Retrieving sea ice drag coefficients and turning angles from in situ and satellite observations using an inverse modeling framework

Author Posting. © American Geophysical Union, 2019. This article is posted here by permission of American Geophysical Union for personal use, not for redistribution. The definitive version was published in Journal of Geophysical Research-Oceans 124(8), (2019): 6388-6413, doi: 10.1029/2018JC014881.For ice concentrations less than 85%, internal ice stresses in the sea ice pack are small and sea ice is said to be in free drift. The sea ice drift is then the result of a balance between Coriolis acceleration and stresses from the ocean and atmosphere. We investigate sea ice drift using data from individual drifting buoys as well as Arctic‐wide gridded fields of wind, sea ice, and ocean velocity. We perform probabilistic inverse modeling of the momentum balance of free‐drifting sea ice, implemented to retrieve the Nansen number, scaled Rossby number, and stress turning angles. Since this problem involves a nonlinear, underconstrained system, we used a Monte Carlo guided search scheme—the Neighborhood Algorithm—to seek optimal parameter values for multiple observation points. We retrieve optimal drag coefficients of CA=1.2×10−3 and CO=2.4×10−3 from 10‐day averaged Arctic‐wide data from July 2014 that agree with the AIDJEX standard, with clear temporal and spatial variations. Inverting daily averaged buoy data give parameters that, while more accurately resolved, suggest that the forward model oversimplifies the physical system at these spatial and temporal scales. Our results show the importance of the correct representation of geostrophic currents. Both atmospheric and oceanic drag coefficients are found to decrease with shorter temporal averaging period, informing the selection of drag coefficient for short timescale climate models.The scripts developed for this publication are available at the GitHub (https://github.com/hheorton/Freedrift_inverse_submit). The Neighborhood Algorithm was developed and kindly supplied by M. Sambridge (http://www.iearth.org.au/codes/NA/). Ice‐Tethered Profiler data are available via the Ice‐Tethered Profiler program website (http://whoi.edu/itp). Buoy data were collected as part of the Marginal Ice Zone program (www.apl.washington.edu/miz) funded by the U.S. Office of Naval Research. The ice drift data were kindly supplied by N. Kimura. H. H. was funded by the Natural Environment Research Council (Grants NE/I029439/1 and NE/R000263/1). M. T. was partially funded by the SKIM Mission Science Study (SKIM‐SciSoc) Project ESA RFP 3‐15456/18/NL/CT/gp. T. A. was supported at the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration. M. T. and H. H. thank Dr. Nicolas Brantut for early discussions on the implementation of inverse modeling techniques.2020-02-1