Analysis of ICESat Data in East Antarctica using Kriging and Kalman Filter

1
Analysis of ICESat Data in East Antarctica using
Kriging and Kalman Filter
Motivation detect height change surface
characteristics

• An T Nguyen (MIT)
• Thomas A Herring (MIT)

Special thanks to Dr. Zwally and H. Cornejo
2
Introduction

• ICESat
• first laser satellite to study ice sheets
• high accuracy (lt 20cm)
• high spatial resolution (172m spacing)

3
Region of study
4
Kriging / Kalman filter
5
ICESat Data
GLA06 Laser 2a release 24 Laser 3a
release 23 Data editing 1) Saturation
correction 2) Within 0.29o, 0.34o
pointing 3) Gain 13,100 4) Single profile
editing
6
ICESat Data (cont)
7
Preliminary resultsa) dh/dt

• Pointing errors
• ? gives s 7cm/yr
• ? dh/dt results inconclusive at this time.

8
Validation of the technique
9
Results (contd)b) surface features
10
Surface characteristics (contd)
11
Summary

• Kriging/Kalman filter results
• 5cm/yr to 9cm/yr in East Antarctica
• consistent with cross-over analysis
• 5-km DEM removes long wavelength features (gt
  5km)
• Residual analysis structures at shorter
  wavelengths
• time-correlated noise process s12 , s22 , s32,
  b1, b2, wo
• correlation lengths roughness from bs
• instrument noise level roughness, s32
• pointing errors still dominate, 7cm/yr
• in progress to model pointing biases

12
BLANK