Engineering assessment of ice gouge statistics from the Canadian & American Arctic Oceans

Caines, Jonathan V. M. (Valance Maxwell) (2009) Engineering assessment of ice gouge statistics from the Canadian & American Arctic Oceans. Masters thesis, Memorial University of Newfoundland.

[English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (14MB)

Abstract

An engineering assessment of seabed ice gouging has been conducted for the American Beaufort, Canadian Beaufort, and Chukchi Seas. This assessment was limited to compilation of historical public domain ice gouge data and statistics. Data bias, correlation, and regional ice gouge measurement and analysis procedures used in probabilistic assessment of ice gouge geometry and recurrence rate estimates have been evaluated through investigation of previous studies and available data sets. -- The United States Geological Survey has collected a significant amount of ice gouge data through numerous seabed survey programs conducted in the American Beaufort and Chukchi Seas. The interpreted data is available in the public domain as numerous open-tile t -port publications. Historical Canadian Beaufort Sea ice gouge data was collected through Geological Survey of Canada and the Program for Energy Research and Development research initiatives. Interpreted data was archived in the SCOURBASE and ECHOBASE databases and is updated in the NEWBASE database through ongoing studies; however, data interpretation was contracted to Canadian Seabed Research and is not publicly available. Therefore, numerous summary reports and subsets of the Canadian Beaufort Sea ice gouge databases, available through Environmental Studies Research Fund (ESRF), have been utilized in this work. -- Research has indicated that seabed soil conditions limit ice gouging processes, with deeper gouge depths generally occurring in weak marine silts and clays. Dynamic ice gouge infilling processes are influenced by seabed sediment properties, general sediment deposition rates, water depth, gouge geometry, and bathymetry, although waves and currents are the dominant infilling mechanisms. Ice gouge infilling processes, minimum gouge depth cut-offs, and class range sizes contribute to interpretation subjectivity, bias, and perceived differences between regional ice gouge data collections. These processes were reviewed in this work, but were not integrated in statistical and probabilistic analyses. -- Investigated ice gouge depth statistical distributions included the gamma, Weibull, and exponential forms. In contrast with many early investigators (i.e., Lewis, 1977a; 1977b; Weeks et al., 1983; Lanan et al., 1986) who recommended the single-parameter exponential distribution as an effective and conservative probabilistic ice gouge model, His study has found the three-parameter gamma and/or Weibull distributions to more appropriately model ice gouge depth data from each region. However, both of these distributions may be reduced to the exponential form under specific conditions. Available ice gouge depth data sets were analyzed as mixed distributions during this thesis, with fixed probabilities of exceedence assigned to shallow gouge depth data and continuous distributions fit to the distribution tails. The mixed distributions were not associated with gouge depth resolution cut-offs, but due to large amounts of shallow gouge depth data in discrete bins. These discrete data bins were characteristics of the available data used for analysis and may be associated with data bias and uncertainty in the ice gouge process. Previous researchers (i.e., Nessim & Hong, 1992) have analyzed entire ice gouge depth data distributions as continuous. By analyzing available gouge depth data sets as mixed distributions, this study has removed bias and uncertainty introduced by the large amounts of shallow gouge depth data. Goodness-of-fit assessments were based on comparison of the fitted distributions and empirical cumulative distribution functions with data histograms and cumulative distributions, respectively. Assessment using probability plots and formal goodness-of-fit tests was not conducted since the available data sets were too large to produce meaningful results. -- Analysis was conducted for investigation of ice gouge parameter correlation, including ice gouge depth, width, and water depth relationships. In general, ice gouge depths exhibited positive relationships with associated water depths. Other parameters were also examined, including gouge widths and lengths, but did not show any relationship. Analysis of dominant ice gouge orientation data indicated a general northeast - southwest ice gouging direction in each analyzed region, thus suggesting that gouges are not necessarily formed orthogonal to the shoreline. -- Additional work is recommended to address ice gouge modeling issues such as considerations for ice gouge infilling processes, gouge correlation with geotechnical and environmental data, and analysis of gouge depth and width correlations.

Actions (login required)

View Item