This repository provides tools for converting data from META3.1EXP DT into labeled outputs compatible with EddyNet, a deep learning model designed for eddy detection.
The input data is sourced from sea surface height datasets provided by Copernicus.
-
Input Preparation
Place the downloaded META3.1EXP DT dataset in the root directory alongside all provided scripts. -
Script Execution Workflow
The generated outputs will be stored in theData
folder. Each script is tailored to process either anticyclonic or cyclonic eddies, indicated by the following suffixes:ac
: For anticyclonic eddiesc
: For cyclonic eddies
-
Execution Order
Execute the scripts with the same suffix in the following order:-
area_limit
-
grid_trans
-
data_generate
-
For test datasets in the
grid_trans
step, an additional suffixtest
is included. -
To output data as a dictionary instead of an array, use
data_generate_dict
.
-
This pipeline ensures accurate and efficient conversion of eddy data for use with EddyNet.