5c9c63a04dc0d9a2f531343a9f5d1d6b.ppt
- Количество слайдов: 12
RMIB GERB Processing status S. Dewitte Royal Meteorological Institute of Belgium
Overview u Processing status. u Product quality status. u Data release plan. GIST, London, 14 -16/12/2005
Processing status u Reprocessing June, July, Dec 2004 for testing new spectral response: V 999, available for GIST members. u Operational processing restarted after eclipse safety period: V 2, close to V 999 (geolocation degraded). GIST, London, 14 -16/12/2005
Major changes compared to pre eclipse V 2 u New spectral response, calibration, unfiltering. u Activation improved LOS non repeatability correction (not in NRT V 2). u Activation full thermal ADM correction (addition hot desert) u Dedicated clear sky ocean unfiltering and desactivation aerosol correction. GIST, London, 14 -16/12/2005
Geolocation u V 999 quality visually satisfactory for resolution enhancement to 3 x 3 SEVIRI resolution. u Same quality is achievable in NRT (see presentation Luis Gonzalez). Ø OK for release GIST, London, 14 -16/12/2005
Spectral response & SW calibration u SW scene dependent differences with CERES have gone away in V 999. u GERB SW now about 6% higher then CERES. Ø Can be released with realistic error estimate. GIST, London, 14 -16/12/2005
Unfiltering u No show stoppers (see presentation Nicolas Clerbaux). Ø OK for release GIST, London, 14 -16/12/2005
Clouds u Degraded quality in sun glint and sunset region, inherent to use visible channels (see presentation Alessandro Ipe). Ø OK for release u Possible future improvement using infrared channels. GIST, London, 14 -16/12/2005
SW ocean asymmetry u Main identified cause: SEVIRI SW BB error propagates to small non linearity in GERB processing (see presentation Cedric Bertrand) Ø Solution possible for release GIST, London, 14 -16/12/2005
Aerosol u Include in GERB products: – Cloud/cirrus/dust information – Aerosol optical depth over ocean GIST, London, 14 -16/12/2005
HDF 5 update u Currently used version 1. 44 has bug. Ø Update to version 1. 6 Ø Users need to make update as well to be able to read new HDF files (old files will still be readable) GIST, London, 14 -16/12/2005
Data release plan u Approval of data release options at this meeting. u Need 2 months to buy new computers for NRT geolocation. u If new NRT geolocation operational: freezing of algorithm, use – for NRT + 40 day validation. – for reprocessing backwards in time. – Start GERB 2 commissioning. u Document algorithm and validation in series of reviewed papers. GIST, London, 14 -16/12/2005
5c9c63a04dc0d9a2f531343a9f5d1d6b.ppt