Just Accepted
This article has been peer reviewed and accepted for publication. It is in production and has not been edited, so may differ from the final published form.
Seasonal sea level forecasts for the Australian coast
Abstract
Seasonal-to-interannual anomalies in coastal water levels, associated with climate drivers such as the El Nino - Southern Oscillation (ENSO) and other atmospheric and oceanographic processes, contribute to coastal sea level extremes by raising the baseline sea level on top of which storm surges and tides act. These anomalies are potentially predictable months in advance, information that could be used to construct an early warning system for coastal flooding hazards. With this aim in mind, we present a comprehensive skill assessment of Australian coastal sea level ensemble forecasts made with the Australian Bureau of Meteorology's seasonal prediction system, ACCESS-S2, between 1 and 8 months into the future. ACCESS-S2 has skill on the Australian north and west coasts, where sea level anomalies associated with tropical climate drivers can reach $\pm$20~cm. Inclusion of the impact of atmospheric surface pressure variations on sea level (the ``inverse barometer") as a post-processing step increases skill around most of the coast, particularly in the east. Forecast skill metrics that incorporate ensemble uncertainty information, such as the Continuous Ranked Probability Score and the Brier Skill Score for median, upper tercile and upper decile exceedances, show significant improvements over reference anomaly persistence or climatology forecasts. Forecast skill arises largely from ACCESS-S2's ability to forecast ENSO and its connection to Australian sea level via oceanic teleconnections, with other modes of variability playing a minor role. These seasonal sea level forecasts will be used for probabilistic sea level products with various applications, including improving community resilience to coastal flooding hazards.
ES24047 Accepted 29 April 2025
© CSIRO 2025