ESTIMATING SURFACE ELEVATION BIASES FROM SUBSURFACE SCATTERED PHOTONS FOR LASER ALTIMETERS

Abstract

Three decades of satellite observations have revealed rapid changes in Earth’s cryosphere associated with anthropogenic climate change, including decreased extent and volume of Arctic sea ice, mass loss from the Greenland Ice Sheet, mass loss in West Antarctica and the Antarctic Peninsula, and increased outlet glacier discharge in Greenland and Antarctica. NASA’s ICESat-2 mission will continue observing these rapid changes by measuring individual photons’ round-trip travel times from the satellite to Earth’s surface, providing precise estimates of surface elevation, and subsequent mass change for ice sheets and sea ice freeboard in Earth’s polar regions. This study investigates the potential bias in ICESat-2 surface elevation estimates from photons that have volume scattered in snow by: (1) measuring the transmission of green light through snow, (2) developing a method capable of characterizing the effects of volume scattered photons recorded by laser altimeters, (3) applying this method to laboratory measurements of volume scattered photons using the simulation laser altimeter for ICESat-2, and (4) simulating volume scattered photon rage biases using a photon tracking Monte Carlo model.

Transmission measurements show that green light attenuates by one order of magnitude every centimeter in the first four centimeters of snow, suggesting that detecting volume scattered photons originating from laser altimeters is unlikely after photons travel more than a few centimeters in snow. Laboratory measurements using ICESat-2’s simulation laser altimeter MABEL (Multiple Altimeter Beam Experimental Lidar), show volume scattered photon return biases of 5 – 10 cm. However, these laboratory measurements revealed a previously unidentified drift in MABEL’s ranging on the order of 5 cm, potentially overestimating the volume scattering bias.

Simulations from a single-photon tracking Monte Carlo model developed for this study reveal that approximately 95% of backscattered photons accrue path lengths less than 5 cm. This suggests that while statistically possible for photons to accrue large path lengths, the likelihood of laser altimeters detecting these photons is small.

The results from this work demonstrate that volume scattered photons may be measured by photon counting laser altimeters, but will produce little bias in derived elevation estimates due to their low frequency of measurement.