Abstract: Increasing exposure to coastal flood hazards will potentially induce an enormous socio-economic toll on vulnerable communities. To accurately characterize the hazard, we must consider both natural water level variability and climate change-induced sea-level rise. In this study, we develop a paleo-proxy-based reconstruction of coastal flood events over the last 500 yr to capture natural water level variability and superimpose that reconstruction onto expected sea-level rise to explore interannual and multidecadal variability in plausible future coastal flood risk. We first develop reconstructions of leading principal components (PCs) of sea surface temperature anomalies from 1500 CE onwards, using tree-ring, coral, and sclerosponge chronology-based El Niño Southern Oscillation reconstructions as predictors in a wavelet autoregression model. These reconstructions of PCs are then used in a stochastic water level emulator to develop ensemble simulations of hourly still water levels (SWLs) in the San Francisco Bay. The emulator accounts for multiple relevant processes, including monthly mean sea level (MMSL) anomalies, storm surge, and tide, all varying at different timescales. Accounting for natural variability in water levels over 1500?2000 CE increases coastal flood risk beyond that suggested by instrumental records alone. When superimposed on 0.22 m of sea-level rise (approximately the amount experienced over the previous century), the simulations show that while high tides and large storm surges cause the smaller extreme SWLs, the larger extreme SWLs occur during concurrent high MMSL, high tides, and significant storm surges. Our findings thus highlight the need to consider natural water level variability for coastal adaptation and planning.

Autoría

Descargar referencia