Yavor, please introduce yourself.
My name is Yavor Kostov. I’m an ocean/ice modeler at the British Antarctic Survey.
Tell us about your professional and academic career before becoming part of the OCEAN:ICE community.
I studied Mathematics as an undergraduate at Pomona College in California, where my senior thesis analyzed bifurcations in an idealized model of ENSO. To put it more plainly, I explored abrupt transitions between different possible regimes of the El Niño/La Niña phenomenon. I then moved on to do a Ph.D. in Climate Physics at MIT. As a doctorate student, I joined a large collaboration exploring the impact of greenhouse gases, the ozone hole, and atmospheric variability on the Southern Ocean and beyond. I then came to the UK for my postdoctoral projects at Oxford and then Exeter University, where I focused more on the ocean circulation and water masses in the North Atlantic. During my first postdoc, I tutored a physics course on fluids, nonlinear dynamics, and chaos. I also co-supervised an Oxford Ph.D. student researching the Beaufort Gyre and Arctic sea ice.
What do you do within OCEAN:ICE?
Within OCEAN:ICE, I model iceberg dynamics and iceberg grounding in the Amundsen Sea using the NEMO ocean model. I introduce new physically-motivated algorithms in the model code, verify them, and test them. I simulate icebergs in the vicinity of Bear Ridge, where they tend to ground and block the westward drift of sea-ice. The interaction between grounded icebergs and sea ice in the Amundsen Sea affects the regional stratification and the mixing of water masses. The latter is very important because warm deep water masses may destabilize the ice shelf in the basin.
What have you enjoyed about OCEAN:ICE so far?
I have really enjoyed familiarizing myself with old interdisciplinary literature on iceberg grounding. Some of it includes engineering studies from the 1980s, when they tried to protect undersea infrastructure from iceberg damage. I also found an old manuscript, dating back to the 19th century, and it speculates on iceberg interaction with topography.
Tell us about a skill or trait unique to you that you would like to share?
This is not a unique skill. However, I am part of a community of climate scientists who use a technique called “algorithmic differentiation” in sensitivity analysis. This tool allows us to generate code that corresponds to the first derivative of a simulated ocean index with respect to boundary conditions and model parameters. This technique has been widely used to assimilate data into computer models, but our community of enthusiasts applies it to study the sensitivity of the ocean circulation and ocean properties. I also use that method for causal attribution in model simulations. That is, if I am interested in what causes certain historical changes in an ocean index, I can attribute them to particular changes of the conditions at the ocean surface.
Stay tuned on our social media channels (X, Mastodon, LinkedIn and Facebook) for more of the series of OCEAN:ICE 'Researcher in the Spotlight' articles.