1. Satellite images: the participants should use all free-accessible satellite data (e.g. Landsat, Sentienl-1/2) in the test regions, the data pre-processing methodology are not limited, either. We will provide 15-day composited Sentinel-2 time series data in the Iran and Sudan test regions. For the test region in Afghanistan, the participants will have to collect the imagery independently. The participants need to share the availability of the data they used and how the pre-processing is conducted, if they used more data than the official provided data set.

2. Training samples: We will provide a limited number of training samples for cropland mapping, participants can also collect some training samples by themselves as cropland extent can be visually interpreted from Google Earth and Satellite images. Particularly, for the cropland mapping in Afghanistan, training samples were collected in April 2022, and the model is used to identify cropland extent in the same period.

We recommend that you use Google Earth Engine, as it easy to use and the data is public and open to anyone. The provided Colab starter notebook uses Google Earth Engine to download satellite data for modelling. In order to use GEE, you first need to have a Google account, and then you have to register for GEE after which you will get a token which you will use to download data.