The Coupled Earth Reanalysis system [CERA] Patrick Laloyaux Acknowledgments: Eric de Boisséson, Magdalena Balmaseda, Dick Dee, Peter Janssen, Kristian Mogensen, Jean-Noël Thépaut and Reanalysis Section November 19, 2014 Patrick Laloyaux (ECMWF) CERA November 19, 2014 1 / 15
Outline 1 Operational ECMWF reanalysis system 2 Coupled earth reanalysis system 3 Assessment of the performance 4 Test run for CERA20C 5 Conclusions and future steps Patrick Laloyaux (ECMWF) CERA November 19, 2014 2 / 15
Table of Contents 1 Operational ECMWF reanalysis system 2 Coupled earth reanalysis system 3 Assessment of the performance 4 Test run for CERA20C 5 Conclusions and future steps Patrick Laloyaux (ECMWF) CERA November 19, 2014 2 / 15
Context and purpose of reanalysis for climate studies Patrick Laloyaux (ECMWF) CERA November 19, 2014 3 / 15
Operational reanalysis systems at ECMWF Uncoupled assimilation system Atmospheric reanalysis [ERA-INTERIM]: consistent with the atmospheric-wave-land model air-sea interface is prescribed Ocean reanalysis [ORAS4]: consistent with the ocean-ice model forced by the atmospheric reanalysis The atmos reanalysis has a limited effect on the ocean reanalysis The ocean reanalysis has no effect on the atmos reanalysis Patrick Laloyaux (ECMWF) CERA November 19, 2014 4 / 15
Observations considered in reanalysis production Extended Climate Reanalysis Ocean observation type CTD/ARGO/Moorings XBT HADISST2 Ocean variable Temperature Salinity Temperature Sea surface temperature Satellite Era Reanalysis Ocean observation type CTD/ARGO/Moorings XBT HADISST2 Ocean variable Temperature Salinity Temperature Sea surface temperature Atmospheric observation type Atmospheric variable Atmospheric observation type Atmospheric variable Weather stations Surface pressure Weather stations Surface pressure Ships and buoys Surface pressure 10m wind direction/force Ships and buoys Radiosondes and profilers Aircraft report Satellite sounding Surface pressure 10m wind direction/force upper air wind direction/force upper air temperature Specific humidity Upper air wind direction/force Temperature Radiance Radio occultation bending angle 10m wind direction/force Atmospheric motion winds Wave observation type Wave variable Wave observation type Wave variable Altimeters Wave height Land observation type Land variable Land observation type Land variable Weather stations 2m Temperature 2m relative humidity Snow depth Satellite sounding Snow cover Patrick Laloyaux (ECMWF) CERA November 19, 2014 5 / 15
Table of Contents 1 Operational ECMWF reanalysis system 2 Coupled earth reanalysis system 3 Assessment of the performance 4 Test run for CERA20C 5 Conclusions and future steps Patrick Laloyaux (ECMWF) CERA November 19, 2014 5 / 15
The Coupled Earth Reanalysis System - ERACLIM2 project CERA system Coupled earth model: 1-hour atmosphere-ocean coupling frequency Atmosphere: 1.125 with 137 levels Ocean: 1 with 42 levels (first layer of 10 meters) Atmospheric and Ocean reanalysis: consistent with the coupled earth model consistent air-sea interface Patrick Laloyaux (ECMWF) CERA November 19, 2014 6 / 15
Assimilation in the CERA system CERA system Sequential approach: A common 24-hour assimilation window observation misfits computed by the coupled model Increments computed separately Two outer iterations in 4DVAR and 3DVAR: ocean-atmosphere correlations in the assimilation cycle Patrick Laloyaux (ECMWF) CERA November 19, 2014 7 / 15
Model level Assimilation in the CERA system CERA system Sequential approach: A common 24-hour assimilation window observation misfits computed by the coupled model Increments computed separately Two outer iterations in 4DVAR and 3DVAR: ocean-atmosphere correlations in the assimilation cycle Time step : 0h atmospheric temperature increment -0.55-0.45-0.35-0.25-0.15-0.05 0.05 0.15 0.25 0.35 0.45 0.55 20 40 60 80 0 2010080103 tn Z-0.00000 195801 mean 20 40 Depth (m) 60 80 100 2000 3000 4000 5000 6000 180 160W 140W 120W 100W 80W Longitude latitudes in [-1.0, 1.0] - (7 points) (): Min= -0.01, Max= 0.39, Int= 0.10-0.55-0.33-0.11 0.11 0.33 0.55 ocean temperature increment Patrick Laloyaux (ECMWF) CERA November 19, 2014 7 / 15
Model level Assimilation in the CERA system CERA system Sequential approach: A common 24-hour assimilation window observation misfits computed by the coupled model Increments computed separately Two outer iterations in 4DVAR and 3DVAR: ocean-atmosphere correlations in the assimilation cycle Time step : 6h atmospheric temperature increment -0.55-0.45-0.35-0.25-0.15-0.05 0.05 0.15 0.25 0.35 0.45 0.55 20 40 60 80 0 2010080103 tn Z-0.00000 195801 mean 20 40 Depth (m) 60 80 100 2000 3000 4000 5000 6000 180 160W 140W 120W 100W 80W Longitude latitudes in [-1.0, 1.0] - (7 points) (): Min= -0.01, Max= 0.39, Int= 0.10-0.55-0.33-0.11 0.11 0.33 0.55 ocean temperature increment Patrick Laloyaux (ECMWF) CERA November 19, 2014 7 / 15
Model level Assimilation in the CERA system CERA system Sequential approach: A common 24-hour assimilation window observation misfits computed by the coupled model Increments computed separately Two outer iterations in 4DVAR and 3DVAR: ocean-atmosphere correlations in the assimilation cycle Time step : 12h atmospheric temperature increment -0.55-0.45-0.35-0.25-0.15-0.05 0.05 0.15 0.25 0.35 0.45 0.55 20 40 60 80 0 2010080109 tn Z-0.00000 195801 mean 20 40 Depth (m) 60 80 100 2000 3000 4000 5000 6000 180 160W 140W 120W 100W 80W Longitude latitudes in [-1.0, 1.0] - (7 points) (): Min= -0.00, Max= 0.39, Int= 0.10-0.55-0.33-0.11 0.11 0.33 0.55 ocean temperature increment Patrick Laloyaux (ECMWF) CERA November 19, 2014 7 / 15
Model level Assimilation in the CERA system CERA system Sequential approach: A common 24-hour assimilation window observation misfits computed by the coupled model Increments computed separately Two outer iterations in 4DVAR and 3DVAR: ocean-atmosphere correlations in the assimilation cycle Time step : 18h atmospheric temperature increment -0.55-0.45-0.35-0.25-0.15-0.05 0.05 0.15 0.25 0.35 0.45 0.55 20 40 60 80 0 2010080115 tn Z-0.00000 195801 mean 20 40 Depth (m) 60 80 100 2000 3000 4000 5000 6000 180 160W 140W 120W 100W 80W Longitude latitudes in [-1.0, 1.0] - (7 points) (): Min= -0.00, Max= 0.38, Int= 0.10-0.55-0.33-0.11 0.11 0.33 0.55 ocean temperature increment Patrick Laloyaux (ECMWF) CERA November 19, 2014 7 / 15
Model level Assimilation in the CERA system CERA system Sequential approach: A common 24-hour assimilation window observation misfits computed by the coupled model Increments computed separately Two outer iterations in 4DVAR and 3DVAR: ocean-atmosphere correlations in the assimilation cycle Time step : 24h atmospheric temperature increment -0.55-0.45-0.35-0.25-0.15-0.05 0.05 0.15 0.25 0.35 0.45 0.55 20 40 60 80 0 2010080121 tn Z-0.00000 195801 mean 20 40 Depth (m) 60 80 100 2000 3000 4000 5000 6000 180 160W 140W 120W 100W 80W Longitude latitudes in [-1.0, 1.0] - (7 points) (): Min= -0.01, Max= 0.33, Int= 0.10-0.55-0.33-0.11 0.11 0.33 0.55 ocean temperature increment Patrick Laloyaux (ECMWF) CERA November 19, 2014 7 / 15
Table of Contents 1 Operational ECMWF reanalysis system 2 Coupled earth reanalysis system 3 Assessment of the performance 4 Test run for CERA20C 5 Conclusions and future steps Patrick Laloyaux (ECMWF) CERA November 19, 2014 7 / 15
Performance of the CERA system in the Tropics for September 2010 CERA system & Uncoupled assimilation system Same resolution and code version Temperature analysis RMSE (dashed lines) Temperature background RMSE (solid lines) Improved fit to radiosonde observations (selected above ocean and coast) 13321 (6) -10 1 24591 (11) Ocean depth (m) -10 2 25582 (11) 22360 (7) 22995 (11) 19330 (4) 26420 (22) 25574 (22) 17898 (12) 31495 (18) 27561 (21) -10 3 21714 (18) 19440 (15) 14072 (21) 5682 (8) Improved fit to buoy and probe observations 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Patrick Laloyaux (ECMWF) CERA November 19, 2014 8 / 15
Use of near-surface observations in the CERA system Impact of scatterometer data (10-meter wind observations) during the typhoon Phailin The second-strongest tropical cyclone in India formed: 4 October 2013 dissipated: 14 October 2013 damage: $696 million fatalities: 45 CERA analysis for mean sea level pressure over the bay of Bengal 09 October 2013 10 October 2013 11 October 2013 12 October 2013 Patrick Laloyaux (ECMWF) CERA November 19, 2014 9 / 15
Scatterometer data on the 11 November 2013 Ascending pass on the 11 October 2013 ASCAT-A ASCAT-B OSCAT Wind observations assimilated by the CERA system Wind observations rejected (first guess departure too large) Patrick Laloyaux (ECMWF) CERA November 19, 2014 10 / 15
Impact of scatterometer data on 10-meter winds 6 observations per day: wind speed wind direction Observations 16 14 12 10 8 6 4 2 0 01 Oct 2013 08 Oct 2013 15 Oct 2013 10-meter wind speed Patrick Laloyaux (ECMWF) CERA November 19, 2014 11 / 15
Impact of scatterometer data on 10-meter winds 6 observations per day: wind speed wind direction Observations CERA CERA/NOSCATT 16 14 12 10 8 6 4 2 0 01 Oct 2013 08 Oct 2013 15 Oct 2013 10-meter wind speed CERA system Patrick Laloyaux (ECMWF) CERA November 19, 2014 11 / 15
Impact of scatterometer data on 10-meter winds 6 observations per day: wind speed wind direction Observations CERA CERA/NOSCATT 16 14 12 10 8 6 4 2 0 01 Oct 2013 08 Oct 2013 15 Oct 2013 10-meter wind speed CERA system Observations UNCPL UNCPL/NOSCATT 16 14 12 10 8 6 4 2 0 01 Oct 2013 08 Oct 2013 15 Oct 2013 10-meter wind speed Uncoupled assimilation system Positive and similar impacts of scatterometer data on atmospheric 10-meter winds Patrick Laloyaux (ECMWF) CERA November 19, 2014 11 / 15
Cold wake 27.75 28 28.25 28.5 28.75 29 29.25 29.5 29.75 30 CERA analysis of the Sea Surface Temperature 27.75 28 28.25 28.5 28.75 29 29.25 29.5 29.75 30 27.75 28 28.25 28.5 28.75 29 29.25 29.5 29.75 30 27.75 28 28.25 28.5 28.75 29 29.25 29.5 29.75 30 10 October 2013 11 October 2013 12 October 2013 13 October 2013 The CERA system (1 ) produces the cold wake after the tropical cyclone Patrick Laloyaux (ECMWF) CERA November 19, 2014 12 / 15
Impact of scatterometer data on ocean temperature at 40-meter depth From a 10-day frequency (0-2000m) to a 3-hour frequency (0-300m) Observations 29 28 27 26 Thu10 Fri11 Sat12 Sun13 Mon14 Tue15 Patrick Laloyaux (ECMWF) CERA November 19, 2014 13 / 15
Impact of scatterometer data on ocean temperature at 40-meter depth From a 10-day frequency (0-2000m) to a 3-hour frequency (0-300m) Observations CERA CERA/NOSCATT 29 28 27 26 Thu10 Fri11 Sat12 Sun13 Mon14 Tue15 CERA system Patrick Laloyaux (ECMWF) CERA November 19, 2014 13 / 15
Impact of scatterometer data on ocean temperature at 40-meter depth From a 10-day frequency (0-2000m) to a 3-hour frequency (0-300m) Observations CERA CERA/NOSCATT Observations UNCPL UNCPL/NOSCATT 29 29 28 28 27 27 26 26 Thu10 Fri11 Sat12 Sun13 Mon14 Tue15 CERA system Thu10 Fri11 Sat12 Sun13 Mon14 Tue15 Uncoupled assimilation system Positive impact of scatterometer data at 40-meter depth in the CERA system......but not perfect (10-meter layers, nudge to a daily SST product) Patrick Laloyaux (ECMWF) CERA November 19, 2014 13 / 15
Table of Contents 1 Operational ECMWF reanalysis system 2 Coupled earth reanalysis system 3 Assessment of the performance 4 Test run for CERA20C 5 Conclusions and future steps Patrick Laloyaux (ECMWF) CERA November 19, 2014 13 / 15
CERA20C - Extended Climate Reanalysis Configuration: 1 CERA system assimilation of 10-meter winds, mslp, temperature and salinity profiles HadISST2 for the SST nudging from 1900 to 2010 Test run over 2 years (2009-2010) providing forcing fields for offline CARBON reanalyses Tools to monitor the CERA20C production (e.g. mslp statistics from buoys) Observation statistics 300 250 200 150 100 50 0 50 100 mean(fg_depar) mean(an_depar) rms(fg_depar) rms(an_depar) Number of observations 16000 15000 14000 13000 12000 1 0 9000 8000 7000 Jan 2009 Feb 2009 Mar 2009 Apr 2009 May 2009 Jun 2009 Jul 2009 Aug 2009 Sep 2009 Oct 2009 Nov 2009 Dec 2009 Jan 2010 Feb 2010 Mar 2010 Apr 2010 May 2010 Jun 2010 Jul 2010 Aug 2010 Sep 2010 Oct 2010 Nov 2010 Dec 2010 Patrick Laloyaux (ECMWF) CERA November 19, 2014 14 / 15
Table of Contents 1 Operational ECMWF reanalysis system 2 Coupled earth reanalysis system 3 Assessment of the performance 4 Test run for CERA20C 5 Conclusions and future steps Patrick Laloyaux (ECMWF) CERA November 19, 2014 14 / 15
Conclusions and future steps The CERA system: improves the fit to atmospheric and ocean temperature observations in the Tropics represents better the air-sea interaction during a typhoon encouraging results for the production of coupled reanalysis Future steps: Cray migration Upgrade IFS to cycle 41R1 Increase ocean vertical resolution Start production of CERA20C Patrick Laloyaux (ECMWF) CERA November 19, 2014 15 / 15