Recently KNMI released this report http://bibliotheek.knmi.nl/knmipubIR/IR2018-05.pdf
Quite an intresting read.
In this study a new classification has been developed to classify Sudden Stratospheric Warmings
(SSWs) based on their vertical depth. This new classification was developed because the depth of a
SSW has been found to be important for the magnitude of tropospheric impact (Gerber et al., 2009;
Palmeiro et al., 2015), and the official SSW classification does not tell anything about the vertical
extent of a SSW.
The new classification adapted from a previously developed classification of Kramer (2016; hereafter
referred to as K16)). It prescribes that the zonal mean zonal wind between 60N and 70N should
reverse over a depth of at least 80 hPa between 10 and 100 hPa for at least two days in a 5-day
period. This classification was termed a ‘Deep Stratospheric Warming’ (DSW).
In the stratosphere, the new DSW classification was compared to both the SSW-classification and the
K16-classification. However, the K16-classification included events that were too weak to be
considered important for tropospheric impact. Coupled to the fact that K-16-events did not contain
any information about the vertical depth of SSWs, this classification was not taken into account in
Compared to the official SSW classification, it appeared that DSWs were less frequent; whereas SSWs
occurred about 6 times per decade, DSWs were found to occur only 4 times per decade.
Furthermore, the SSW and DSW classifications were connected: most DSWs occurred a few days to
weeks after the SSW date. This was explained by the behavior of SSWs and DSWs: first a warming at
10 hPa took place (the SSW date). After that, the wind reversal extended downward in an irregular
fashion, during which after some SSWs at some point the wind reversal extended sufficiently far to
the lower stratosphere to cause a DSW to be classified.
Connected to the differences in classification date, the DSWs were rarely located close to the
moment of rapid warming in the stratosphere. Thus, unlike SSWs, the DSWs did not show the
‘sudden warming’ behavior. Rather, (most of) the DSWs seemed to be a result warm anomalies in the
mid-stratosphere that developed sometime during or after the SSW and subsequently downwelled
on a timescale of about a week.
In terms of upper tropospheric impact, the DSW classification resulted in a similar response
compared to the SSW classification: anomalously high temperatures near the pole and a southward
displaced jet stream. Furthermore, the jet stream was found to become slightly more meridionally
oriented in the midlatitudes and slightly more zonally oriented in the subtropics, but this signal was
fairly weak. The main difference between the DSW and SSW was that the DSW effects were mostly
stronger in magnitude than the SSW effects. Thus, a DSW results in stronger upper tropospheric
impacts compared to a SSW.
A similar story was found in terms of surface impacts. Both the SSW and DSW classifications resulted
in positive pressure anomalies over the pole and generally negative pressure anomalies in the
midlatitudes (indicative of a negative Arctic Oscillation); but the anomalies were more pronounced
after a DSW. The surface temperature response after both a SSW and DSW was chaotic, though,
showing that the DSW classification did not result in a stronger or more structured surface
Finally, a thermodynamic perspective was presented to explain why the DSW classification resulted
in a stronger tropospheric response compared to the SSW classification method (only in terms of a -
AO and a weaker jet stream in the midlatitudes). The cause was speculated to be that the warming
accompanied by a DSW extended further to the upper troposphere. This is important for two
1) A deeper warming causes the pressure surfaces to be less tilted from the midlatitudes to the
pole, resulting in a more pronounced weakening of midlatitude westerlies, including the
polar jet stream.
2) A warming that extends more towards the troposphere exerts more pressure on the
underlying air, because the density of air closer to the troposphere is higher. This results in a
higher surface pressure. `
In short, the new DSW classification enables to better assess the zonal mean timing and intensity of
tropospheric impact compared to the SSW classification. However, no specific conclusions could be
drawn about the zonally asymmetric effects of DSWs.