Geostatistical mapping and spatial variability of surficial sediment types on the Beaufort Sea shelf based on grain size data

• Main Authors: Jerosch, Kerstin, Kostylev, V. E., Blasco, S. M.
• Format: Conference Object
• Language: unknown
• Published: 2012
• Subjects: Beaufort Sea; Mackenzie river
• Online Access: https://epic.awi.de/id/eprint/30640/; https://epic.awi.de/id/eprint/30640/1/IPY_2012_Montreal_Poster_KJ.pdf; https://hdl.handle.net/10013/epic.39580; https://hdl.handle.net/10013/epic.39580.d001

The nearshore Beaufort Sea is a sensitive marine environment that is also the focus of oil and gas exploration. Offshore, the Beaufort Sea contains large potential reserves of hydrocarbons. Any future exploitation of these resources will present unique engineering challenges and will require an understanding of the processes that govern stability, nearshore morphology and sediment properties in the extensive shallow coastal zone of the Beaufort Sea shelf. Knowledge of the surficial sediment distribution is, therefore, necessary to provide a framework for understanding sediment stability, sediment transport, platform foundation conditions and to balance engineering challenges with environmental concerns, resource development and precautionary sustainable management. We describe an approach for a quality controlled mapping of grain sizes and sediment textures for the Beaufort Sea shelf in the Canadian Arctic. The approach is based on grain size data sampled during the period 1969-2008. A replenishment of grain size data since the 1980’s, as well as the consideration of correlating parameters (bathymetry, slope and sediment input) to a cokriging algorithm, amends the former way of mapping the surficial sediments of the Beaufort Sea shelf. Subsequent to data processing and applying autocorrelation, four single grids (clay, silt, sand and gravel) were generated from grain size data by ordinary kriging and cokriging. Cokriging also considered parameters that influence sediment texture such as bathymetry, slope, cost distance from the Mackenzie River and data anisotropy (directional dependency). The cokriging algorithm expressed as a variogram was quality controlled by cross-validation and predicted standard errors (PSEs). PSE values express a maximum deviation of modeled from the real values and therefore help to estimate the quality in these regions regarding the interpolation results for each grain size range. A sediment type classification scheme applied to the set of clay, silt, sand and gravel content maps ...