Evolution of sea-level trends along the Norwegian coast from 1960 to 2100

Abstract

A first national analysis of the evolution of sea-level rates along the Norwegian coast for the period 1960–2100 has been accomplished by exploring tide-gauge records, relative sea-level projections, and detection techniques for acceleration. Firstly, sea-level rates for the two study periods 1960–2020 and 1991–2020 were estimated. Along the Norwegian coast, relative sea-level rates show significant spatial variation due to glacial isostatic adjustment. Moreover, the coastal average sea-level rate for the period 1991–2020 is significantly higher than for the period 1960–2020. Accelerations were then estimated for all combinations of start years and study periods longer than 30 years by including quadratic coefficients in regression models. It was found that the estimates strongly depend on the study period and do not provide confident estimates of climate change driven variation in the sea level along the Norwegian coast. Secondly, non-linear trends in relative sea level were reconstructed from Singular-Spectrum Analysis, which at several tide gauges revealed low rates in the 1970s, maximum rates around 1990, and declining rates thereafter. From the reconstructed trends, significant positive acceleration in the relative sea level was estimated for the period 1960–2020, while accelerations less than zero were detected for the period 1991–2020. However, the estimates for the recent period appear not robust due to the influence of decadal and multidecadal variation characterizing Norwegian tide-gauge records. Finally, by artificially extending the tide-gauge records by projections, the time when unprecedented high sea-level rates emerge was identified. With projections calculated for the intensive emission scenario RCP8.5, the climate signal of the relative sea level emerges at earliest in the late 2030s. The time of emergence is typically 5 to 10 years later if reanalysis of sea-level pressure and wind speed are used to reduce decadal variation in the tide-gauge records. This is because the meteorological regressors, as a side effect, introduce serial correlations that lead to rate estimates with enlarged standard errors. An important implication of the findings is that a possible absence of record high sea-level rates in the 2020s and 2030s does not falsify that relative sea level along the Norwegian coast is consistent with projections of RCP8.5.