Validating Cryosat ice thickness data

1
Validating Cryosat ice thickness data
Seymour Laxon
2
CryoSat Mission Requirements
The CryoSat mission is intended to provide
measurements of mass and thickness fluctuations
of the Earths land and marine ice fields to the
wider scientific and applications community. This
wider community is concerned to receive
measurements that require no further processing
for their application and which are supported by
estimates of their uncertainty
3
CryoSat-1 October 2005
4
CryoSat-2 Launch 2009
5
Satellite Altimetry - Measurement Principle
Freeboard
6
Origin of Radar Altimeter Sea Ice Echoes

• Co-incident AATSR imagery reveals the origin of
  Diffuse and Specular echoes over sea ice
• Diffuse echoes originate from ice floes
• Specular echoes originate from leads
• Gaps are caused by Complex echoes which are
  excluded

7
Echo Origin Determination

• Diffuse echoes originate from the Open Ocean or
  Ice Floes with Sea Ice areas
• SSM/I ice concentration is used to discriminate
  diffuse ocean and ice returns
• Specular Echoes are assumed to originate from
  leads and thin ice

8
Repeat Profile Analysis

• Up to 60 repeat profiles are analysed along each
  of the 501 orbit tracks
• Ocean returns are used to construct a mean sea
  surface profile
• Residual height profiles are used to determine
  ice freeboard

9
Altimeter elevation profile
Sea level
Sea ice freeboard
10
Freeboard to Thickness Conversion

• Conversion assumes reflection from the ice/snow
  interface
• Conversion to thickness using climatology of snow
  depth/densities Warren, 1999

11
Submarine Validation 2000
Beaufort Sea October 1996
12
Kwok
Comparison of Submarine and Altimeter Thickness
PDF
13
Validation The need for averaging
14
Error Co-Variance
?geophys
10
Error in Ice Thickness (metres)
1.0
?noise
0.1
?bias
1s 1km
106s 102km
108s 104km
Time Space
15
Error Co-Variance
?geophys
10
Error in Ice Thickness (metres)
1.0
?noise
0.1
?bias
1s 1km
106s 102km
108s 104km
Time Space
16
Error Co-Variance
?geophys
10
Error in Ice Thickness (metres)
1.0
?noise
0.1
?bias
1s 1km
106s 102km
108s 104km
Time Space
17
Error Co-Variance
?geophys
10
Error in Ice Thickness (metres)
1.0
?noise
0.1
?bias
1s 1km
106s 102km
108s 104km
Time Space
18
Conceptual Experiment Design
Example Level 2 Sea ice geometric and
penetration model error

• Assess practicality and identify missing
  capability e.g. ASIRAS.
• Identify and contact important groups and
  planning time-scales e.g. Alfred Wegener
  Institute 2-3 year planning horizon for polar
  activity.
• Identify practical locations e.g. Arctic Ocean N.
  and W. of Greenland is accessible and gives
  access to strong ice concentration variations.
• Identify experimental complexity and novelty and
  assess need for pre-launch trials e.g. LARA
  (2002) and CryoVEx (2003) campaigns.
• Identify and implement requirements on
  ground-segment capability.

19
Autosub Under Ice Greenland 2004
Wadhams, et al., 2006
20
Conclusions

• Techniques to derive Cryosat sea ice measurements
  is identical to current altimeters (ERS/Envisat)
• (We dont know any different)
• Geophysical uncertainties will be the same
• SSH determination
• Ice/Water/Snow density
• Errors will depend on spatial/temporal averaging
• This is not SSMI!
• Design of averaging scheme application dependent
• There is still a lot of work to do
• Independent ( timely) validation data of great
  use

21
Issues for AUV

• Use of historical data
• We may be able to learn a lot from previous
  deployments
• Winter deployments
• Validation data are most useful during the winter
• AUV range
• Data are only valid some way (1-200km) into the
  ice pack
• AUV as part of an overall suite of measurements
• Best combined with in-situ and airborne data
• UCL is very happy to collaborate and participate
  in proposals and field work activities where
  useful

22
Envisat Near Real Time Ice Freeboard
5-7-April-2007
Ice Freeboard (cm)
23
Further Reading http//www.esa.int/esaLP/LPcryosa
t.html
24
First (Equal?) ERS-ULS Antarctic Thickness