Improving tropical cyclone rapid intensification forecasts with
satellite measurements of sea surface salinity and calibrated machine
learning
Abstract
Forecasting rapid intensification (RI) of tropical cyclones (TC) is a
mission known for large errors. One under-researched factor that affects
TC intensification is sea surface salinity (SSS), which is an important
proxy for density stratification in certain ocean regions and can affect
the surface enthalpy flux under a strengthening hurricane. To
investigate the effects of SSS on TC RI, we use a previously built
statistical model consisting of a variety of machine learning (ML)
methods. A calibrator was trained on top of the ML models to correct
probability forecasts. The ML model performance is improved with the
addition of SSS in the Eastern North Pacific and the Caribbean subregion
of the North Atlantic. Limited improvement is found in the Western North
Pacific. In the Indian Ocean, SSS is also notably correlated with RI
occurrence, but the TC samples are not sufficient to train ML models.