The Density-Dimension Algorithm: Detection of complex cryospheric and climatic change signatures in NASA ICESat-2 satellite laser altimeter data.

Abstract

According to the International Panel of Climate Change (IPCC), glacial acceleration is the largest source of uncertainty in sea-level-rise (SLR) assessment. There are only four types of glacial acceleration, of which surging is the least understood one.  The current surge of Negribreen, Svalbard, demonstrates that mass loss from a single Arctic glacier in only 3 months during surge can equal 1% of global annual SLR. A surge is a rapid acceleration to 200 times the glacier's normal velocity, which leads to heavy crevassing.
 
NASA ICESat-2, launched in September 2018, carries a micro-pulse photon-counting multi-beam laser altimeter, the Advanced Topographic Laser Altimeter System (ATLAS). Because ATLAS records returns from every single photon, separation of signal and background is an ill-posed problem, as background changes drastically from night-time to day-time. To solve this problem, we have developed the density-dimension algorithm (DDA-ice) for surface-height determination of crevassed and otherwise complex ice surfaces. The DDA-ice is an auto-adaptive algorithm that resolves surface heights at the 0.7m resolution of the sensor and can be run for the entire Greenland Ice Sheet on the NASA cloud in 10% of real time (data acquisition). Thus we are now able to provide the high-resolution geophysical data needed to study, simulate and understand complex glacial acceleration processes. The DDA-ice forms the core of a cyberinfrastructure that will be shared according to OSS/OS principles.

Bio

Ute Herzfeld is a geomathematician, computer scientist, mountaineer and Arctic glaciologist. At the University of Colorado Boulder, Dr. Herzfeld is the Director of the Geomathematics, Remote Sensing and Cryospheric Sciences Laboratory and Research Professor in the Department of Electrical, Computer and Energy Engineering. New in January 2021, Ute is also a research faculty and member of the graduate committee in the Department of Computer Sciences, which motivates the invitation for this colloquium. Dr. Herzfeld has a doctoral degree in mathematics with secondary subjects in applied mathematics and geosciences and developed the first one of her data analytic software systems during an expedition to the Antarctic aboard RV POLARSTERN. Dr. Herzfeld is a member of NASA's ICESat-2 Science Team and each summer - except 2020 - leads a research expedition to an Arctic glacier system. The computer science focus of her work is in computational algorithms, including auto-adaptive algorithms and ML, for processing of big data sets and geophysical information extraction.

https://cuboulder.zoom.us/j/190280621