Beth Caissie



Research Interests

Sea ice is a defining feature of our planet.  It serves as habitat for some of the most spectacular animals on Earth—polar bears, walrus, and ice seals—as well as for some of the smallest primary producers:  sea ice diatoms. Declining sea ice has the potential to further increase global temperatures through the ice-albedo feedback.  Dark open waters will absorb incoming radiation in contrast to the white surface of sea ice, which reflects solar radiation.  Because of this, it is critical to be able to model how sea ice will decline as temperatures rise.  One way to improve sea ice models is by looking at how sea ice changed in the past when the earth warmed due to natural forcings.   My research aims to add to our understanding of the temporal and spatial distribution of sea ice, especially in the Bering and Chukchi seas, particularly during past periods of climate warming. Some of the research questions that I am addressing include:

  • As the Earth went in and out of glacial-interglacial cycles, did the growth and decline of sea ice lead or lag the growth and decline of ice sheets?
  • Does sea ice duration decrease at a constant rate, or are there thresholds that lead to abrupt switches in sea ice extent or duration? 
  • During warming, are there periods of time that are characterized by re-advances of ice or does ice retreat in a linear fashion?

Specifically, my dissertation is a database of surface sediments from the Bering and Chukchi seas.  I have characterized diatom assemblages and measured certain morphological traits of diatoms in these core tops.  I am using statistical methods and machine learning techniques (modern analog technique, transfer functions, naïve Bayes classification, and artificial neural networks) to relate these sediment components to satellite-derived sea ice duration for each site.  Once this mathematical relationship is established, it will be able to be applied down core to reconstruct sea ice duration in the past.  As I continue with this work in the future, I hope to be able to expand my diatom work into earlier time periods, such as the Pliocene and use diatom assemblages and morphology as indicators of other environmental variables.  I am also keen to push the limits of temporal resolution when it comes to diatom reconstructions. Finely laminated intervals, possibly varves, found in the Bering Sea during interglacials (even as far back as Marine Isotope Stage 31) mean that it might be possible to see decadal-scale changes during previous warm, dysoxic periods.





Warming the Arctic: Sea Ice and Climate Change

Geosciences Grad Student Homepage  
Geosciences Department