Ice Sheet Topography from a New CryoSat-2 SARIn Processing Chain, and Assessment by Comparison to ICESat-2 over Antarctica

Jérémie Aublanc,Nicolas Picot,Pierre Thibaut,Amandine Guillot,François Boy

doi:10.3390/rs13224508

Abstract

In this study, we present a new level-2 processing chain dedicated to the CryoSat-2 Synthetic Aperture Radar Interferometric (SARIn) measurements acquired over ice sheets. Compared to the ESA ground segment processor, it includes revised methods to detect waveform leading edges and perform retracking at the Point of Closest Approach (POCA). CryoSat-2 SARIn mode surface height measurements retrieved from the newly developed processing chain are compared to ICESat-2 surface height measurements extracted from the ATL06 product. About 250,000 space–time nearly coincident observations are identified and examined over the Antarctic ice sheet, and over a one-year period. On average, the median elevation bias between both missions is about −18 cm, with CryoSat-2 underestimating the surface topography compared to ICESat-2. The Median Absolute Deviation (MAD) between CryoSat-2 and ICESat-2 elevation estimates is 46.5 cm. These performances were compared to those obtained with CryoSat-2 SARIn mode elevations from the ESA PDGS level-2 products (ICE Baseline-D processor). The MAD between CryoSat-2 and ICESat-2 elevation estimates is significantly reduced with the new processing developed, by about 42%. The improvement is more substantial over areas closer to the coast, where the topography is more complex and surface slope increases. In terms of perspectives, the impacts of surface roughness and volume scattering on the SARIn mode waveforms have to be further investigated. This is crucial to understand geographical variations of the elevation bias between CryoSat-2 and ICESat-2 and continue enhancing the SARIn mode level-2 processing.

Highlights

Monitoring the continental ice sheet is a crucial issue to understand and evaluate the impacts of global warming
Where ∆h are the CryoSat-2 and ICESat-2 co-located elevation differences. Both median bias and Median Absolute Deviation (MAD) allow mitigating the impact of large outliers in the statistics
The median bias, MAD, and outliers’ ratio of the CryoSat-2 and ICESat-2 elevation differences are shown in Figure 4, represented as a function of the nadir surface slope derived from Reference Elevation Model of Antarctica (REMA)

Summary

Introduction

Monitoring the continental ice sheet is a crucial issue to understand and evaluate the impacts of global warming. In this context, the polar ice sheet melting needs to be thoroughly surveyed as it dominates uncertainties in the projected sea level [1]. Among the different techniques employed, spatial altimetry is a powerful tool to monitor the ice sheet mass balance by measuring the ice topography, and converting its evolution in time into mass change. Since the early 1990s, spatial altimetry has provided a continuous time series on mass change rate, mostly composed of data acquired in the conventional Low. Resolution Mode (LRM) from ERS 1 and 2, ENVISAT, and, more recently, CryoSat-2 and SARAL [2,3].

Results

Discussion

Conclusion