Closed zxdawn closed 2 years ago
Note that it's better to have two variables called dbscan_label
and lightning_label
. lightning_label
is paired with the values of lightning mask
, while the dbscan_label
is the original clustered label by DBSCAN which can help us check the clusters.
The DBSCAN lightning labels have both clean and polluted lightning, while the lightning mask has distinguished them and grouped them using wind data:
https://github.com/zxdawn/S5P-LNO2/blob/a3404ca04403a71924fa91e7177e2d64783ba37d/main/s5p_lnox_functions.py#L312-L343
Because the number of lighting labels represents the lightning count, it should be updated with the lightning mask. Otherwise, we don't know how much lightning happened in each lightning mask.