- Link to replacement WX site
2002
Some dates removed - available by E mail only.
Using data available morning of 23rd March 2002 comparison day 4,5 and 8+)
Using data available morning of 13th March 2002 comparison to day 6 (Also Day 8-12)
Using data available morning of 7th March 2002 comparison to day 6 (Day 8-12 ready)
Using data available morning of 26rd Feb 2002 comparison to day 6/7 (8-12 ready)
DT 23rd Feb 2002 did anyone model spot the deepening low at day 3
DT 21st Feb 2002 Comparing UK, NOGAPS, EC, MRF JMA and ENS day 4 to 6 - not a good effort!
DT 21st Feb 2002 - ****DAY 9-12 ADDED 7/3/02****
DT 16th Feb 2002 Comparing UK, NOGAPS, EC, MRF and ENS day 4 to 7 SNOW EVENENT day 6/7 was indicated by MRF
A selection from 1999 or before
A look at 23/1200 T+36 and SSMI rain rates?
A look at 17/1200 T+36. How good are SSMI rain rates?
BOXING DAY STORM - A WEATHER AND OZONE EVENT
Summary
1997 forecast output suggested that forecasts were only reliable during mobile situations to 3 days and sometimes less. MRF model seemed to get developing lows 1 run before the others. Uk Met had been better spotting High pressure development.1998 saw an improvement in UKmet, following introdution of hi-res global, but some deterioration in AVN and MRF. NOGAPS seems very volitile - see individual comparisons for details, but can give very good detail. Low res DWD and CANADA output is normally not evaluated. EC failed to live up to its reputation based on my comparisons which usually look at situations where error is expected. JAPAN, recently available, seems to be improving.
1999 Two cases caused concern (17 FEB 99 and 28 JAN 99, "newruns"). There was good agreement between the models at short range (less than t+36) but this agreement turned out to be incorrect in detail, i.e. the weather was different!. Could there have been problems with data in the north and east Atlantic ?
2000/2001 saw increased model resolution and for UK MET and generally big improvements in forecastability for a few days ahead.
In 2002 I will look at days 4 to 6 and MRF 8 to 12 and predicability using ENS.
Information is sourced directly from WWW sites. Any comment is a private statement and does not represent the view of any organisation.
1998 saw an improvement in UKmet, following introdution of hi-res global, but some deterioration in AVN and MRF. NOGAPS seems very volitile - see individual comparisons for details, but can give very good detail. Low res DWD and CANADA output is normally not evaluated. EC failed to live up to its reputation based on my comparisons which usually look at situations where error is expected. JAPAN, recently available, seems to be improving.
1999 Two cases caused concern (17 FEB 99 and 28 JAN 99, "newruns"). There was good agreement between the models at short range (less than t+36) but this agreement turned out to be incorrect in detail, i.e. the weather was different!. Could there have been problems with data in the north and east Atlantic ?
2000/2001 saw increased model resolution and for UK MET and generally big improvements in forecastability for a few days ahead. In 2002 I will look at days 4 to 6 and MRF 8 to 12 and predicability using ENS.
Information is sourced directly from WWW sites. Any comment is a private statement and does not represent the view of any organisation.
BACK TO INDEX
Last layout update on 02/02/02