Date

Call for Abstracts
Ocean Sciences Meeting 2022

27 February-4 March 2022
Honolulu, Hawaii

Abstract submission deadline: Wednesday, 29 September 2021, 11:59 p.m. Eastern Daylight Time

For more information and to submit an abstract, go to:
https://www.aslo.org/osm2022/


Ocean Sciences Meeting 2022 invites abstract submissions for scientific sessions, innovative sessions, town halls, and other auxiliary events. This conference will convene 27 February-4 March 2022 in Honolulu, Hawaii.

Conveners of the following virtual session invite abstracts:

OD15: Analyzing Sea Ice with Machine Learning Techniques
Conveners: Thomas Y. Chen and John Calhoun

As climate change accelerates rapidly throughout the world, research has shown that the regions that are the most severely affected are the Arctic and the Antarctic. Therefore, it is especially important to monitor geophysical parameters at the polar regions to assess the impacts of climate change. One such parameter is sea ice drift, which yields insights into the rates of global warming as well as key impacts on oceans and their ecological systems. Furthermore, analyzing the rate of melting sea ice aids in the understanding of climate change; sea ice is also important in reflecting solar energy. Changes in sea ice can disrupt normal ocean circulation. More generally, because sea ice is present in highly dynamic environments that involve winds and ocean currents, it can yield insights about how society can work to implement policies to both mitigate and adapt to climate change. Recently, machine learning methods have emerged as a key asset in assessing sea ice and deriving the insights that come with it. In terms of techniques, these range from simple linear regression to random forest ensemble models (RFs), and from support vector machines (SVMs) to artificial neural networks (ANNs). For example, harnessing convolutional neural networks (CNNs) on multitemporal satellite imagery data can yield results on the rate and severity of ice melt. Parameters to consider include concentration, thickness, drift velocity, and more, with machine learning-ready data collected from buoys, satellites, drones, in addition to feet on the ground. The goal of this session is to foster the interdisciplinary collaborations that are necessary at this exciting intersection. While artificial intelligence is crucial for the future of this area of study, conveners recognize the unique importance of domain-specific knowledge. Therefore, conveners seek contributions from both the computer science and cryospheric sciences communities.

For questions about this session, contact:
Thomas Chen
Email: thomaschen7 [at] acm.org