Post on 12-Jan-2016
Predicting global mean temperature
1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007
-0.5
0
0.5
-0.5
0
0.5
Ensemble size is 5 2m Temp obs:ECMWF forecasts, mean for months 2-13, plotted at centre of verification period
70N-70S 2m Temp forecast anomalies
ec_erai ec_era4 hadcru3 Fc f3yi/m3
MAGICS 6.12n c re s s ida - ne t Fri Ja n 9 18:26:45 2009
Developments at ECMWF
• Merge of monthly forecast into EPS– Medium-range EPS is now continuous with monthly forecast– Resolution “steps down” from T399 to T255 after 10 days– Hindcasts also have high resolution for first 10 days– Forecasting system changes every few months – hindcasts
produced in near-real time. – Runs to 15 days every day; to 32 days only once per week– Ocean coupling switched off for first 10 days – for the
moment– Experimental runs to 45 days
– A first step towards unified NWP/seasonal prediction ??
Tim Stockdale, WGSIP12
Monthly All India Precipitation
VEPS
SEAS3
Correlation with IMD station data- 1991-2007
June July Aug. Sept
Month
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Correlation
(Slides from Frederic Vitart)
Interannual variability of Atlantic hurricanes 1991-2007
August September October
VAREPS
Seas - Month 1
Seas – Month 2
0.77
0.31
-0.03
0.37
0.33
0.23
0.69
0.18
0.18
Correlation with HURDAT
Global Scores2mTm- Ensemble mean correlation with
analysis
August 1981-2007
SEAS3- Month 2
VEPS June 2MTM
80°S80°S
70°S 70°S
60°S60°S
50°S 50°S
40°S40°S
30°S 30°S
20°S20°S
10°S 10°S
0°0°
10°N 10°N
20°N20°N
30°N 30°N
40°N40°N
50°N 50°N
60°N60°N
70°N 70°N
80°N80°N
160°W
160°W 140°W
140°W 120°W
120°W 100°W
100°W 80°W
80°W 60°W
60°W 40°W
40°W 20°W
20°W 0°
0° 20°E
20°E 40°E
40°E 60°E
60°E 80°E
80°E 100°E
100°E 120°E
120°E 140°E
140°E 160°E
160°E
<0.2 0.2..0.3 0.3..0.4 0.4..0.5 0.5..0.6 0.6..0.7 0.7..0.8 0.8..0.9 0.9..0.95 > 0.95
June 2MTM
80°S80°S
70°S 70°S
60°S60°S
50°S 50°S
40°S40°S
30°S 30°S
20°S20°S
10°S 10°S
0°0°
10°N 10°N
20°N20°N
30°N 30°N
40°N40°N
50°N 50°N
60°N60°N
70°N 70°N
80°N80°N
160°W
160°W 140°W
140°W 120°W
120°W 100°W
100°W 80°W
80°W 60°W
60°W 40°W
40°W 20°W
20°W 0°
0° 20°E
20°E 40°E
40°E 60°E
60°E 80°E
80°E 100°E
100°E 120°E
120°E 140°E
140°E 160°E
160°E
<0.2 0.2..0.3 0.3..0.4 0.4..0.5 0.5..0.6 0.6..0.7 0.7..0.8 0.8..0.9 0.9..0.95 > 0.95
Developments at ECMWF• Seasonal prediction “System 4”
– New ocean model – NEMO– New ocean assimilation system – 3D/4D VAR– New coupling interface – OASIS3– Better treatment of lakes and sea-ice
– Hope to have improved stratosphere (extra resolution, volcanic aerosols, improved gravity wave parameterization)
– Consistent land surface initialization (stand-alone analysis system)
– (Latest IFS cycle is much more active in tropics, generally more accurate, but has problems with easterly bias in equatorial winds, at least at TL159)
Developments at ECMWF• Decadal prediction
Only as part of EU-funded projects:
– ENSEMBLES– THOR– COMBINE
• EC-EARTH– European group adapting the ECMWF model for climate change
work– Attempt to keep technical infrastructure integrated with ECMWF– We may get feedback from some aspects of their work