Seasonal Forecasting of Sea Level Anomalies in a Multi-model Prediction Framework

Abstract:

Coastal high water events are increasing in frequency and severity as global sea levels rise. With higher relative sea levels, coastal flooding and erosion are occurring more often during periods of higher astronomical tides. If combined with above-normal seasonal sea levels, often associated with climate-driven variability in the ocean, coastal impacts becomes more severe. Such total high water events expose coastlines to potential risks, yet no seasonal prediction of sea level anomalies exists globally. Advancements in forecasting seasonal climate variability using coupled ocean-atmosphere models, which have the ability to assimilate and predict sea level, provide the opportunity to predict the potential for future high water events several months in advance for many parts of the world. Here, we construct a 10-model ensemble of retrospective forecasts with lead times up to 11 months and then compare the predicted sea levels with observations from satellite-based altimetry measurements and shore-based tide gauges. Forecast skill tends to be highest in the tropical and subtropical open oceans, whereas the skill degrades in the higher latitudes and along some continental coasts. For most locations, multi-model averaging improves the skill compared to individual models. Although, we find that the best forecasts typically come from models with more accurate initializations of the sea level anomaly and also having relatively higher horizontal resolutions in the ocean. The assessment suggests that there is an opportunity for skillful seasonal sea level forecasts in many parts of the world. We also identify places with no skill, which to improve on, will require advancements in future-generation forecast systems.