The Need for SPIRIT DEMs to Quantify Antarctic Ice Sheet Discharge Robert Bindschadler NASA & UMBC

Antarctic Ice Sheet Discharge • 33 largest, most active basins account for 840 Gt/a (Rignot and Thomas, 2002) • This equals less than half of total accumulation flux (1811 Gt/a) • Sparse ice thickness data at glacier grounding lines • Coastal basins – Smaller areal percentage (~25%) – High accumulation – Most of the perimeter

(adapted from Rignot and Thomas, 2002)

Antarctic Surface Accumulation and Ice Discharge: ASAID (IPY Project #88) In a nutshell: • Map grounding line from satellite imagery • Determine elevation along grounding line • Convert elevation to ice thickness • Apply velocity to calculate discharge

A slightly larger nutshell: • Map grounding line from satellite imagery • Map hydrostatic line from grounding line and hydrostatic points • Determine elevation along grounding line • Compare all available DEMs to choose “best” elevation at each point • Convert elevation to ice thickness • Apply velocity to calculate discharge

Ice Sheet Margin Tidal flexure begins Break in surface slope

Grounding line from Landsat imagery Full tidal range Hydrostatic line from ICESat profiles

(Fricker and Padman, 2006)

Kelly Brunt has provided F, I and H points around the Antarctic perimeter

New ASAID Grounding Line of Antarctica

9 contributions from 7 countries involving 20 scientists 3,574,365 points @15 m resolution

53,609 km long

Hydrostatic Line in process

Elevation Determination • Customized software to ingest all available DEMs • Shaded relief representations of DEMs to judge regional strength – Compared to Landsat image

• Display elevation profiles along grounding line – Includes crossing ICESat elevation as “truth”

• Select “best” elevation from candidate DEMs over grounding line segment – Assign confidence rating (1 to 5) – Single operator (for consistency)-Amy Wichlacz

Photoclinometry

--- ASAID --- MOA

RAMP

--- ASAID

Altimetry

--- ASAID

X ICESat --- ASTER --- Triangulation --- Photoclinometry --- Altimetry --- RAMP

50m 5 km

Photoclinometry

--- ASAID --- MOA

Altimetry

RAMP --- ASAID

--- ASAID

X ICESat --- ASTER --- Triangulation --- Photoclinometry --- Altimetry --- RAMP

50m

5 km

Photoclinometry --- ASAID --- MOA

Altimetry

RAMP --- ASAID

--- ASAID

X ICESat --- ASTER --- Triangulation --- Photoclinometry --- Altimetry --- RAMP

X ICESat --- ASTER --- Triangulation --- Photoclinometry --- Altimetry --- RAMP

Lessons Learned • DEMs vary A LOT at the coast • Altimetry and RAMP tend to overestimate elevation • Photoclinometry is best (when it works) – Crevasses, open water, other albedo changes violate photoclinometry assumption of constant albedo

• Photogrammetry is a very strong method – Especially where photoclinometry is poorest – ASTER mask causes frequent gaps in coastal elevations

What about SPIRIT DEMs?

Landsat

X ICESat --- SPIRIT --- ASTER --- Ocean

SPIRIT and ASTER elevations along grounding line are very similar SPIRIT DEM

SPIRIT DEMs add detail and fill gaps with higher quality elevations Landsat

SPIRIT DEM

Photogrammetric DEMs SPIRIT coverage is severely limited in Antarctica—BUT DATA EXIST! ASTER G-DEM masked out many critical coastal—AND RECOVERY OF ELEVATIONS IS NOT POSSIBLE!

Solutions • The most significant improvement to Antarctic ice discharge calculations is improved elevations at the grounding line and hydrostatic line – ASTER G-DEM version 2 is expected in 2011, but it is not clear they will apply a more appropriate mask – SPIRIT has the data and could produce the necessary DEMs

Conclusions • Antarctic ice discharge calculations are important but remain incomplete • Methods and data exist to improve discharge calculations • Completing SPIRIT DEMs of the Antarctic coast will have immediate impact on improving discharge calculations

Photoclinometry • Assumes a diffusive surface of constant albedo θ surface

DN = A cosθ + B image brightness

scaling coefficient

surface slope

scattering

Antarctic Surface Accumulation and Ice Discharge: ASAID (IPY Project #88) • Benchmark data sets – – – – –

Grounding line position Surface Elevation in vicinity of GL Ice Thickness seaward of GL Surface Velocity in vicinity of GL Discharge Flux across GL

• Utilize new LIMA imagery • Train young scientists in analyzing multiple satellite data sets • Engage international partners • Release data (and supporting documentation) to data centers for unrestricted use

Revised Approach 1.

Delineate grounding line –

Visual inspection of Landsat imagery (15 m) • • •

Compare to MODIS-derived grounding line (Scambos et al., 2007) Compare to ICESat elevation profiles Compare to key tidal flexure points (provided by Kelly Brunt)

1a. Determine “hydrostatic line” by tidal flexure in both ICESat and InSAR 2. Determine elevation in vicinity of grounding line –

Examine all available DEMs • • • • • •

3.

Convert elevation to ice thickness – –

4.

Photoclinometry (uses ICESat elevation control and Landsat image brightness) RAMP Altimetry (Bamber’s blend of ICESat and radar altimetry) ASTER G-DEM SPIRIT Triangular interpolation

Apply hydrostatic equilibrium condition Fails to include zone of possibly intense basal melt between grounding and hydrostatic lines

Multiply ice thickness by column-averaged velocity – –

Velocities derived from InSAR Surface velocity equals column-average for floating ice