Investigating Sea-Level Acceleration Along the East Coast of North America

Abstract:

A number of researchers have reported on accelerated sea level along the east coast of North America, particularly in the northeast. We have previously modeled sea-level rates and accelerations from the last half of the 20th and early 21st centuries inferred from tide gauges in this region using steric sea-level changes, gravitationally self-consistent sea-level changes that includes self-attraction and loading (SAL), and glacial isostatic adjustment (GIA). We have found that, whereas the spatial variability of sea-level rates is dominated by GIA, the observed accelerations are not explained by these processes. In this talk, we first further investigate the observed accelerations, which we took to be constant during the study period. We have found, however, some evidence that the accelerations began in the timeframe 1990-2000. For example, the figure below shows the root-mean-square (rms) residual after removing a best-fit model wherein the acceleration was zero before the indicated year, for the Boston tide gauge, having one of the longest tide-gauge records. The minimimum rms residual occurs in the year 2000. A Monte Carlo simulation (red curve) shows that no time-dependence is expected from white noise. Evaluation of the statistical significance of these results has been difficult, since the postfit residuals are dominated by interannual variability. We will utilize time-dependent models for dynamic sea-level changes (including steric changes), GIA. For the Greenland ice mass, we will combine estimates of Greenland ice-mass variability obtained from recent analyses of GRACE data with long-term climate models for Greenland (from, e.g., RACMO) to calculate long-term sea-level impact. Comparing these models with tide-gauge data will yield insights into the nature and timing of accelerated sea level in this region. We will also discuss the implciations of these models for long-term global sea-level change.