Breaker, L.C. and Ruzmaikin, A., 2013. Estimating rates of acceleration based on the 157-year record of sea level from San Francisco, California, U.S.A.
The question of the acceleration of global sea level rise has gained increasing attention because the present rate of sea level rise is relatively small in comparison to the rates that are predicted to occur in the near future. Recent measurements have come under scrutiny on how to correctly analyze and interpret regional estimates of acceleration. In this context, we employ the Ensemble Empirical Mode Decomposition (EEMD), a data-adaptive method developed for the analysis of nonstationary and nonlinear data to estimate acceleration over the 157-year record of sea level from San Francisco, California. We define sea level acceleration (SLA) as the mean of the second differences of the residual from an EEMD. Using the residual provides a means by which to reduce or eliminate the contaminating influence of decadal and longer-period oscillations in sea level that are folded into estimates obtained using the conventional approach. For the entire record, a value of 0.011 ± 0.003 mm/y2 was obtained for the acceleration and its uncertainty, compared with 0.013 mm/y2, using the conventional approach. The effect of record length is examined by estimating the accelerations for truncated versions of the record, one starting in 1900 and a second in 1925. The accelerations differed in each case from the conventional values, as expected, because the methods are based on different definitions of SLA.