No abstract

Nine potential vorticity (PV) regimes over East Asia are identified by applying a Self‐Organizing Map and Hierarchical Ascendant Classification regime analysis to the daily PV reanalysis fields on the 300 K isentropic surface for December–March 1948–2014. According to the surface temperature anomalies over East Asia, these nine regimes are further classified into three classes, i.e., cold class (three regimes), warm class (four regimes), and neutral class (two regimes). The PV‐based East Asian winter monsoon index (EAWMI) is used to study the relationship between PV distributions and the temperature anomalies. The magnitude of cold (warm) anomalies over the land areas of East Asia increases (decreases) quasi‐linearly with the EAWMI. Regression analysis reveals that cold temperature anomalies preferentially occur when the EAWMI exceeds a threshold at ∼0.2 PVU (where 1 PVU ≡ 10−6 m2 K kg−1 s−1). PV inversion uncovers the mechanisms behind the relationships between the PV regimes and surface temperature anomalies and reveals that cold (warm) PV regimes are associated with significant warming (cooling) in the upper troposphere and lower stratosphere. On average, cold regimes have longer durations than warm regimes. Interclass transition probabilities are much higher for paths from warm/neutral regimes to cold regimes than for paths from cold regimes to warm/neutral regimes. Besides, intraclass transitions are rare within the warm or neutral regimes. The PV regime analysis provides insight into the causes of severe cold spells over East Asia, with blocking circulation patterns identified as the primary factor in initiating and maintaining these cold spells.

No abstract

To examine the combined effects of the different spatial patterns of the Arctic Oscillation (AO)‐related sea level pressure (SLP) anomalies and the El Niño–Southern Oscillation (ENSO)‐related sea surface temperature (SST) anomalies on the wintertime surface temperature anomalies over East Asia, a nonlinear method based on self‐organizing maps is employed. Investigation of identified regimes reveals that the AO can affect East Asian temperature anomalies when there are significant SLP anomalies over the Arctic Ocean and northern parts of Eurasian continent. Analogously, ENSO is found to affect East Asian temperature anomalies when significant SST anomalies are present over the tropical central Pacific. The regimes with the warmest and coldest temperature anomalies over East Asia are both associated with the negative phase of the AO. The ENSO‐activated, Pacific‐East Asian teleconnection pattern could affect the higher latitude continental regions when the impact of the AO is switched off. When the spatial patterns of the AO and ENSO have significant, but opposite, impacts on the coastal winds, no obvious temperature anomalies can be observed over south China. Further, the circulation state with nearly the same AO and Niño3 indices may drive rather different responses in surface temperature over East Asia. The well‐known continuous weakening (recovery) of the East Asian winter monsoon that occurred around 1988 (2009) can be attributed to the transitions of the spatial patterns of the SLP anomalies over the Arctic Ocean and Eurasian continent, through their modulation on the occurrences of the Ural and central Siberian blocking events.

The cooccurrence of wintertime temperature anomalies over eastern Asia and eastern North America is examined. The winter days during 1948–2014 are assigned to nine regimes by applying the self‐organizing map clustering method to the area‐averaged land surface temperature anomalies over these two regions. About half of the winter days are associated with concurrent temperature anomalies. The occurrence of the concurrent/nonconcurrent regimes is closely related to the large‐scale circulation conditions. The Eurasian teleconnection pattern and the Pacific‐North American teleconnection pattern are two dominant large‐scale circulation modes associated with the cooccurrence of the temperature anomalies, through their impacts on the intensities of the corresponding troughs. The precursor analysis reveals that the lead time of the early signals for the concurrent cold anomalies is about 4 days longer than that for the concurrent warm anomalies. In addition, the average lead time of the precursor signals for the wintertime temperature anomalies over eastern Asia is longer than that over eastern North America.