Doctoral dissertation estimating sea ice thickness from satellite Earth observation data

Pulse Peakiness parameter from CryoSat-2 satellite in January 2013 (colour) which can be used for sea ice classification. Black line indicates the expected border between multi-year ice (north of Greenland and Canada) and first-year-ice. Over smooth first-year-ice the radar altimeter is very peaky while over rough multi year ice it becomes more diffuse resulting in a lower Pulse Peakiness value.Pulse Peakiness parameter from CryoSat-2 satellite in January 2013 (colour) which can be used for sea ice classification. Black line indicates the expected border between multi-year ice (north of Greenland and Canada) and first-year-ice. Over smooth first-year-ice the radar altimeter is very peaky while over rough multi year ice it becomes more diffuse resulting in a lower Pulse Peakiness value.Dr. Marta Zygmuntowska has defended her degree of philosophiae doctor (PhD) entitled Arctic sea ice altimetry – advances and current uncertainties at the University of Bergen. Her doctoral study has been completed at the Nansen Center under the CISAR project - CryoSat land and sea ice studies in the Arctic funded by the Research Council of Norway and in 2014 by the Trond Mohn grant.

Dr. Marta Zygmuntowska addresses one of the most prominent features of global climate change - the reduction in Arctic sea ice thickness. The main source of information to derive sea ice thickness on an Arctic wide scale is radar altimetry from polar orbiting satellites. In this study state of the art satellite Earth observation data from satellite mission such as the US ICESat and the European CryoSat-2 are investigated.

Current estimates of ice thickness from satellite Earth observation data are associated with high uncertainties. The altimeter measures the distance from the satellite to the sea ice surface and to the surrounding water surface from an altitude of 700 kilometers above the Earth.

By calculating the elevation difference between ice and ocean surfaces the sea ice freeboard can be estimated, the part of the ice above the water level. The freeboard is closely related to the ice thickness, but pending on ice type characteristics and snow cover. In her thesis she presents a new quantification of uncertainties in Arctic sea ice thickness and volume (i.e. thickness multiplied by area) and presents the main sources of these uncertainties. Furthermore, she explores the possibility for sea ice type classification based on data from radar altimeters, which can be used to improve the estimates of sea ice thickness.

Assessment of uncertainties in the estimation of the sea ice thickness and volume in the Arctic are quantified using freeboard retrievals from ICESat and investigating different assumptions on snow depth, sea ice density and ice area. These geophysical properties of the sea ice and snow cover are important when converting freeboard measurements from altimeters in order to estimate the sea ice thickness and volume. She has shown that these parameters have an influence on the overall mean, the year-to-year variability, and on the long-term trends of the ice thickness estimates. The overall uncertainties appears to be larger than suggested in previous studies, and the recent reported dramatic reduction in Arctic sea ice volume between the ICESat (2003-2008) and CryoSat-2 (2010 -2012) periods appears to be smaller. The total uncertainty in sea ice volume is estimated to be around 13% during the cold season. Uncertainties in sea ice area are of minor importance for the estimates of sea ice volume. The uncertainty in snow depth characterization contributes up to 70% of the total uncertainty, and the sea ice density related to different ice types to be up to 30–35%.

Sea ice volume, its uncertainty and changes over the last decade. Green full line indicates our best estimate and its uncertainties of about 13%. For comparison previous results are shown: Black dashed line shows ICESat results from JPL/ Kwok et al. (2008) and grey dashed line CryoSat results from UCL/ Laxon et al. (2013). Green dashed line shows results for sea ice volume when the same values for snow depth and ice density are used as for CryoSat-2 data, indicating that the decline may have been less dramatic than reported previously.Sea ice volume, its uncertainty and changes over the last decade. Green full line indicates our best estimate and its uncertainties of about 13%. For comparison previous results are shown: Black dashed line shows ICESat results from JPL/ Kwok et al. (2008) and grey dashed line CryoSat results from UCL/ Laxon et al. (2013). Green dashed line shows results for sea ice volume when the same values for snow depth and ice density are used as for CryoSat-2 data, indicating that the decline may have been less dramatic than reported previously.Information about sea ice type can be utilized to characterize the snow and ice properties, which are essential for converting freeboard measurements into ice thickness. Radar altimeter data have accordingly been analyzed over different Arctic sea ice regimes in order to develop a method for sea ice classification of first and multi-year ice. In a first case study the radar signals from an airborne altimeter have been investigated. In 80% of the analyzed airborne data first and multi-year-ice is classified correctly.  For satellite data from CryoSat-2 the classification is more difficult as different ice types can be found in the larg satellite footprint. Additionally, the signal is found to be mainly sensitive to surface roughness - a surface characterization that is dependent on several environmental factors and not only the age of sea ice.

The PhD thesis Arctic sea ice altimetry – advances and current uncertainties is based on one published, one in press  and one to be submitted papers:

Paper I: Zygmuntowska, M.; Rampall, P.; Ivanova, N.; Smedsrud, L.H., 2014: Uncertainties in Arctic sea ice thickness and volume: New estimates and implications for trends.  The Cryosphere, 8, 705 - 720, 2014.

Paper II: Zygmuntowska, M.; Khvorostovsky, K.; Helm, V. and Sandven, S., 2013: Waveform classification of synthetic aperture radar altimeter over Arctic sea ice. The Cryosphere, 7, 1315-1324, 2013.

Paper III:Zygmuntowska, M. and Khvorostovsky, K.: Analysis of CryoSats radar altimeter waveforms over different Arctic sea ice regimes. Manuscript to be submitted, 2014.

AttachmentSize
PhDthesisMartaZygmuntowska.pdf11.34 MB