We know that the El Niño Southern Oscillation (ENSO) peaks in December--January--February (DJF) but the teleconnection effect in the Amundsen Sea is the strongest in June--July--Auguest (JJA). In order to understand this, we need to first understand the mechanism of the teleconnection in detail. Previous studies has suggested that the ENSO teleconnection to Antarctica is via modulation of the Pacific-South America pattern (PSA). This is a a positive--negative--positive pressure wavetrain to runs from the tropics to the Amundsen Sea region (and onwards to the Weddell Sea), see Schneider et al 2006. Previous studies has also hypothesised that the teleconnections is due to Rossby wave generation and propagation, see Lanchlan-Cope et al 2006. However, most studies have only analysed a small amount of years (or even individual years) due to the lack of observations -- this makes it difficult to make any statistically significant statements in this area of high variability. Moreover, many previous studies do not use Rossby wave methods but rely on the geopotential height as an proxy.
We follow the methods of Li et al 2014 which applies Rossby wave methods to analyse the teleconnections from the atlantic. We found that due to differences in the jet positions between summer and winter, the Rossby wave source only appears in JJA. Furthermore, in DJF, the atmosphere does not allow propagation of the Rossby waves into the Amundsen Sea low area. Therefore, this explains the seasonality difference in the teleconnection.
This figure shows the Rossby wave source anomaly in extreme El Niño (+3.0K Niño 3.4) for DJF (top panel) and JJA (bottom panel). We can clearly see a strong Rossby wave source in JJA but not in DJF
A flowchart summarising the differences in the teleconnection between DJF and JJA. Inspired by Simpkins et al 2014.
The linearity of ENSO and the teleconnection to the Amundsen Sea region
The linearity of ENSO has been explored by many previous studies. However, the linearity of the ENSO teleconnection has been scarcely explored. Many previous studies look at the composite or linear regression of these teleconnections. While this is correct to first order and is the simplest way to understand these teleconnections. These methods may be slightly misleading. We found that the changes due to a La Niña are significantly different to changes caused by a El Niño. In certain circumstances, the non-linearity may not only be the magnitude of the anomaly but also the sign of the anomaly.
This figure shows the mean sea level pressure anomaly for extreme La Niña (-3.0K in Niño 3.4 region) in the left panel and extreme El Niño (+3.0K in Niño 3.4 region) in the right panel. We can clearly see the negative anomaly for the La Niña is not at the same positon, magnitude or structure of the El Niño.