15 Years of progress in Radar Altimetry Symposium Ocean surface topography science team (OSTST) International Doris Service (IDS) Workshop, Argo Workshop 13-18 March 2006, Venice, Italy Alexei V. Kouraev, Benoit Legrésy, Frédérique Remy Studies of Austfonna ice cap (Svalbard) using radar altimetry with other satellite techniques Laboratoire d'etudes en Geophysique et Oceanographie Spatiales (LEGOS), Toulouse France State Oceanography Institute, St. Petersburg branch, St. Petersburg, Russia
Austfonna - largest ice cap in the Eurasian Arctic AUSTFONNA Austfonna - 8120 km 2, Vestfonna - 2500 km 2 Etonbreen 0 m 767 m Digital Elevation Model (DEM), courtesy of the Norwegian Polar Institute (NPI), Tromsoe, Norway
ETONBREEN: RECENT CHANGES FROM SPOT 1 km Etonbreen + Basin 03: continuous rapid retreat 1-1987 April 15-21 2-1988 July 28 3-1991 July 18-29 4-1993 July 19 5-1998 March 28 1 2 3 4 5 Etonbreen DATA: SPOT 1-4 imagery for 1987, 1988, 1991, 1993, 1998 from ISIS project 0506-778. Period Change, km2 Rate, km2/yr 15/04/1987-28/07/1988-0.75-0.58 28/07/1988-29/07/1991-3.04-1.01 29/07/1991-19/07/1993-1.42-0.72 19/07/1993-28/03/1998-2.25-0.48 Basin 03
DATA: ENVISAT RA2 data for cycles 10 to 40 (Oct 2002 - Aug 2005). ICE2 retracker. Data quality control: Non-blank values of backscatter and range in Ku band, heights more than -1 m above the geoid. Number of observations 25-31 20-24 15-19 10-14 ENVISAT AUSTFONNA COVERAGE Eastern coast - few data (altimeter locked), central part - OK 4500 4000 3500 3000 2500 2000 1500 1000 500 0 Number of observations: monthly changes for cycles 10 to 36 1 2 3 4 5 6 7 8 9 10 11 12 Number of observation decreases in summer (melting)
Orbit changes across track are much larger than distance between 18 Hz measures (1500 m vs 400 m) CHOICE OF REFERENCE POINTS FOR TEMPORAL ANALYSIS Use of latitude or longitude for study of temporal changes is misleading 1500 1500 1500 1500 1500 1500 1500 1500 1500 m m m m m m m m m Solution: Choosing one cycle as reference and referring data for each cycle to the nearest reference point 400 m Location of 18 Hz ENVISAT points for cycles 10 to 36. Track 20.
Height: good relation to DEM, temporal variability low on top, high on slopes SPATIAL DISTRIBUTION: MEDIAN VALUES Backscatter: low values for ice, low temporal variability Haight (Ku) Backscatter (Ku)
Latitude 80.4 80.2 79.9 79.7 ENVISAT GR*1000 Monthly values GR= (TB365 - TB238) / (TB238 + TB365) JAN FEB MAR SNOW DEPTH GR ratio: linear relation to snow depth 79.4 79.2 80.4 80.2 APR MAY JUN ENVISAT: maximal values NE, GPR - NE and SE Latitude 79.9 79.7 79.4 79.2 80.4 JUL AUG SEP 50 40 30 20 10 NPI Ground Penetrating Radar (GPR) snow depth (2004) and interpolated map. Courtesy of A. Taurisano, NPI 80.2 0 Latitude 79.9 79.7 79.4-10 -30-50 79.2 80.4 OCT NOV DEC 80.2 Latitude 79.9 79.7 79.4 79.2 18 20 22 24 26 Longitude 18 20 22 24 26 Longitude 18 20 22 24 26 Longitude
D 2005.2 C B Backscatter (S) D C A B TEMPORAL VARIABILITY ALONG THE TRACK Backscatter: high values in summer near the ice edge (melting), low values in winter on top dtb - strong melting signal in summer A 2005.2 dtb (TB23.8-TB36.5) D C B A 2005.0 2005.0 2004.8 2004.6 2004.8 2004.6 SUMMER 30000 2004.4 12 2004.4 2000 1750 1500 2004.2 2004.0 8 4 2 2004.2 2004.0 1250 1000 800 600 400 200-1000 2003.8 2003.6 2003.4 2003.2 2003.0-480 -470-460 -450-440 -430-420 -410-400 -390-380 -370-360 0-2 -4-6 -8-12 -16 2003.8 2003.6 SUMMER 2003.4 2003.2 2003.0-480 -470-460 -450-440 -430-420 -410-400 -390-380 -370-360
D C B HEIGHT PROFILE - MEAN AND ANOMALIES A TOP A B C D A B C D Mean height (mm relative to reference ellipsoid) in Ku band for track 853. Opposite changes (swinging mode) before and after top of AID. No real seasonal signal, but something intriguing is evident... Anomalies in mm (from median values for cycles 11 to 31) of height in Ku band for track 853.
Artefact related to changes in orbit over topography! INFLUENCE OF ORBIT CHANGE ON ALTIMETRIC MEASURES) Anomalies of Height in Ku band, color scale is the same as previous slide D C B A
Sloping and hilly terrain: complicated radar echo SLOPE AND TOPOGRAPHY INFLUENCE Nadir line (shortest distance in theory) R A Shortest distance in reality Case 2: first echo comes not from the nadir point Simulation of radar return echo at various time steps for different topography
Overestimation on slopes SLOPE AND CURVATURE INFLUENCE Theoretical line (Measured = DEM) Good agreement, topography influence is low Altimeter locked Comparison of measured height in Ku band vs NPI DEM
Slope, cycle 18 SLOPE AND CURVATURE CORRECTIONS Altimeter X and Y position DEM DEM subset in +-20 km direction from X and Y Curvature, cycle 18 Approach: for a given DEM subset approximate (least square solution) Hdem=a*x2+ b*y2+ c*x*y + d*x + e*y + f Coefficients (a,b,c,d,e,f) Slope and curvature values for a given DEM subset
DEM SOURCE 2: SAR INTERFEROMETRY (1995-1996) Amplitude Coherence DEM Three ERS-1 and ERS-2 tandem pairs from SPRI November 1995, December 1995, January 1996
DEM SOURCE 3: SPOT STEREO Austfonna (Etonbreen region): stereo pairs for 1987 (SPOT1) and 1991 (SPOT2) from the ISIS project 0506-778
ITERATIVE APPROACH: SIMULATED AND MEASURED RA DATA DEM SIMULATOR ENVISAT Correcting errors due to DEM inclination plane SIMULATED WAVEFORM COMPARISON MEASURED WAVEFORM DEM CORRECTIONS QC, DATA SELECTION AND INTERPRETATION