Prsentation PowerPoint

1
L2P products in MEDSPIRATION   Pierre Le Borgne,
Gérard Legendre, Météo-France/CMS,
Lannion Sophie Casitas, Alain Coat Actimar, Brest

• Content
• Examples
• Comments on content
• Production chain
• Comments on production
• Conclusions

2
Medspiration L2P products

• PRODUCT (from ) Producer (over ) Coverage
• AMSRE REMSS Atlantic
• TMI REMSS Atlantic
• AATSR ESA/SOC Atlantic
• AVHRR_G (N-16 -17) NAVOCEANO/Podaac Atlantic
• AVHRR_L (N-16 -17) NAVOCEANO /Podaac Medit.
• NAR (N-16 -17) EUMETSAT/OSI SAF Medit.
• SEVIRI EUMETSAT/OSI SAF Atlantic

3
Medspiration zones
Atlantic
Mediterranean
4
Content
Main SST information SST, SST_dtime SSESD,
SSESs DT_analysis Proximity confidence

• Coordinates
• Regular (SEVIRI, AMSRE, TMI)
• Non regular swath (ATSR)
• Non regular projection (NAR)
• None (AVHRR_G, AVHRR_L)

Flags Rejection , confidence
Ancillary data SSI, wind_speed, Sea_Ice_fraction,
AOD sources Dtime_from_sst
5
Example 1/3
SSESD, SSESs, Proximity confidence
DT_analysis
6
Example 2/3
SST
Rej. Bit 6 (land)
Rej. Bit 2 (IR cloudy)
Rej. Bit 7 (spare)
7
Example 3/3
Conf. Bit 0 (sidelobe)
Conf. Bit 1 (rain cont.)
Conf. Bit 3 (high wind)
8
Ancillary aerosols (mask or flag?)
AOD in Jan. 05 (Courtesy of S. Good, Univ.
Leicester)
TMI-SEVIRI in Jan.05 (Courtesy of S. Good, Univ.
Leicester)
(Courtesy of R. Legeckis, NOAA)
9
Ancillary wind
Wind
SST CMS last analysis
10
Comments on Content
SST, SST_dtime Mask rejection flag (producer
decision) Common approach?
Complementary fields should answer the user
questions (IR) - Accuracy? - Contaminated by
cloud? aerosols? diurnal warming?
But - Proximity confidence do not answer all
questions - SSESD, SSESs reflect departure from
definition
Then - Confidence flag (producer
expertise) -DT_analysis and ancillary value
(user expertise) ..are useful

• One month of beta testing is quite short
• Users please report any problem

11
L2P chain (1/3)Overview
12
Admiss. cutoff time (18 to 36h T1 0600 to 2400)
L2P chain (2/3)Ingestion
13
QC cutoff time (17h T1 0500)
Prod. cutoff time (18h T1 0600)
L2P chain (3/3)L2P production
14
Comments on Production

• Merging of data (SST ancillary) solved by
  cutoff times (absolute)
• the metrics should reflect that.
• There are too many sources of data (Last
  analysis Ancillary)
• 1)Native (when avail.) 2)Standard 3)Backup
• There should not be multiple L2P producers for
  one dataset
• L2P (core) at the (L2) producer level,
  DT_analysis and ancillary value at the GDAC level
  if necessary

15
Conclusions

• L2P not easy to produce
• additional work on the convergence of contents
  is needed
• Feedback from users (SOC, March 2005) quite
  positive

• Agenda (L2P in EURDAC)
• Summer 2005
• - AATSR L2P global and produced at Ifremer
• Last analysis from CMS climatology to FNMOC
  analysis
• Aerosols from NOAA to NAAPs (under
  investigation)
• Ice from Nise to OSI SAF
• Extension of the fine resolution area to
  European waters
• mid 2007
• Introduction of major modifications (if any)
• Production of Metop L2P

Producers are people!
16
Ancillary aerosols Testing sources
METOSAT-7 Colour composition
NAAPS
NESDIS