HR: 1330h
AN: C32A-0431    [PDF]
TI: ICESat Waveform Ground Processing Algorithm
AU: Roberts, L
EM: leeanne.roberts@gsfc.nasa.gov
AF: Raytheon ITSS NASA/Goddard Space Flight Center, Ocean and Ice Branch, Code 971, Greenbelt, MD 20771 United States
AU: Zwally, H
EM: zwally@icesat2.gsfc.nasa.gov
AF: NASA/Goddard Space Flight Center, Ocean and Ice Branch, Code 971, Greenbelt, MD 20771 United States
AU: * Brenner, A C
EM: anita@icesat2.gsfc.nasa.gov
AF: Raytheon ITSS NASA/Goddard Space Flight Center, Ocean and Ice Branch, Code 971, Greenbelt, MD 20771 United States
AU: Saba, J
EM: jack.saba@gsfc.nasa.gov
AF: Raytheon ITSS NASA/Goddard Space Flight Center, Ocean and Ice Branch, Code 971, Greenbelt, MD 20771 United States
AU: Yi, D
EM: donghui.yi@gsfc.nasa.gov
AF: Raytheon ITSS NASA/Goddard Space Flight Center, Ocean and Ice Branch, Code 971, Greenbelt, MD 20771 United States
AB: The shape of the ICESat laser-altimeter waveforms represents the interaction of the laser pulse with the surface-height distribution of the Earth's surface, which may be complex due to multiple reflecting surfaces of varying shapes within the laser footprint. Therefore, the ICESat waveforms are processed on the ground to determine the location on the waveform that represents the surface elevation and to derive characteristics of the variable surface height distributions. The transmitted pulse has a Gaussian shape and the return pulse from single-reflecting surfaces is also usually Gaussian in shape. The observed Gaussian shape of the returns confirms the assumption that the surface within the laser footprint can be modeled as a combination of a smooth-sloping surface and a rough surface with a random distribution of heights. Derived parameters include: mean surface elevations, pulse amplitude, pulse width, and the signal to noise ratio. The mean surface elevation is represented by the location of the center of the Gaussian fit. The combined effect of surface slope and surface roughness is calculated from the spreading of the pulse width. Multiple Gaussian functions are used to model waveforms resulting from multi-layer surfaces such as vegetated land or the edge of icebergs. Although the centroid of the waveform is sometimes used to represent a mean surface, the centroid is strongly influenced by asymmetry in the tails of the waveform and/or the limits over which the centroid is calculated. Over multi-layer surfaces, it is more useful to identify the individual layers and their associated mean elevations, which is done by multiple Gaussian fitting. Distortion of the waveform shape by non-surface characteristics, such as atmospheric forward scattering and detector/amplifier saturation, causes the centroid of the waveform to misrepresent the actual mean surface. These effects can be diminished by fitting a Gaussian to the return, and using the centroid of the Gaussian to determine the mean surface elevation. We present algorithms that use single or double Gaussians to fit the return waveform and show how the mean elevation and surface characteristics are calculated from the functional fit. The initial estimates and covariance matrix are set to optimize the fit to the leading edge of the return waveform corresponding to the largest Gaussian peak. Over ice surfaces, two Gaussian peaks are allowed to account for the extended tail of the returns that have high forward scattering components, or two distinct surfaces in the footprint. Over land, up to six Gaussian peaks are allowed. The algorithm was fine tuned using the first 36 days of data, which included returns over the ice regions with high detector/amplifier saturation and strong atmospheric forward scattering.
DE: 1640 Remote sensing
DE: 1827 Glaciology (1863)
DE: 4556 Sea level variations
DE: 5462 Polar regions
DE: 5464 Remote sensing
SC: Cryosphere [C]
MN: 2003 Fall Meeting