资源环境科技发展态势分析平台

Cyclones are an important component of Antarctic climate variability, yet quantifying their impact on the polar environment is challenging. We assess how cyclones which pass through the Bellingshausen Sea affect accumulation over Ellsworth Land, West Antarctica, where we have two ice core records. We use self-organizing maps (SOMs), an unsupervised machine learning technique, to group cyclones into nine SOM nodes differing by their trajectories (1980-2015). The annual frequency of cyclones associated with the first SOM node (SOM1, which generally originate from lower latitudes over the South Pacific Ocean) is significantly (p < 0.001) correlated with annual accumulation, with the highest seasonal correlations (p < 0.001) found during autumn. While significant (p < 0.01) increases in vertically integrated water vapor over the South Pacific Ocean coincide with this same group of cyclones, we find no indication that this has led to an increase in moisture advection into, nor accumulation over, Ellsworth Land over this short time period.

Plain Language Summary Cyclones are an important component of Antarctic climate variability, yet quantifying their impact on the polar environment is challenging. We assess how cyclones which pass through the Bellingshausen Sea affect snowfall over Ellsworth Land, West Antarctica, where we have two ice core records. We use a novel machine learning technique to group cyclones differing by their pathways. The annual frequency of cyclones associated with cyclones which generally originate from lower latitudes over the South Pacific Ocean is significantly correlated with annual snowfall, with the highest seasonal correlations found during autumn. While significant increases in water vapor over the South Pacific Ocean coincide with this same group of cyclones, we find no indication that this has led to an increase in the transport of moisture into, nor accumulation over, Ellsworth Land over this short time period.