Data for Prediction of Sea Ice Motion with Convolutional Long Short-Term Memory Networks

5
1 rating - Please login to submit your rating.

Abstract 

This is the images and the image masks used in the paper "Z. Petrou and Y. Tian, Prediction of Sea Ice Motion with Convolutional Long Short-Term Memory Networks,IEEE Transactions on Geoscience and Remote Sensing."

Instructions: 

The dataset contains the images and the corresponding sea-ice masks used in the experiments in the manuscript, for the AMSR-E, AMSR2, and ASCAT sensors. The processed png 8-bit images are provided. In particular:

- AMSRE_2011.zip contains the data from the AMSR-E sensor:
  - images: the directory contains the brightness temperature images
    Naming convention: AMSRE_YYYYMMDD.png, where YYYY, MM, and DD stand for the year, month, and day of the month, respectively;
  - masks: contains the sea-ice masks (sea ice: 1, non-sea ice: 0)
    Naming convention: AMSRE_YYYYMMDD_mask.png

- AMSR2_2018.zip contains the data from the AMSR2 sensor:
  - images: the directory contains the brightness temperature images
    Naming convention: AMSR2_YYYYMMDD.png, where YYYY, MM, and DD stand for the year, month, and day of the month, respectively;
  - masks: contains the sea-ice masks (sea ice concentration: 1-255, non-sea ice: 0)
    Naming convention: AMSR2_YYYYMMDD_mask.png

- ASCAT_2017.zip contains the data from the ASCAT sensor:
  - images: the directory contains the brightness temperature images
    Naming convention: ASCAT_YYYYDOY.png, where YYYY is the year and DOY the day of the year (e.g. Feb 19 is 050, Feb 20 is 051, etc.);
  - masks: contains the sea-ice masks (non-ice ocean: 0, sea-ice: 128, land: 255)
    Naming convention: ASCAT_YYYYDOY_mask.png