Machine learning-driven skillful decadal predictions of the German Bight
storm surge climate
Abstract
The German Bight coastline is regularly affected by storm surges driven
by extratropical cyclones. Decadal-scale predictions of the local surge
climate would foster coastal protection and decision making in affected
areas. We therefore examine the prediction skill of the
Max-Planck-Institute Earth System Model (MPI-ESM) decadal prediction
system for three storm surge metrics at Cuxhaven, Germany: the annual
upper percentiles of surge heights and durations and the annual number
of surges. To avoid dynamical downscaling from the coarse model output
to local surge heights, we use machine learning and train a neural
network on observed surge heights and reanalyzed fields of mean
sea-level pressure (MSLP). We apply this network to MSLP output of our
prediction system to generate decadal predictions of surge heights at
Cuxhaven. We find that the prediction system falls short of generating
skillful predictions for individual lead years, but can predict certain
multi-year averages of surge metrics skillfully.