Sea Ice Exchange across the Canadian Arctic Archipelago from Enhanced AMSR-E Imagery

1
Posters on Applications of AMSR-E

• 1) Canadian Airborne and Microwave Radiometer and
  Snow Survey campaigns in Support of International
  Polar Year.
• 2) New sea ice algorithm
• Does not use a single tie point to represent a
  given ice surface but uses a distribution for
  values. It then solves the radiometric equations
  1000 times with random values from each
  distribution.
• Using an optimization procedure to solve the
  radiometric equations to estimate sea ice
  concentration rather than solving the equations
  deterministically.
• Generic in that can be applied to any combination
  of remote sensing data (combine scatterometer and
  passive microwave).
• 3) Canadian Operational Ice Services Satellite
  data assimilation Program

2
Sea Ice Exchange between the Canadian
Archipelago, the Arctic Ocean and Baffin Bay
using Enhanced AMSR-E Imagery

• Tom Agnew1, Andrew Lambe2, Linda Enciu2 and David
  Long3,
• 1 Environment Canada, Toronto, Canada
• 2University of Toronto, Toronto, Canada
• 3Brigham-Young University, U.S.A.

3
Freshwater Return via Fram Strait and the
Archipelago
North Pacific

• Fram Strait FW
• 2300 km3 liquid
• 2400 km3 sea ice

• Archipelago FW ?
• 3200 km3 liquid

• sea ice ? small

Fram Strait
4
Spatially Enhanced AMSR-E

• Daily AMSR-E Tb products use drop-in-the-bucket
  method of combining swath data into a daily
  average product projected on the polar grid.
• Spatial image enhancement developed by Prof.
  David Long at BYU called Scatterometer Image
  Reconstruction (SIR). The method uses knowledge
  of the scan geometry and the antenna pattern of
  the sensor to increase the spatial resolution of
  the data and multiple estimates of Tb from
  different orbits to reduce pixel noise.
• AMSR-E resolution
• - 89 GHz 3 km compared to 6 km
• - Similar percentage improvements in the other
  channels
• - Available at (ftp//ftp.scp.byu.edu/pub/amsre)
• - Data has been processed from June 2002 to end
  of 2007
• By processing only the ascending orbits (daytime)
  and descending orbits (nighttime) get two
  products each day. Each pixel in the image is
  valid for a specific local time with a
  discontinuity along the 180 longitude.

5
89 GHz False Color Image (RH, GH, BV)
AMSR-E 89 GHz drop-in-the bucket Resolution 6
km. Each pixel is a daily average.
Enhanced AMSR-E 89 GHz Resolution 3 km. Each
pixel is valid for a specific time.
6
Animation of Enhanced AMSR Imagery
7
Sea ice Transport in the Archipelago
8
Estimating Sea Ice Motion from AMSR-E

• Maximum cross correlation (MCC) between pairs of
  satellite images is used to estimate daily sea
  ice motion. From this enhanced imagery we can get
  an independent estimate of ice motion every 15
  km.
• By combining sea ice motion and sea ice
  concentration we can get an estimate of ice area
  flux.
• Analyzed 5 years from September 2002 to June 2007
  however because of increased atmospheric moisture
  in the summer months we cannot get ice motions in
  July and August

9
AMSR Sea ice Motion using MCC
January 7-8, 2003
10
Fluxgates surrounding the Archipelago
AG
MS
JS
LS
Flux in the unit normal direction is negative
11
Daily ice area flux through AG and MS gates
flux into Archipelago - flux into Arctic
Ocean
flux into Archipelago - flux into Arctic
Ocean
Negative flux mean export into the Arctic Ocean
12
Daily ice area flux through QEI-S and N gates
Negative flux means export into the Arctic Ocean
13
Daily ice area flux through LS and JS gates
flux into Baffin Bay - flux into
Archipelago
Positive flux means export into Baffin Bay
14
Total 10-month Area Fluxes (103 km2) each year
Year AG MS QEI-S QEI-N LS JS
2002/03 -77 -39 7 -5 63 -1
2003/04 -29 1 -4 -6 79 3
2004/05 -31 -10 -10 8 58 1
2005/06 -20 -12 2 8 63 0
2006/07 -53 -1 4 -1 77 -2
5-year average -42 -12 0 0 68 0
Estimated ice thickness (m) 1 2.5 3.4 3.4 1.5 1.5
Volume flux (km3) -42 -30 0 0 102 0
15
Yearly Sea Ice Production and Export in the
Canadian Archipelago

• The Canadian Archipelago is a region
• of net sea ice production and export into
• Both the Arctic Ocean and Baffin Bay.
• Over the 2002 to 2007 period the
• Archipelago produced and export
• 122,000 km2/yr of sea ice area (174 km3)
• of sea ice.

Export into Baffin Bay
Export into Arctic Ocean
16
Conclusions

• Enhanced AMSR-E imagery can be used to estimate
  sea ice motion in the main channels of the
  Archipelago. Increased atmospheric absorption
  prevents estimation in July and August.
• The largest flux variability and largest fluxes
  occur through Amundsen and MClure Strait which
  exports ice the Arctic Ocean and Lancaster Sound
  with a net export of ice into Baffin Bay. Daily
  area fluxes can be as large as /- 2500 km2.
• For this 5-year period, 54000 km2 or 72 km3 of
  sea ice is exported into the Arctic Ocean each
  year and 68000 km2 or 102 km3 into Baffin Bay.
  This sea ice is generated within the Archipelago
  itself mainly from the system of stationary and
  transient polynyas which form each winter. There
  is very little direct transport of sea ice from
  the Arctic Ocean through the Archipelago to
  Baffin Bay.
• These ice fluxes are considerably less than the
  flux of sea ice through Fram Strait ( 2800 km3
  yr-1 ), the southward ice area transport Baffin
  Bay
• 530 x 103 km2 .

17
Conclusions

• Future work analyzed enhanced SSM/I imagery
  (from 1988) estimate ice area fluxes for a much
  longer period.

18
THANK YOU
Acknowledgements Canadian IPY US
NSF/SEARCH NSIDC IABP
19
AMSR Sea ice Motion using MCC
March 3-4, 2003
20
AMSR Sea ice Motion using MCC
March 3-4, 2003
21
Local Pixel Time for Nighttime Orbits
22
Animation of Enhanced AMSR Imagery
23
Advanced Microwave Scanning Radiometer for EOS
(AMSR-E)

• For Arctic monitoring and research, satellite
  microwave has an advantage over other satellite
  sensors because it can see through cloud and
  during 24-hour darkness.
• AMSR-E is a passive sensor with resolution has
  been improved to 3 km and it has daily repeat
  coverage of the entire Arctic. From it we can get
  daily sea ice motion and sea ice concentration
  estimates in the main channels of the Canadian
  Archipelago.
• Maximum cross correlation (MCC) between pairs of
  satellite images is used to estimate daily sea
  ice motion. From this enhansed imagery we can get
  an independent estimate of ice motion every 15
  km.
• By combining sea ice motion and sea ice
  concentration we can get an estimate of ice area
  flux.
• Analyzed 5 years from September 2002 to June 2007
  however because of increased atmospheric moisture
  in the summer months we cannot get ice motions in
  July and August

24
Cross gradient pressure /ice flux relationship