Jump to content


  • Content Count

  • Joined

  • Last visited

Community Reputation

2,333 Exceptional

1 Follower

Recent Profile Visitors

22,904 profile views
  1. The nature of 'downwelling' has been described by a number of theories but whatever the downward influence on the troposphere doesn't alter the fact that the pattern of tropospheric heights will be always be seen in the stratosphere to some extent.
  2. Of course the trop imprints up into the strat.
  3. The closest 30-day analogues of 10mb 60°N zonal wind to 11/01/19 (MERRA2 data) are 20/02/01, 15/01/04 and 20/01/13 - This chart includes the GEOS forecast period to 19/01/19 during which 2004 deviates somewhat, but 2001 and 2013 remain closest analogues. From these dates, the AO chart to 11/01/19 shows closest similarity to 2013 but this then undergoes a couple of weeks in +ve territory for the forecast period while 2001 and 2004 have some fairly negative AO - Similar to the AO, the lower strat 150mb wind is closest to 2013 whereas 2001 and 2004 weakened from stronger values but notable in the near term forecast period is that these analogues all display more marked weakening than 2019. At this pressure level it is not always clear if the effects are working downwards or up, but if downwelling is suspected, it is quite poor at the moment and in the short term.
  4. The GFS parallel has attracted some attention in recent runs with increased meridionality towards the end which can be seen with reversed tropospheric zonal flow at 60°N. However the interesting thing is that this is led from the troposphere to the strat as can be seen in this plot of zonal wind at various pressure levels - There is a marked dip around Jan 22nd which propagates from the lower trop with a minimum at 30mb 51 hours later. This is not to say that the SSW hasn't potentially caused this in the first place, but concluding that these events are the result of SSW is fraught with uncertainty.
  5. How and where are you measuring this reversed tropospheric flow? It may be locally reversed associated with increased meridionality and quasi-stationary blocks but at 60°N for example, the longest continuous reversed mean 500mb zonal flow (1979-2014) in winter was 7 days up to 25/12/00, at 850mb 16 days up to 31/12/00 - there was no technical SSW at that time (close - u1060 down to ~2.8 m/s on 25/12/00). Looking at 30-day mean zonal flow at 850mb, for January it has been negative in 2010, 2001, 1985, 2011 and 2013 - only 1985 followed an SSW (2013 was preceding). In February negative 30-day mean in 1985, 2010, 1980, 1986, 2007, 1994, 1979. Of these 1985 and 2010 followed full SSW, though of course 2010 was already low as shown above. Flow anomalies may be significant, but there isn't wholesale tropospheric reversal at 60°N as seen at 10mb. The predisposition and variability of the troposphere would appear to be more important.
  6. That's seems compelling, but let's see equivalent anomaly correlation charts from any tele expert forecast.
  7. A few from MERRA (1) data - 18/01/03 1 day 09/02/10 3 days 24/03/10 3 days 01/01/85 5 days 22/02/79 8 days 08/12/87 9 days 21/02/89 9 days (initially - after 5 day break another 15 days) 11/02/01 14 days 12/02/18 17 days 26/02/99 21 days 06/01/13 22 days 24/01/09 29 days Edit: as mentioned previously and as tweeted below, this is not a typical split though, it is a displacement / wave 1 event.
  8. Further to the above the ECMWF have created a "stratosphere task force" to improve forecasts - they have an in depth (very!) report from June 2018 with download link here - https://www.ecmwf.int/en/elibrary/18259-report-stratosphere-task-force An obvious challenge to start with is that verification is against observations which may have significant errors. Less obvious is that increasing horizontal resolution degraded some performance -
  9. Here is their latest technical memorandum from October 2018 - "Evaluation of ECMWF forecasts, including the 2018 upgrade" https://www.ecmwf.int/sites/default/files/elibrary/2018/18746-evaluation-ecmwf-forecasts-including-2018-upgrade.pdf 50 hPa verification to day 5 only.
  10. There you go, CET anomalies for October 0.0, November +1.8, December +2.2 - Nov + Dec in particular good predictions matching the statistical analysis from a 5/6 month lead time. December came in at 6.9°C, another 6+ month making it 6 from the 13 coolest May/June SSTs in the area in the charts above, with 12/13 milder than average and only 2009 cooler.
  11. Wasn't sure where to post this, but there has been a lot of criticism of Chris Fawkes mild ramping, however on News 24 he just said -
  12. The 06z GFS para has really gone off on one, here are u1060 of the last 3 runs, the 06z has it's lowest of it's runs I've seen so far, -15.5 at T384 - The 850 mb zonal wind at 60°N is forecast to fall to -8.0 m/s at one point - this is lower than anything in the NCEP R2 reanalysis from Jan 1979 - Dec 2014 (struggling for more data with the government shutdown). The lowest in that period was -7.9 m/s from 09/02/85 - yes there was a notable SSW beforehand but over a month earlier. Interesting how the AO doesn't match this zonal wind especially well - the CPC AO value was -3.44 (NAO -1.06) whilst three weeks earlier on 19/01/85 the record lowest AO of -6.23 was recorded (NAO -1.64) with a 850 60°N wind of -6 m/s. Meantime, the reduction of wave propagation by the easterly flow allows the 1 mb wind to get back up to 40 m/s, quite some wind shear as it only reaches 4.5 m/s at 3 mb (Damn NaNs spoil plot!)
  13. GFS op, FV3 and NASA GEOS all go for reversal from Jan 1st midday onwards - these are 3 hourly instantaneous values, in each case if averaged for a daily value for the reanalyses then the reversal is from Jan 2nd.