Retrieval of sea ice ridging index from S1 SAR and AMSR2 PMW data

Installation

conda env create -f environment.yml
conda activate sar2sir_inference

Preparation

Change the following paths in the config file sir_is2_inference.py

• inference_options['cnn_path'] - location of the best_model_sir_is2_09_00054_weights.pth
• inference_options['output_dir'] - location of output
• train_options['s1_dir'] - location of S1 files
• train_options['amsr2_dir'] - location of AMSR2 files

Running

python infer.py path_to_config_file.py [--force]

Note

Parts of the code are taken from oficial MMSeaIce repository