An Observing System for Canadian Arctic Ice Caps

1
An Observing System for Canadian Arctic Ice Caps

• Martin Sharp
• Earth and Atmospheric Sciences
• University of Alberta

2
Queen Elizabeth Islands 110,000 km2 ice
Grant
Agassiz
Muller
Likely to respond more quickly than Greenland
ice sheet (response times on order of centuries
to a millennium)
Prince of Wales
Steacie
Sydkap
Manson
Behaviour very poorly known
Devon
3

• The Arctic climate is changing and is expected to
  continue to change
  

Mean Annual Air Temperature 2060-2089 c.f.
1961-1990 Source Polar Research Group, U.Illinoi
s
Arctic Annual Temperature Trends 1954-2003
Source J.Walsh
4
The Problem

• Is the area and volume of Canadas Arctic
  glaciers and ice caps responding to these climate
  changes - and how rapidly ?
  
• What is the contribution of these changes to
  global sea level?
  
• How important are the different mass loss
  processes - surface melt and iceberg calving?

5
An Observing System

• Gradual development since 1994
• Modeled on PARCA - mix of remote sensing
  (airborne satellite), in situ observations and
  modeling
  
• Done with multiple increments of short term
  funding and few people
  
• Key differences from Greenland - scale of ice
  masses, max elevations (minimal dry snow zone),
  extensive surface melt, more complex topography
  

6
What do we need to know?

• Ice extent - multiple epochs
• Surface elevation - multiple epochs
• Ice thickness and bed topography
• Surface velocity (and its temporal variability)
• Surface mass balance (including inter-annual
  variability)
  
• Iceberg calving flux

7
Ice Extent

• Data sources 160k aerial photography (1959/60)
  and derived 1250k NTS maps (digital) Landsat7
  ETM and ASTER (1999-2004)
  
• Alternatives DISP (1960s), MODIS, LandSat
• Accuracy (line placement due to digitizing, snow,
  cloud, shadow masking) AP 5-75m ETM (15-120m)
  
• Problems
• N. Ellesmere not covered in recent period
• ice outline errors in NTS maps
• limited ground control for AP
• lack of knowledge of short term variability in
  calving front positions

8
The Whole Picture
1960-1993 only
9
Surface Elevation (1)

• Data sources
• CDED (based on 1250k NTS sheets and 1959/60 AP)
  
  
• 1995 2000 airborne data (Abdalati, Dowdeswell)
  
  
• GLAS (post 2003)
• Alternatives kinematic GPS from traverses,
  InSAR, ASTER
  
• Vertical Accuracy
• CDED - /-20-50m
• Abdalati (ATM)
• Dowdeswell (radar terrain clearance and aircraft
  GPS) /- 7m
  
• GLAS
• InSAR 30-40m

10
Surface Elevation (2)

• Issues
• Poor quality of 1959/60 data due to limited
  ground control and poor contrast above snowline
  
• limited airborne coverage (ATM often misses major
  outlet glaciers)
  
• spacing of GLAS orbits and limited of crossing
  points per ice cap (esp. in S)
  
• InSAR DEMs pick up subtle detail but have poor
  absolute accuracy
  
• only possible to detect large dynamically driven
  changes
  
• Needs improved ground control to redo
  photogrammetry (potential accuracy /- 5m) and
  improve absolute accuracy of InSAR DEMs more
  extensive ATM coverage

11
IceSat Orbits Oct/Nov 2003
Devon Ice Cap CDED - GLAS 2003 NB Only 16 orbit
crossing points
12
Ice Thickness and Bed Topography

• Data Sources Dowdeswell (2000) airborne RES
  (Devon, Manson, POW, Agassiz)
  
• Alternatives Older data (Robin, Clarke,
  Koerner) limited U.Kansas data (1995) flow
  model inversion
  
• Accuracy 8m at crossing points
• Issues
• No recent coverage for Axel, N.Ellesmere, Sydkap,
  smaller ice masses
  
• Limited to along flow line on outlet glaciers
• Only Devon has dense (10km) grid

13
Other approaches inversion of numerical flow
models ?

14
Surface Velocity

• Data sources ERS 1/2, RadarSat (InSAR/speckle
  tracking) Landsat 7 ETM, ASTER, AP (feature
  tracking/IMCORR)
  
• Alternatives balance velocities, static GPS,
  optical surveying
  
• Accuracy InSAR - 10 on outlets Speckle tracking - limited validation data) ImCorr - 1 pixel
  rectification error
• Issues
• InSAR - accuracy of external DEMs (CDED)
• poor coherence (time period between images, high
  outlet velocities)
  
• lack of ascending and descending orbit coverage
  (need to project look-direction velocities)
  
• seasonal variability unknown due to lack of
  summer coverage
  
• lack of time series
• lack of data for balance velocity calculations

15
Surface Velocities from Interferometry and Spec
kle Tracking
Balance Velocities
Devon Ice Cap Flow Field
16
Surface Mass Balance

• Approach combine ice core analyses and
  modelling
  
• Problems
• upscaling/downscaling climate input data
  (ppt/temp)
  
• lack of on-ice temperature data
• poor knowledge of accumulation rate fields
• lack of validation data for models
• Issues Coarse resolution of passive microwave
  data limits use for melt detection and
  accumulation measurement on ice caps (lack of dry
  snow zone is also problematic for accumulation
  measurement)
• Possibilities QuikScat for facies mapping, melt
  detection and possibly melt intensity measurement
  (critical as inter-annual MB variability is
  mainly due to summer balance variability)

17
Lapse Rate Issues
John Evans Glacier, Ellesmere Island
Evolution of surface air temperature field during
2002 melt season
18
Air temperature versus QuikScat backscatter
Station at 1300m, Leffert Glacier,
Prince of Wales Icefield,
Ellesmere Island
19
QuikScat - 2000-2004 Melt Duration Climatology
20
2001 Warm year
2002 Cold Year
Lapse rate implications?
21
Iceberg Calving Flux

• Approach compute from surface velocity and ice
  thickness at grounding line
  
• Problems
• poor InSAR coherence due to fast flow near
  grounding lines
  
• patchy data from ImCorr/feature tracking
• poor knowledge of temporal variability in
  velocity
  
• lack of cross-glacier thickness data near
  grounding line
  
• floating tongues not well known (bottom melt
  issue)

22
Primary Needs (1)

• Improved ground control for topographic mapping
• Continued high resolution visible imagery for
  extent (Quickbird/Ikonos acquisition?)
  
• High accuracy surface topography for inversion
  modelling, balance velocities, volume-area
  scaling
  
• New thickness datasets, especially for Axel
  Heiberg and N.Ellesmere, and across outlet
  glacier grounding lines
  
• Targeted interferometry mission plus year-round
  GPS measurements on outlet glaciers
  
• Identify grounding lines and floating tongues

23
Primary Needs (2)

• Extended net accumulation dataset from shallow
  ice cores
  
• compute balance velocity fields
• develop climatology
• understand patterns of inter-annual variability
• Explore potential of RCMs/GCMs for accumulation
  simulation
  
• Extend air temperature and summer melt
  measurements
  
• ground truth active microwave data
• understand lapse rate variability
• model validation
• Continued collection of high resolution active
  microwave data