on the path  and strength  of the Caribbean  Current  (and  the associated eddies) and hence on regional sea level and sea level extremes.

References:

Church, J., et al. (2013), 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, Cambridge Univ. Press, Cambridge, U.K.

Den Toom, M., Dijkstra, H. A.,   Weijer, W., Maltrud, M. E., Hecht, M. W. and Van Sebille, E. (2014).   Response   of   a  Strongly   Eddying   Global   Ocean   to   North   Atlantic   Freshwater Perturbations, J. Phys. Oceanography, in press, available at http://www.staff.science.uu.nl/~dijks101/SCENES.

Torres,  R. R. and Tsimplis,  M. N. (2013).  Sea-level  trends and interannual  variability  in the

Caribbean Sea. J. Geophys. Res-Oceans 118, 2934–2947.

Weijer, W., Maltrud, M. E., Hecht, M. W., Dijkstra, H. A. and Kliphuis, M. A. (2012). Response of the Atlantic Ocean circulation  to Greenland  Ice Sheet melting in a strongly-eddying  ocean model. Geophys. Res. Lett 39, L09606.