Sea Ice Koopman Mode Analysis
Abstract
The program goal was to apply Koopman Mode Analysis (KMA) to sea ice data sets to study their temporal and geographic behavior over multi-year time scales. Work in this period primarily consisted of prediction of future sea ice concentration values using Koopman reconstruction techniques; Prediction was performed for varied input year ranges between 3 and 38 years, and for prediction years between 1982 and 2017, and for the entire data available for each hemisphere and regions within each hemisphere; Results shown include prediction of spatial distribution of sea ice concentration, prediction of pixel-by-pixel deviation between predicted and actual results, and comparisons of the mean of the predicted and actual data for each region.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 30, 2018
- Accession Number
- AD1061584
Entities
People
- Ignor Mezic
- James Hogg
- Maria Fonoberova