SST Data Continuity: The Impact of a data gap between AATSR and SLSTR

1
SST Data Continuity The Impact of a data gap
between AATSR and SLSTR

• Gary Corlett(1), Hugh Kelliher(2), David
  Llewellyn-Jones(1), Dave Poulter(3), John
  Remedios(1), Ian Robinson(3), Werenfrid Wimmer(3)

(1) University of Leicester (2) Space
ConneXions (3) National Oceanography Centre,
Southampton
2
Background

• SST is an Essential Climate Variable (ECV) of the
  Global Climate Observing System (GCOS)
• The (A)ATSR SST record has been defined as a
  Fundamental Climate Data Record (FCDR) within the
  GCOS Implementation Plan
• The GCOS Climate Monitoring Principles (GCMPs)
  include a requirement for an overlap between
  successive instruments in a series.
• AATSR is expected to operate until 2010, at which
  time Envisat is expected to run out of fuel but
  the Sea and Land Surface Temperature Radiometer
  (SLSTR) on Sentinel 3 will not fly before 2012,
  so there may be an enforced break in the SST
  climate record, with no overlap between
  instruments.
• This raises two fundamental SST data continuity
  questions
• Can the SST data from Sentinel 3 be tied to the
  same absolute temperature reference as the SST
  data record in the (A)ATSR archive?
• How might the data gap between ENVISAT and
  Sentinel 3 be filled using alternative sources of
  SST data with the minimum loss of quality?

3
GCOS GCMP

• GCOS defines 20 GCMP for satellite climate data
  records
• 10 for all data records
• 10 just for satellites
• ATSR-series adheres to many GCMP. Examples
  include
• On-board calibration
• Overlap period between sensors to ensure
  homogeneity
• Ground-based monitoring throughout lifetime of
  satellite
• Full list available in
• Implementation plan for the Global Observing
  System for Climate in Support of the UNFCCC,
  GCOS-92, October 2004 (WMO/TD No. 1219)

4
Aim and Strategy

• Aim
• To ensure that the SST data record provided by
  the (A)ATSR-SLSTR series continues to fulfil the
  needs of international Earth observation
  initiatives, including GCOS, GEOSS, GMES, GHRSST,
  CEOS/IVOS, IPCC and UNFCCC
• Strategy
• To assess how satellite and in situ data can best
  be used to tie the Sentinel 3 SST data to the
  AATSR SST data by identifying possible sources of
  SST data that might be available to bridge the
  gap between the end of the Envisat mission (or
  failure of AATSR) and the start of the Sentinel 3
  mission
• To devise a strategy for national and
  international agencies to respond to the
  challenge of maintaining the (A)ATSR series as a
  climate standard for SST
• The strategy will recommend a programme of work
  to be implemented before and during an interval
  when no ATSR-class sensor is operating, with the
  aim of securing the integrity of the SST climate
  record from 1991 to at least 2021

5
Alternate SST Data Sources

• To identify potential sources of SST data,
  especially in the 2010 to 2012 timeframe, when
  the gap is expected to occur in ATSR SST data
  record
• To look at both current and planned
  instrumentation, and their availability, through
  consulting within the GHRSST-PP community and
  mission scientists for planned instrumentation.
• Several different types of data sources,
  including
• Infrared satellite instruments
• Microwave satellite instruments
• In situ data
• To make recommendations as to the suitability of
  these alternatives for bridging the gap between
  AATSR and Sentinel 3

6
Possible Sources of Data (May 2007)

• There are two polar-orbiting IR satellite data
  sets that may be used to cover the period from
  2010-2012, AVHRR (on MetOp) and VIIRS on NPP
• The availability of the MODIS dataset through
  2010-2012 cannot be confirmed at this time
• It is noted that VIIRS will sample at 0530 and
  1330 and not at the 1000-1030 sampling time of
  the ATSRs
• There will be data from geostationary IR
  satellites such as SEVIRI, GOES and MTSAT
• It is likely that there could be only one MW SST
  data set available for the period 2010-2012,
  provided by the CMIS instrument on NPOESS
• The CMIS instrument will not sample at the same
  time as the ATSRs
• The availability of the AMSR-E dataset through
  2010-2012 cannot be confirmed at this time
• In situ radiometers

7
The Need for Long-term in situ Radiometer Data

• The importance of continuing the M-AERI and ISAR
  in situ data records is noted, as they may
  provide the key link from AATSR to Sentinel 3,
  and are required to ensure the ATSR data record
  continues to adhere to the GCMPs. Therefore, we
  recommend that the AATSR and Sentinel 3 funding
  bodies should provide a way to ensure the
  continuity of these datasets as a high priority.

8
Analyse Suitability of Alternative Sources of SST

• To use the GHRSTT-PP High Resolution Diagnostic
  Data Set (HR-DDS) system at NOCS to perform a
  global-wide comparison between AATSR data and
  other satellite derived SST products to identify
• The magnitude of the bias and uncertainties
• Their geographical and seasonal variability
• To examine how much deviation there is between
  SST recorded by the different satellite SST
  products, when sampled coincidentally at each of
  the HR-DDS sites
• To identify the scale of the mismatch between
  different sensors and AATSR, and the extent to
  which this varies with geographical location

9
Compare Validated Performance of Potential
Sources of SST Data with that of AATSR

• To employ the high quality in situ skin
  measurements, presently used to validate the
  AATSR products, as a means of determining the
  errors of the other satellite SST products
• Uses 1 year of in situ skin temperature
  measurements obtained from the ISAR instrument
  operated by NOCS on the PO car ferry Pride of
  Bilbao over the English Channel and Bay of Biscay
• To provide an objective comparison between the
  performance of the alternative sources of SST
  data and AATSR performance with reference to the
  in situ data
• To provide a first indication of the bias
  corrections that would have to be applied to the
  alternative SST data to make it consistent with
  AATSR-measured SST data in that region.

10
Validation with ISAR

• Validation area - match-ups for SEVIRI 2006

11
AMSR-E
Grade Match-up Criteria
1 Coincidence of ISAR and AATSR sample within 2000 s time window and 1 km search radius in space.
2a Temporal match within 2000 s and spatial match within 20 km
2b Temporal match within 2 hrs and spatial match within 1 km
3 Temporal match within 2 hrs and spatial match within 20 km.
12
AVHRR - NAR 18
13
SEVIRI
14
Summary of ISAR Validation

• In the ISAR validation area (English Channel and
  Bay of Biscay) the match-ups for the infrared
  sensors (SEVIRI and NAR17/18) yield very good
  agreement in the grade 1 and 2b cases. The
  microwave sensor (AMSR-E) does not produce
  similarly good match-ups statistics, although it
  is not clear whether that is due to the data
  product or because of sensor limitations close to
  coastal areas.
• The main reason for the good validation
  performance of the infrared sensors is a stable
  atmosphere in the ISAR validation area in 2006.
  Therefore the static single view atmospheric
  correction of theses sensors does almost as well
  as the dual view dynamic atmospheric correction
  of AATSR. However the comparison of the
  validation results of the different sensors will
  look very different in atmospheric conditions
  with a high aerosol count.

15
Analyse Long-term Variability of SST Data Quality

• To identify
• Regions of the world where alternative SST data
  sources have a quality comparable to AATSR,
• Where new validation activities will be required
  to ensure the best chance for an accurate tie
  between AATSR and Sentinel 3 SST data.
• Data sets analysed
• AMSR-E v5, AVHRR Pathfinder v5, MODIS Aqua, MODIS
  Terra, Reynolds OI v2

16
Analysis Areas
17
AVHRR PF v5 Mean Difference
18
AVHRR PF v5 Absolute Difference
19
AMSR-E v5 Mean Difference
20
AMSR-E v5 Absolute Difference
21
Formulate a Plan for Maintaining SST Data Quality

• To formulate a draft plan for discussion with
  Defra, NERC, ESA and GHRSTT-PP on what needs to
  be put in place to maintain the quality and
  continuity of the (A)ATSR data set using
  alternative sources of SST data where appropriate
• The plan will include
• An outline of the steps that need to be taken
• Recommendations for a programme of work that will
  be required
• Not yet complete but will be circulated to
  GHRSST-PP science team for comment

22
Latest News from ESA

• ESA are currently assessing options for extending
  lifetime of ENVISAT
• Summary presentation given in Montreux
• Large fuel requirement needed for deorbiting at
  EOL
• Get free station keeping by deorbiting prior to
  OCM
• Retain 5 min MLST of ascending node
• Could extend lifetime of ENVISAT up to 2 years
• Currently being assessed by ESA and instrument
  QWGs

23
Acknowledgements

• Robert Murphy
• Peter Minnett
• Vincent Salomonsen
• Chelle Gentemann
• Ian Barton
• Antony Rea
• BNSC