This one-hour webinar will start at 9:00 am (Alaska Daylight Time/UTC−08:00) on Tuesday, 13 July 2021.
This webinar is designed for the sea-ice research community and others interested in information about applications of machine learning in sea-ice forecasting. While this is an open event, attendees are advised that the discussions will largely be of a technical nature.
The rapid decline of Arctic sea ice and the implications for a broad array of stakeholders have spurred a decade of research activity into sea ice predictability and prediction. In this talk, I will introduce a newly developed dynamical seasonal prediction system, GFDL-SPEAR, and assess the performance of this system for regional sea ice predictions. I will discuss the key physical sources of Arctic sea ice predictability, routes to improving sea ice predictions, and fundamental limits on prediction skill. Advancing dynamical Arctic sea ice prediction capabilities will require coordinated efforts between the modeling, observational, and data assimilation communities.
Mitch Bushuk is a research scientist at the Geophysical Fluid Dynamics Laboratory, where he works on sea ice predictability and polar climate. He received his PhD in Atmosphere-Ocean Science and Mathematics from New York University in 2015, and his B.Sc. in Mathematics and Physics from the University of Toronto in 2009.
This webinar focused on the use of Machine Learning and Artificial Intelligence, on how such approaches can be applied in cryospheric research, as well as related challenges and limitations. Discussion included the use of advanced machine learning algorithms in climate science with attention on how to improve the prediction of future sea-ice. Participant questions and answers from speakers followed the presentations.
This one-hour webinar started at 8:00 am (Alaska Daylight Time)
The webinar has been archived and is available online.
Note: In addition to the Zoom presentation, a free livestream of this webinar was broadcast on YouTube — a facilitator provided support to those participants to submit written questions to the speakers.
This webinar was designed for the sea-ice research community and others interested in information about applications of machine learning in sea-ice forecasting. While this was an open event, attendees were advised that the discussions would largely be of a technical nature.
Elizabeth A. Barnes, Associate Professor, Department of Atmospheric Science, Colorado State University
Morteza Karimzadeh, Assistant Professor, Department of Geography, Colorado University Boulder
Tom R. Andersson, Data Scientist at the British Antarctic Survey Artificial Intelligence Lab
Elizabeth (Libby) Barnes is an associate professor of Atmospheric Science at Colorado State University. She joined the CSU faculty in 2013 after obtaining dual B.S. degrees (Honors) in Physics and Mathematics from the University of Minnesota, obtaining her Ph.D. in Atmospheric Science from the University of Washington, and spending a year as a NOAA Climate & Global Change Fellow at the Lamont-Doherty Earth Observatory. Professor Barnes' research is largely focused on climate variability and change and the data analysis tools used to understand it. Topics of interest include earth system predictability, jet-stream dynamics, Arctic-midlatitude connections, subseasonal-to-seasonal (S2S) prediction (she is currently Task Force Lead for the NOAA MAPP Subseasonal-to-Seasonal (S2S) Prediction Task Force), and data science methods for earth system research (e.g. machine learning, causal discovery). She teaches graduate courses on fundamental atmospheric dynamics and data science and statistical analysis methods.
Morteza Karimzadeh is an assistant professor in the Department of Geography and affiliate assistant professor in the Department of Information Science at the University of Colorado (CU) Boulder. He joined CU from Purdue University, where he was a postdoctoral scientist at the School of Electrical and Computer Engineering. Morteza is a geospatial data scientist, with research cutting across geographic information retrieval, machine learning, geovisualization, and visual analytics. His primary research focuses on method development, spanning various domains including social media analytics, environmental data fusion and analysis, energy (resilience and production), situational awareness, precision agriculture, and digital humanities. His approach to research and development is human-centered, from visual design to ground truth creation, algorithm integration and evaluation, to domain deployment and field studies.
Tom Andersson is a Data Scientist at the British Antarctic Survey (BAS) Artificial Intelligence Lab and holds a Master's degree in Information Engineering from the University of Cambridge. Tom's research at BAS involves the application of cutting-edge machine learning algorithms to remote sensing and climate model data. Tom's work is in collaboration with the Alan Turing Institute in London. For more information, see: Tom Andersson’s website.
Understanding Stakeholder Information Needs for Sea-Ice Forecasting
This webinar included an overview of stakeholder groups and their information needs, examples of how stakeholder groups deal with sea-ice and relate to sea-ice information, and discussion of how information gathered through stakeholder engagement can be of use to the sea-ice forecasting community. Time for participant questions followed the presentations.
Note: In addition to the Zoom presentation, a free livestream of this webinar was broadcast on YouTube — a facilitator provided support to those participants for submiting written questions to the speakers. There was a delay in the relay from the Zoom presentation to the livestream on YouTube.
This webinar was designed for the sea-ice research community and others interested in information about understanding stakeholder needs for sea-ice forecasting. While this was an open event, attendees were made aware that the discussions would largely be of a technical nature.
Hajo Eicken, Director, International Arctic Research Center
Joseph Little, Experimental Arctic Prediction Initiative, International Arctic Research Center/School of Management
Zeke Baker, Postdoctoral Research Associate, University of Oklahoma and National Weather Service-Alaska
Marta Terrado, Science Communication Specialist, Barcelona Supercomputing Center
Hajo Eicken is Professor of Geophysics and Director of the International Arctic Research Center at the University of Alaska Fairbanks. His research focuses on sea ice geophysics, Arctic coastal processes, and their importance for human activities and ecosystems. In Alaska he has helped lead efforts to advance collaborative research with Indigenous knowledge holders and to enhance use of scientific data by Arctic communities and government agencies. For more than a decade, he has worked with colleagues to establish a sea-ice observatory at Utqiaġvik/Pt. Barrow. Other collaborative efforts include his involvement in helping launch the Arctic Sea Ice Outlook and Sea Ice Prediction Network, his co-leadership of the Arctic Observing Summit, and service as Chair of a National Academies Standing Committee on Offshore Science and Assessment.
Joseph Little is Professor of Economics and Director of the Economics Program in the School of Management at the University of Alaska Fairbanks. His research interests include Applied microeconomics, environmental economics, and non-market valuation. As a Project Team member with SIPN2, his work draws from an online survey of the Bering Sea fixed gear fleet to evaluate stakeholder preferences for seasonal scale sea-ice prediction.
Zeke Baker is a sociologist whose research uses historical and qualitative methods to understand the development and use of environmental/climate science, especially insofar as climate knowledge is embedded in social relationships of power. He is a Postdoctoral Research Associate with the University of Oklahoma Cooperative Institute for Mesoscale Meteorological Studies and works with the National Oceanic and Atmospheric Administration (NOAA) in Anchorage, Alaska, leading a project that evaluates how different groups use and interpret marine forecasts in the Bering Sea region.
Marta Terrado has an experience of more than 10 years in agriculture, water management and ecosystem services research. She is Science Communication Specialist at BSC’s Earth Science Department, supporting activities on communication, dissemination and user engagement. Working in the co-production of climate services, Marta facilitates knowledge transfer for climate change adaptation at the science-stakeholders interface. She has a PhD in Earth Sciences (University of Barcelona) and a Master’s degree in Geographical Information Systems (Polytechnic University of Catalonia).
An Overview of European Union-Funded Project APPLICATE
This webinar provided an overview on the EU-Funded H2020 project APPLICATE, whose main goal is to advance our capability to predict the weather and climate in the Arctic and beyond. Ortega presents a few examples on the seasonal prediction activities carried out within the project. These included an analysis on how the different forecast errors are developed in the EC-Earth system, a multi-model comparison of predictive skill in all the seasonal forecast systems participating to the Consortium, results from empirical statistical models used for benchmarking, and experiments exploring the added-value of increasing both the atmospheric and oceanic resolution on seasonal prediction.
Pablo Ortega, Earth Science Department, Barcelona Supercomputing Center
Pablo Ortega is currently co-leading the Climate Prediction Group of the Barcelona Supercomputing Center. His research is focused on climate variability and predictability in the North Atlantic region, and he’s particularly interested in the role that both the ocean and sea ice play on climate prediction at seasonal-to-decadal timescales.
The webinar focused on one of NASA’s science objectives for the ICESat-2 altimetry mission—to provide observations to quantify changes and to add to previous satellite and airborne records of freeboard, thickness, and sea surface height of the ice-covered Arctic and Southern Oceans (e.g., from ICESat, Operation IceBridge, and CryoSat-2, etc.). ATLAS, a multi-beam photon-counting lidar, the sole instrument on the ICESat-2 observatory, launched in September 2018, provides a rich altimetric dataset of multiple profiles of the ice and ocean surfaces. In this talk, Ron Kwok talked about the capabilities of the multi-beam instrument based on data acquired thus far over the Arctic and Antarctic ice covers. In particular, he showed the precision in the retrieved surface heights over relatively flat surface, the spatial resolution of the height estimates, the time-varying freeboard estimates and sea surface height anomalies over a seasonal cycle, and assessments of the retrievals when compared with airborne and field acquisitions.
Webinar Presenters: Ron Kwok, NASA Jet Propulsion Laboratory and the ICESat-2 Team
Ron Kwok is a Senior Research Scientist at the Jet Propulsion Laboratory, California Institute of Technology, Pasadena California. His research interests include the mass and energy balance of the Arctic and Southern Ocean ice cover and the role of the sea ice in global climate, with current focus on the analysis of thickness, small-scale sea ice kinematics, time varying gravity from various spaceborne and airborne instruments.
An Overview of MOSAiC: The Multidisciplinary Drifting Observatory for the Study of Arctic Climate
This webinar will provide an overview of the interdisciplinary international MOSAiC program and ongoing work on the causes and consequences of an evolving and diminished Arctic sea ice cover. MOSAiC is a large interdisciplinary international program addressing an overarching question of "What are the causes and consequences of an evolving and diminished Arctic sea ice cover?" The centerpiece of MOSAiC is a year-long drift experiment from September 2019 to September 2020. In MOSAiC, models are being used to inform observations and observations will be used to enhance models. MOSAiC is taking an interdisciplinary approach with elements investigating the atmosphere, ice, ocean, ecosystem, and biogeochemistry. Observations will be conducted on three primary scales; the central observatory (5 km), the distributed network (50 km), and the large scale (1000s km).
Donald Perovich is a Professor at the Thayer School of Engineering, Dartmouth College. His research is focused on the geophysics of sea ice, with particular emphasis on electromagnetic, thermodynamic, and morphological properties. A central element of his work is observing and understanding the role of the sea ice albedo feedback in the Arctic climate system.
This webinar provided an overview of the SIPN2 main activities and an overview of the SIPN Data Portal for sea ice prediction. Uma Bhatt presented an overview of the SIPN2 project goals and an update on related activities to improve Arctic sea ice forecasts using a multi-disciplinary approach that includes modeling, new products, data analysis, and scientific networks. Cecilia Bitz presented an overview of the Data Portal for SIPN Forecasts project, funded by the Office of Naval Research and the development of products including model visualization and access to data processing. Time for participant questions will follow the presentation.
Uma S. Bhatt, University of Alaska Fairbanks, Geophysical Institute Cecilia Bitz, University of Washington, Program on Climate Change
Uma S. Bhatt, SIPN2 Principal Investigator (PI), is Chair of Atmospheric Sciences at University of Alaska Fairbanks, Geophysical Institute and Director of NOAA Cooperative Institute for Alaska Research (CIFAR). Her research on climate variability aims to understand how climate system components impact one another.
Cecilia Bitz, SIPN2 Leadership Team member, is a professor in the Atmospheric Sciences Department and Director of the Program on Climate Change and part of the Future of Ice Initiative, all at University of Washington. Her research interests include the role of sea ice in the climate system and high-latitude climate and climate change and global coupled climate modeling. Including integrations at very high resolution.