51ÁÔÆæÈë¿Ú

Abstract:

In recent years, numerical weather prediction models have begun to show notable levels of skill at predicting the average winter North Atlantic Oscillation (NAO) when initialised one month ahead. At the same time, these model predictions exhibit unusually low signal-to-noise ratios, in what has been dubbed a `signal-to-noise paradox'. We analyse both the skill and signal-to-noise ratio of the Integrated Forecast System (IFS), the European Center for Medium-range Weather Forecasts (ECMWF) model, in an ensemble hindcast experiment. Specifically, we examine the contribution to both from the regime dynamics of the North Atlantic eddy-driven jet. This is done by constructing a statistical model which captures the predictability inherent to to the trimodal jet latitude system, and fitting its parameters to reanalysis and IFS data. Predictability in this regime system is driven by interannual variations in the persistence of the jet latitude regimes, which determine the preferred state of the jet. We show that the IFS has skill at predicting such variations in persistence: because the position of the jet strongly influences the NAO, this automatically generates skill at predicting the NAO. We show that all of the skill the IFS has at predicting the winter NAO over the period 1980-2010 can be attributed to its skill at predicting regime persistence in this way. Similarly, the tendency of the IFS to underestimate regime persistence can account for the low signal-to-noise ratio, giving a possible explanation for the signal-to-noise paradox. Finally, we examine how external forcing drives variability in jet persistence, as well as highlight the role played by transient baroclinic eddy feedbacks to modulate regime persistence.

Abstract:

The polar regions are warming rapidly, exceeding the global average due to greenhouse gas emissionsfrom human activities. Using CMIP6 simulations, HadCRUT5 data, and MERRA-2 aerosol records, thisstudy shows that major 20th-century volcanic eruptions, Agung (1963), El Chicho´n (1982), and Pinatubo(1991) temporarily offset polar warming through sulfate-induced cooling. The Agung eruption suppressedArctic surface air temperature anomalies for approximately 4-5 years (1964–1968), El Chicho´n for asimilar duration (1983–1987), and Pinatubo for about 4-5 years (1992–1996), demonstrating a consistentmulti-year cooling response following each event.Without these eruptions, Arctic surface air temperatureanomalies would have increased by up to 0.5 °C, whereas volcanic activity caused short-term coolingranging from –0.07 °C to –0.4 °C. Arctic minimum temperatures (Tmin) are rising faster than maximumtemperatures (Tmax), indicating asymmetric warming. In contrast, Antarctic responses were weaker,moderated by oceanic and atmospheric dynamics. Collectively, these eruptions produced episodic multi-year cooling that temporarily interrupted the ongoing Arctic warming trajectory, with warming resumingas volcanic aerosol forcing diminished. Spatial patterns reveal pronounced warming in the central Arctic,Greenland, and Siberia in the absence of volcanic forcing, highlighting the role of eruptions in modulating polar warming and providing insights relevant to future geoengineering considerations.