Rapid change of snow surface properties at Vostok, East Antarctica, revealed by altimetry and radiometry

Abstract

We present results of snow surface properties using the ENVISAT dual frequency altimeter at S (3.2 GHz) and Ku (13.6 GHz) bands and the AMSR-E microwave radiometer at frequencies ranging between 6 and 36 GHz in the Vostok region, East Antarctica. The altimetric time series observed between 2002 and 2008 show variations at 3 different time scales (daily, seasonal and inter-annual), that correlate directly with variations in the snow surface properties. In this study we focus on the analysis of the rapid daily event, occurring on February 14th 2005, that created a jump of the backscatter coefficient of up to 5.3 dB at the S band and 2.5 dB at the Ku band. The ratio of V/H-polarization brightness temperature slowly decreased in December and January 2005, and suddenly increased on February 14th 2005.

The origin of this rapid event is investigated using AWS data from Vostok station, altimetric and radiometric data simultaneously. Both snow surface density and roughness are found to vary during this event. This event is shown to be synchronous with strong wind occuring during a period of anomalous wind direction, and the presence of surface hoar. These particular conditions certainly modified the snow surface roughness and thus impacted the altimetric signal. We finally investigate the impact of this event on the calculation of the regional ice-sheet mass-balance using different corrections of height with echo shape variations. It is shown to be negligible only if the full echo shape correction (Legresy et al., 2006) is used.

Introduction

The upper 100 m of the Antarctic ice-sheet consists of dry-snow, with seasonal melting of the surface only occurring near the coasts (Picard et al., 2007). The inland snow stays dry, but is subject to seasonal metamorphism (Colbeck, 1983) and compaction (Li & Zwally, 2002) that occurs under temperature forcing, snow accumulation and wind driven process. Antarctic winds distribute the snow over the surface and hence control the surface properties.

A knowledge of the snow surface properties (roughness and density) is of prime importance for two main reasons: 1) these properties are related to climatological parameters, such as wind, precipitation and air temperature, 2) they are key parameters for analyzing remotely sensed data. All microwave sensors are sensitive to centimeter scale surface roughness. The signals from altimeters are particularly sensitive to snow surface density and roughness: smooth surface targets can easily be identified from altimetric signals (Lacroix et al., 2007), and a change of snow surface density will introduce a bias in the radar altimetric altitude estimation (Lacroix et al., 2008).

In situ measurements of snow surface density will typically be limited to pits and shallow cores, drilled mostly during the summer field seasons, whereas measurements of snow surface roughness are almost non-existent. Only sparse measurements of snow density and surface roughness exist over the Antarctic ice-sheet, and are thus insufficient to resolve time and space variability of these parameters. The only realistic way to observe these parameters is via the use of remotely sensed data aided by available in situ data.

We present here an analysis of altimetric data in the Vostok area (East Antarctica). We show that snow properties variations occur at different time scales, and focus on the interpretation of a rapid event occuring on February 14th 2005, based on the analysis of meteorological, altimetric and radiometric data. We then use this event to estimate the impact of snow properties variations on the altimetric measurements.