Variability of Sea-Surface Height (SSH) from ocean dynamic processes is an important component of sea-level change. In this study we dynamically downscale a present-day control simulation of a climate model to replicate sea-level variability in the Northwest European shelf seas. The simulation can reproduce many characteristics of sea-level variability exhibited in tide gauge and satellite altimeter observations. We examine the roles of lateral ocean boundary conditions and surface atmospheric forcings in determining the sea-level variability in the model interior using sensitivity experiments. Variability in the oceanic boundary conditions leads to uniform sea-level variations across the shelf. Atmospheric variability leads to spatial SSH variability with a greater mean amplitude. We separate the SSH variability into a uniform loading term (change in shelf volume with no change in distribution), and a spatial redistribution term (with no volume change). The shelf loading variance accounted for 80% of the shelf mean total variance, but this drops to ~ 60% around Scotland and in the southeast North Sea. We analyse our modelled variability to provide a useful context to coastal planners and managers. Our 200-year simulation allows the distribution of the unforced trends (over 4–21 year) of sea-level changes to be quantified. We found that the 95th percentile change over a 4-year period can lead to coastal sea-level changes of ~ 58 mm, which must be considered when using smooth sea level projections. We also found that simulated coastal SSH variations have long correlation length-scales, suggesting that observations of interannual sea-level variability from tide gauges are typically representative of > 200 km of the adjacent coast. This helps guide the use of tide gauge variability estimates.@en

The effective climate sensitivity (EffCS) of models in the Coupled Model Intercomparison Project 6 (CMIP6) has increased relative to CMIP5. We explore the implications of this for global mean sea-level (GMSL) change projections in 2100 for three emissions scenarios. CMIP6 projections of global surface air temperature are substantially higher than in CMIP5, but projections of global mean thermal expansion are not. Using these projections as input to construct projections of GMSL change with IPCC AR5 methods, the 95th percentile of GMSL change at 2100 only increases by 3–7 cm. Projected rates of GMSL rise around 2100 increase more strongly, though, implying more pronounced differences beyond 2100 and greater committed sea-level rise. Intermodel differences in GMSL projections indicate that EffCS-based model selection may substantially alter the ensemble projections. GMSL change in 2100 is accurately predicted by time-integrated temperature change, and thus requires reducing emissions early to be mitigated.@en

Changes in ocean properties and circulation lead to a spatially non-uniform pattern of ocean dynamic sea-level change (DSLC). The projections of ocean dynamic sea level presented in the IPCC AR5 were constructed with global climate models (GCMs) from the Coupled Model Intercomparison Project 5 (CMIP5). Since CMIP5 GCMs have a relatively coarse resolution and exclude tides and surges it is unclear whether they are suitable for providing DSLC projections in shallow coastal regions such as the Northwestern European Shelf (NWES). One approach to addressing these shortcomings is dynamical downscaling – i.e. using a high-resolution regional model forced with output from GCMs. Here we use the regional shelf seas model AMM7 to show that, depending on the driving CMIP5 GCM, dynamical downscaling can have a large impact on DSLC simulations in the NWES region. For a business-as-usual greenhouse gas concentration scenario, we find that downscaled simulations of twenty-first century DSLC can be up to 15.5 cm smaller than DSLC in the GCM simulations along the North Sea coastline owing to unresolved processes in the GCM. Furthermore, dynamical downscaling affects the simulated time of emergence of sea-level change (SLC) above sea-level variability, and can result in differences in the projected change of the amplitude of the seasonal cycle of sea level of over 0.3 mm/yr. We find that the difference between GCM and downscaled results is of similar magnitude to the uncertainty of CMIP5 ensembles used for previous DSLC projections. Our results support a role for dynamical downscaling in future regional sea-level projections to aid coastal decision makers.@en

Many processes affect sea level near the coast. In this paper, we discuss the major uncertainties in coastal sea-level projections from a process-based perspective, at different spatial and temporal scales, and provide an outlook on how these uncertainties may be reduced. Uncertainty in centennial global sea-level rise is dominated by the ice sheet contributions. Geographical variations in projected sea-level change arise mainly from dynamical patterns in the ocean response and other geophysical processes. Finally, the uncertainties in the short-duration extreme sea-level events are controlled by near coastal processes, storms and tides.@en