The tipping times in an Arctic sea ice system under influence of extreme events

摘要: In light of the rapid recent retreat Arctic sea ice, extreme weather events triggering variability in ice cover has drawn increasing attention. A non-Gaussian $\alpha$-stable Levy process is thought to be an appropriate model describe such event. The maximal likely trajectory, based on nonlocal Fokker-Planck equation, applied a nonautonomous system under noise. Two types tipping times, early-warning time and disaster-happening time, are used predict critical for transition from perennially ice-covered state seasonally ice-free one, respectively. We find that increased intensity results shorter warning melting, enhanced greenhouse effect will intensify this influence, making arrival significantly earlier. Meanwhile, effect, we discover frequency advance which maintained throughout year Ocean. Finally, identify values index $\alpha$ noise $\epsilon$ $\alpha \epsilon$-space can trigger between state. These provide effective theoretical framework studying variations influence events.