Primary Ice Edge extraction tools

Ice edge position on September 22-24 in the year indicated by line color. AMSR-E data processed with the NT2 algorithm.

Purpose

This repository holds a collection of Matlab scripts developed to extract the outermost edge of the sea ice ("the primary ice edge") from satellite-observed sea ice concentration.

Methodology

Given a two-dimensional map of ice concentration, $C(x,y)$, we define the ice edge to be given by a low concentration value, typically $C_0=15%$. The question is how to invert the field to get $(X(C=C_0), Y(C=C_0))$, the spatial position of the corresponding edge.

We use the Matlab function contour to plot the ice concentrations. The function returns a two-dimensional array $(x,y)$ of positions of a given contour (e.g $C_0=15$). These position arrays consist of line segments separated by a distinctive marker. We scan through this array and extract the segments. The endpoints of the segments are examined, and, if they are within max_gap distance (see below) of each other, they are joined together. This allows the

Input data

Scripts are provided for reading several passive microwave ice concentration products, including SSM/I, AMSR-E, and AMSR2. However, you will need to download these files to your local machine.

First, acquire the sea ice concentration data sets

• AMSR2 at 3.125 km analyzed using the ARTIST algorithm: ftp-projects.cen.uni-hamburg.de/seaice/AMSR2/3.125km/. The files are compressed netcdf files with names like Ant_20120801_res3.125km.pyres.nc.gz. After downloading, uncompress these files (e.g. using gzip on linux).
• AMSRE at 6.25 km for the years 2002-2011 analyzed using the ARTIST algorithm: ftp-projects.cen.uni-hamburg.de/seaice/AMSR-E_ASI_IceConc/hdf/s6250/. The files are compressed hdf files with names like asi-s6250-20110204-v5i.hdf.gz
• AMSRE at 12.5 km for the years 2002-2011 analyszed using the NASA Team 2 (NT2) algorithm. Go to https://doi.org/10.5067/AMSR-E/AE_SI12.003 and select the "Download Data" tab. To get a large number of files, choose the "Other Access Options" button.
• SSM/I at 25 km for the years 1978-2018 using the Bootstrap algorithm. Go to https://doi.org/10.5067/7Q8HCCWS4I0R and select the "Download Data" tab. To get a large number of files, choose the "Other Access Options" button.

Usage

The master script is run_iceedge_finder, which calls all other scripts.

Before running the script, set data_dir_root to point to the directory where the downloaded data sets reside. They are assumed to be in subdirectories named AMSR2, AMSEB,AMSRH,SSMI1.

Choose the [data_type], which corresponds to the satellite data source. There are two other parameters that can be chosen smooth_scale and max_gap. smooth_scale is the smoothing scale in kilometers (larger is smoother) and max_gap is the largest gap (in kilometers) that is allowed in a segment before breaking it into two. This allows the segment to continue across a small island or peninusla.

Example Application

These scripts were used to extract the primary ice edge from all four data sets with smooth_scale=500 and max_gap=100.

The output files are stored at from the sample run are archived at Zenodo

References

Cavalieri, D. J., T. Markus, and J. C. Comiso. 2014. AMSR-E/Aqua Daily L3 12.5 km Brightness Temperature, Sea Ice Concentration, & Snow Depth Polar Grids, Version 3. [Indicate subset used]. Boulder, Colorado USA. NASA National Snow and Ice Data Center Distributed Active Archive Center. doi:https://doi.org/10.5067/AMSR-E/AE_SI12.003. Accessed Jan. 7. 2020.

Comiso, J. C. 2017. Bootstrap Sea Ice Concentrations from Nimbus-7 SMMR and DMSP SSM/I-SSMIS, Version 3. [Indicate subset used]. Boulder, Colorado USA. NASA National Snow and Ice Data Center Distributed Active Archive Center. doi:https://doi.org/10.5067/7Q8HCCWS4I0R. Accessed Jan. 7, 2020.