This dissertation attempts to improve the understanding of spatial distribution and evolution of snow-covered sea ice as related to polar bear habitat. This will be accomplished at both the local (i.e. 1m spatial resolution) and regional scales (i.e. 100 km spatial resolution) for various types of first-year sea ice (FYI) through four primary objectives. The first primary objective (i.e. Chapter 3) examines the observed and modeled temporal evolution of snow over smooth FYI, as well as the estimation of on-ice meteorological conditions. Results suggest that increases in observed snowdrifts and changes to the orientation of the drifts are related to snowfall and drifting events. Modeling these changes over time using a spatially distributed snow model is not able to accurately estimate the snow distribution. As well, only the on-ice temperature and humidity can be estimated from land-based station data, limiting the modeling of snow over sea ice. The second primary objective (i.e. Chapter 4) extends this research to rough FYI types, more relevant to polar bear habitat. More specifically this objective studies the spatial pattern of snow distribution over rough ice and ridges and the relationship between ice roughness and meteorological conditions. Results suggest that ice roughness influences the magnitude of snow depth, while the wind direction during periods of snow deposition and/or blowing snow events will impact the spatial pattern. The third primary objective (i.e. Chapter 5) focuses on developing a more feasible method of deriving surface roughness. This objective attempts to use satellite imagery and texture analysis to derive surface roughness for snow-covered sea ice. Results suggest that a Gamma speckle reduction filter, coupled with a grey-level co-occurrence matrix texture measure (Entropy or Angular Second Moment) is able to account for more than 88% of the variability in the surface roughness. The final primary objective (i.e. Chapter 6) examines the temporal evolution and factors controlling the changes in sea ice characteristics over regional scale for a period from 1978 to 2002. Observed anomalies in sea ice characteristics within some of the polar bear subpopulations may be explained by thermodynamic and/or dynamic factors. Results suggest that published reduction in polar bear population and condition within the subpopulations co-occur with these observed changes in sea ice characteristics.

Previous understanding of the winter ice regime in Hudson Bay was limited. This investigation demonstrates the existence of a large coastal polynya in northwestern Hudson Bay. Measuring approximately 600 km long by 60 km wide, this polynya is typically opened and maintained by surface winds for periods of one to five days. Closing mechanisms result from decline in wind maintenance and apparent thermodynamic ice formation. Open water is present on the northwest coast at some size for up to 70% of the winter between ice formation and breakup. Observation of this open water during the period of the winter ice pack indicates that the ‘ice factory’, modelled previously, in northwest Hudson Bay exists. This furthers understanding of the freshwater budgets, stratification, and atmosphere-ocean interactions while presenting new questions about climate change and the future of the Hudson Bay physical system.

Thesis (Ph. D.)--University of Washington, 1997. / Vita. Includes bibliographical references (leaves [121]-127).

Thesis (M.S. in Meteorology and Physical Oceanography)--Naval Postgraduate School, December 1990. / Thesis Advisor(s): Bourke, Robert H. ; Morison, James. "December 1990." Description based on title screen as viewed on April 1, 2010. DTIC Identifier(s): Buoys, Sea Ice, Arctic Buoys, Ice Forecast, Arctic, Theses. Author(s) subject terms: PIPS, Sea Ice, Arctic Buoys, Ice Forecast, Arctic. Includes bibliographical references (p. 106-111). Also available in print.

Thesis (M.S. in Meteorology and Physical Oceanography)--Naval Postgraduate School, June 1990. / Thesis Advisor(s): Nystuen, J.A. ; Bourke, R.H. "June 1990." "MPS-68-90-004." Description based on title screen as viewed on March 24, 2010. DTIC Identifier(s): Radar Images, Sea Ice, Marginal Ice Zones, Ice Classification, Statistical Analysis, Gray Scale, Odden Ice, Theses. Author(s) subject terms: Synthetic Aperture Radar, Sea Ice Classification, Marginal Zone, Gray Level Co-Occurrence Matrices, Texture Statistics, Univariate Statistics, MIZEX '87 SAR Data. Includes bibliographical references (p. 96-98). Also available in print.

Thesis (M.S.)--University of California, Santa Cruz, 1989. / Typescript. Includes bibliographical references.

Thesis (M.Sc.)--Memorial University of Newfoundland, 2004. / Bibliography: leaves 126-131.

The analysis of the sea-ice diatom biomarker IP25 (a mono-unsaturated Highly Branched Isoprenoid (HBI) alkene) in Arctic marine sediments has previously been shown to provide a useful qualitative proxy measure for the past spring sea-ice occurrence. In the Southern Ocean the occurrence and variable abundance of a structurally similar di-unsaturated HBI (HBI diene II) has previously been proposed as a proxy measure of paleo sea-ice extent. However, the use of such biomarker proxies remains under development. In the current study, a number of additional palaeoceanographic developments of HBIs as sea-ice biomarkers in both polar regions has been undertaken. For the Arctic, an investigation into the combined analysis of IP25 and certain phytoplankton biomarkers has been conducted with the aim of providing more detailed and semi-quantitative descriptions of sea-ice conditions in the Barents Sea. In contrast, analysis of HBIs and other lipids within water column, surface sediment and sea-ice samples has been undertaken to provide further insights into the use of HBIs as proxies for Antarctic sea-ice. Analysis of surface sediments from across the Barents Sea has shown that the relative abundances of IP25 and a tri-unsaturated HBI lipid (HBI triene IIIa) are characteristic of the overlying surface oceanographic conditions, most notably, the location of the seasonal sea-ice edge. A semi-quantitative approach, in the form of the PIP25 index, showed a good positive linear relationship between PIP25 indices and spring sea-ice concentration, with a particularly strong relationship found when using HBI triene IIIa (PIIIaIP25) as the open-water counterpart to IP25. The quality of the linear fits were not especially dependent on the balance factor c, used in the PIP25 calculation, which may have important positive consequences for down-core sea-ice reconstruction, and when making comparisons between outcomes from different Arctic regions or climatic epochs. Further, a lower limit threshold for PIIIaIP25 (0.8) might represent a useful qualitative proxy for the past occurrence of summer sea-ice. The re-evaluation of biomarker data from three dated marine sequences in the Barents Sea suggests that the combined analysis of IP25 and HBI triene IIIa can provide information on temporal variations in the position of the maximum (winter) Arctic sea-ice extent, together with more quantitative sea-ice reconstructions. In the Southern Ocean, the distributions of di- and tri-unsaturated HBIs (HBI diene II and HBI trienes IIIa and IIIb) in surface waters were shown to be extremely sensitive to the local sea-ice conditions, consistent with significant environmental control over their biosynthesis by sea-ice diatoms and open water phytoplankton, respectively. Within the water column, the apparent alteration to HBI and other lipid abundances was evident between the photic and benthic parts of the water column, which, along with additional local factors (e.g. polynya formation), may have important implications for paleo sea-ice reconstructions. The sedimentary occurrence and distribution of HBI diene II (termed here as IPSO25) were consistent with the recent identification of the diatom Berkeleya adeliensis Medlin as a source of IPSO25. The tendency for B. adeliensis to flourish in platelet ice, the formation of which is strongly associated with super-cooled freshwater inflow, means that sedimentary IPSO25 may provide a potentially sensitive proxy indicator of landfast sea-ice influenced by meltwater discharge from nearby glaciers and ice shelves. Re-examination of some previous IPSO25 down-core records supports this suggestion, although further down-core analysis is required to confirm this hypothesis. The similar sedimentary distribution relationship between phytoplankton-derived HBI trienes and IPSO25, further indicates that the former may reflect production of these biomarkers by certain diatoms that flourish within the region of the retreating ice edge; however, the source identification of the HBI trienes is still needed to place this interpretation on a firmer footing.

Pressured and ridged ice is a dangerous hazard facing ships in the Arctic. Ships can become stuck or beset in these conditions, which is environmentally and economically costly. Understanding where and when ridges form as a result of pressured ice is important for ensuring safe winter shipping operations; however there have been few studies to date regarding the distribution of ridges and their impacts within a geographic region. The Hudson Strait, which connects Hudson Bay and the Atlantic Ocean, is the site of ongoing winter shipping, where vessels frequently encounter pressured ice. This thesis addresses two questions: where and when do ridges occur in the Hudson Strait and what are their impacts on an ice strengthened vessel traveling through the Strait. To answer the first question, ridges were manually identified in RADARSAT-1 and -2 images during the winter months (December to May) from 1997-2012. Ridge counts and densities for each winter season were calculated and their spatial distribution was mapped. A 30-year sea ice climatology of the Hudson Strait was also created in order to understand ongoing trends in freeze-up and breakup timing in the region. Recurring patterns in the location and timing of ridging were found in the Hudson Strait, specifically in areas where shearing and bottlenecks created pressure. Ridge densities were correlated with sea level pressure, air temperature and wind NCEP reanalysis data to look to for connections between these factors and ridge densities. Some connections were found between freeze-up dates, sea level pressure and ridge densities. The second half of this thesis focuses on how ridges impact the voyage of an ice-strengthened vessel. Log books from the MV Arctic, a cargo ship that makes two winter transits through the Hudson Strait every year, were used to plot the movement of the ship and where and when it became beset. These data were examined for temporal and spatial patterns in besetting events. Most besetting events took place in February and March. They typically occurred in the eastern and western ends of the Strait. These voyages were compared to ridge data from the first half of thesis, and there were good correlations between the presence of high ridge densities and ship besetting events, demonstrating that ridge densities identified in satellite imagery can act as a proxy when forecasting hazardous ice conditions. This research fills an important knowledge gap in understanding where and when pressured ice forms in the Hudson Strait and what factors play a role in creating this hazardous ice condition. It also addresses the impacts that ridges have on ship transits through the Strait.

From the position data of 23 drifters over the four years period, 1982 to 1985, the upper layer kinematics of the Northeast Pacific were investigated. The focus of this study was upon the topographic influence, wind forcing, and Rossby waves. The notions of homogeneity and stationarity were applied to both the Eulerian and Lagrangian analysis. A new computational scheme was proposed and tested in order to explicitly take into account the Lagrangian characteristics of the drifter data. The Eulerian analysis showed that the spatial and temporal distributions of the mean current and the mean wind stress were in good agreement. Three of the four eddy kinetic energy maxima found in the region geographically corresponded to topographic features. Also the velocity field of this region appeared to be strongly inhomogeneous, non-stationary, and anisotropic. The results from the Lagrangian analysis showed that the spectral slope for the periods shorter than 5 days followed the -2 law, and suggested that the direct wind forcing was a dominant mechanism for those periods. The spectrum of the eddy component of the velocity appeared to be white for the periods longer than 10 days. The results showed that the linear Rossby waves were not dominant mechanism for upper layer dynamics of this region. The rotary spectra illustrated some evidence of the rotational preference of the cyclonic over anti-cyclonic motions at a period of 10 days. The new scheme provided meaningful information about the eddy component of the velocity. / Science, Faculty of / Earth, Ocean and Atmospheric Sciences, Department of / Graduate