te Change. 
k, pp. 337- 
> change in 
second half 
, L19604. 
Marsh, S., 
à matching 
anagement. 
z, 63(5), pp. 
Aurray, T., 
of glacier 
atching of 
cience and 
e, P. and 
'hnique for 
esses and 
tical shape 
ment and 
21-928. 
difference? 
ta sets for 
Discussion, 
Videspread 
Peninsula. 
recht, P., 
ven, A.P., 
Modelling 
Dynamics, 
irtographic 
ence and 
>w method 
As without 
fe Sensing, 
ASSESSMENT OF SEA ICE FREEBOARD AND THICKNESS IN MCMURDO SOUND, 
ANTARCTICA, DERIVED BY GROUND VALIDATED SATELLITE ALTIMETER DATA 
D. Price? *, W. Rack?, C. Haas, P. Langhorne ©, O. Marsh? 
* Gateway Antarctica, University of Canterbury, Christchurch, New Zealand — 
(daniel.price, oliver.marsh).pg.canterbury.ac.nz, (wolfgang.rack(a)canterbury.ac.nz) 
"Department of Earth & Atmospheric Sciences, University of Alberta, Canada — (chaas@ualberta.ca) 
‘Department of Physics, University of Otago, Dunedin, New Zealand — (patricia.langhorne@otago.ac.nz) 
Commission VI, WG VIII/10: Cryosphere 
KEY WORDS: ICESat, sea ice freeboard, McMurdo Sound, remote sensing, lidar 
ABSTRACT: 
This investigation employs the use of ICESat to derive freeboard measurements in McMurdo Sound in the western Ross Sea, 
Antarctica, for the time period 2003-2009. Methods closely follow those previously presented in the literature but are complemented 
by a good understanding of general sea ice characteristics in the study region from extensive temporal ground investigations but with 
limited spatial coverage. The aim of remote sensing applications in this area is to expand the good knowledge of sea ice 
characteristics within these limited areas to the wider McMurdo Sound and western Ross Sea region. The seven year Austral Spring 
(September, October, and November) investigation is presented for sea ice freeboard alone. An interannual comparison of mean 
freeboard indicates an increase in multiyear sea ice freeboard from 1.08 m in 2003 to 1.15 m in 2009 with positive and negative 
variation in between. No significant trend was detected for first year sea ice freeboard. Further, an Envisat imagery investigation 
complements the freeboard assessment. The multiyear sea ice was observed to increase by 254 % of its original 2003 area, as 
firstyear sea ice persisted through the 2004 melt season into 2005. This maximum coverage then gradually diminished by 2009 to 20 
% above the original 2003 value. The mid study period increase is likely attributed to the passage of iceberg B-15A minimising 
oceanic pressures and preventing sea ice breakout in the region. 
1. INTRODUCTION 
Quantification of Antarctic sea ice thickness is yet to be 
achieved. This is a significant goal as assessment of holistic 
change to the sea ice cover must account for volume (i.e. extent 
and thickness information must be available). No complete 
investigation has been completed to even produce a baseline 
data set from which change can be assessed. No satellite 
instrumentation is capable of directly retrieving sea ice 
thickness measurements as no technology is capable of 
detecting the ice-ocean interface from such range. Indirect 
methodologies offer the only feasible way of assessing the 
Antarctic sea ice cover holistically. The underlying principle of 
these methodologies is the detection of sea ice freeboard, the ice 
protruding above the ocean surface and subsequent conversion 
to thickness. The use of NASA's Ice, Cloud, and land Elevation 
Satellite (ICESat) Geoscience Laser Altimeter System (GLAS) 
for total freeboard retrieval in the Antarctic has been well 
documented. ICESat offers a great opportunity for assessment 
of sea ice thickness in the Southern Ocean (Zwally et al. 2008). 
However, complications with the determination of sea ice 
freeboard from satellite altimetry methods have inherent errors, 
sometimes to the degree of the desired measurements. Accurate 
determination of freeboard is the prerequisite to thickness 
estimation; any errors in the initial freeboard calculation will be 
manifested in a later thickness calculation. Due to high 
precipitation rates in the Southern Ocean, complications are 
augmented by the common occurrence of negative ice freeboard 
as a result of snow loading. From laser altimetry methods, this 
  
* Corresponding author. 
results in no information being gained on ice freeboard allowing 
no assessment of sea ice mass balance alone. Zwally et al. 2008 
report that interannual changes in estimated thickness are 
mainly representative of snow thickness changes. Errors 
introduced to estimate thickness including nominal snow, ice 
and water density information can be quite significant (Yi et al. 
2011). To further assess methodology accuracy other studies 
(Markus et al. 2011; Worby et al. 2011; Xie et al. 2011) have 
made use of in situ information to support satellite data. 
The Ross Sea is an area of particular interest as it is the only 
sector in the Antarctic which has experienced a significant 
increase in sea ice extent during the satellite period (Comiso 
2010). In the western Ross Sea, two polynyas are also important 
areas for sea ice production, the Terra Nova polynya and 
McMurdo Sound polynya (Comiso et al. 2011). The presence of 
this polynya in the study area furthers the areas importance with 
regard to its role holistically in the Ross Sea. The information 
available on sea ice extent and production is not however 
complemented by any information on thickness trends. 
We assess the feasibility for deriving sea ice freeboard in the 
McMurdo Sound area in the south-western Ross Sea, where 
detailed ground information is available and a general idea of 
sea ice conditions is known (Dempsey et al. 2010). This 
investigation focuses on assessing sea ice freeboard retrieval 
from ICESat from 2003 to 2009. The 2009 data set is 
complemented by field measurements including near-coincident 
helicopter-borne lidar (HL) measurements and ground 
measurements for ice thickness, freeboard and snow depth 
permitting some comparison of freeboard retrieval techniques.