Search for: All records

Abstract Sea surface height observations provided by satellite altimetry since 1993 show a rising rate (3.4 mm yr⁻¹) for global mean sea level. While on average, sea level has risen 10 cm over the last 30 years, there is considerable regional variation in the sea level change. Through this work, we predict sea level trends 30 years into the future at a 2° spatial resolution and investigate the future patterns of the sea level change. We show the potential of machine learning (ML) in this challenging application of long-term sea level forecasting over the global ocean. Our approach incorporates sea level data from both altimeter observations and climate model simulations. We develop a supervised learning framework using fully connected neural networks (FCNNs) that can predict the sea level trend based on climate model projections. Alongside this, our method provides uncertainty estimates associated with the ML prediction. We also show the effectiveness of partitioning our spatial dataset and learning a dedicated ML model for each segmented region. We compare two partitioning strategies: one achieved using domain knowledge and the other employing spectral clustering. Our results demonstrate that segmenting the spatial dataset with spectral clustering improves the ML predictions. Significance StatementLong-term projections are needed to help coastal communities adapt to sea level rise. Forecasting multidecadal sea level change is a complex problem. In this paper, we show the promise of machine learning in producing such forecasts 30 years in advance and over the global ocean. Continued improvements in prediction skills that build on this work will be vital in sea level rise adaptation efforts. 

Abstract The emergence of a spatial pattern in the externally forced response (FR) of dynamic sea level (DSL) during the altimeter era has recently been demonstrated using climate models but our understanding of its initial emergence, drivers, and implications for the future is poor. Here the anthropogenic forcings of the DSL pattern are explored using the Community Earth System Model Large Ensemble (CESM-LE) and Single-Forcing Large Ensemble, a newly available set of simulations where values of individual forcing agents remain fixed at 1920 levels, allowing for an estimation of their effects. Statistically significant contributions to the DSL FR are identified for greenhouse gases (GHGs) and industrial aerosols (AERs), with particularly strong contributions resulting from AERs in the mid-twentieth century and GHGs in the late twentieth and twenty-first century. Secondary, but important, contributions are identified for biomass burning aerosols in the equatorial Atlantic Ocean in the mid-twentieth century, and for stratospheric ozone in the Southern Ocean during the late twentieth century. Key to understanding regional DSL patterns are ocean heat content and salinity anomalies, which are driven by surface heat and freshwater fluxes, ocean dynamics, and the spatial structure of seawater thermal expansivity. Potential implications for the interpretation of DSL during the satellite era and the longer records from tide gauges are suggested as a topic for future research.