Preprocessing observations used for validation

This step aims at connecting the observations to the relevant Sentinel-2 pixels. Information from the intersecting Sentinel-2 tiles are added to the vector of observations. If it contains polygons instead of points, they will be converted to a grid of points corresponding to the centers of each 10 m Sentinel-2 pixel included in the polygon.

Each resulting point includes the following attributes :

• original polygon ID
• id_pixel: ID of the pixel
• name_area : Name of the corresponding area (e.g. Sentinel-2 tile ID)
• epsg: CRS of the associated Sentinel-2 tile as the EPSG integer.

These datapoints are used in the next step to extract Sentinel-2 data.

The points generated are kept in the CRS of the original vector file.

Using local THEIA data

Running this step using a script

Run the following instructions to perform this preprocessing step:

from pathlib import Path

from fordead.validation.obs_to_s2_grid import obs_to_s2_grid

output_dir = Path("<MyOutputDirectory>")
input_dir = Path("<MyInputDirectory>/fordead_data")

obs_path = input_dir / "vector/observations_tuto.shp"
sentinel_dir = input_dir / "sentinel_data/validation_tutorial/sentinel_data/"
preprocessed_obs_path = output_dir / "preprocessed_obs_tuto.shp"


obs_to_s2_grid(
    obs_path = obs_path,
    sentinel_source = sentinel_dir, 
    export_path = preprocessed_obs_path,
    name_column = "id")

NOTE : Set obs_to_s2_grid input variable overwrite = True to re-run this processing stage, otherwise it will raise an error.

Running this step from the command invite

This step can also be ran from the command prompt. The command fordead obs_to_s2_grid -h will print the help information of this step. For example, to use it with the same parameters, the following command can be used:

fordead obs_to_s2_grid --obs_path <MyWorkingDirectory>/vector/observations_tuto.shp --sentinel_source <MyWorkingDirectory>/sentinel_data/validation_tutorial/sentinel_data/ --export_path <MyWorkingDirectory>/vector/preprocessed_obs_tuto.shp --name_column id

Using Planetary Computer

Running this step using a script

Run the following instructions to perform this preprocessing step:

from pathlib import Path

from fordead.validation.obs_to_s2_grid import obs_to_s2_grid

output_dir = Path("<MyOutputDirectory>")
input_dir = Path("<MyInputDirectory>/fordead_data")

obs_path = input_dir / "vector/observations_tuto.shp"
preprocessed_obs_path = output_dir / "preprocessed_obs_tuto.shp"


obs_to_s2_grid(
    obs_path = obs_path,
    sentinel_source = "Planetary", 
    export_path = preprocessed_obs_path,
    name_column = "id")

NOTE : Set obs_to_s2_grid input variable overwrite = True to re-run this processing stage, otherwise it will raise an error.

Running this step from the command invite

This step can also be ran from the command prompt. The command fordead obs_to_s2_grid -h will print the help information of this step. For example, to use it with the same parameters, the following command can be used:

fordead obs_to_s2_grid --obs_path <MyWorkingDirectory>/vector/observations_tuto.shp --sentinel_source Planetary --export_path <MyWorkingDirectory>/vector/preprocessed_obs_tuto.shp --name_column id

OUTPUTS

The vector file containing only points is written in the file fordead_data/vector/preprocessed_obs_tuto.shp given by the parameter preprocessed_obs_path.

Observations outside of the footprint of the Sentinel-2 data are removed from the resulting vector and their ID printed in the command prompt.

The observations vector's polygons for this tutorial are inside two overlapping Sentinel-2 tiles, T31UGP and T32ULU. THEIA data for this tutoriel only contains data from T31UGP, but if Planetary is used as a source, points will be created for each tile unless tile_selection = ["T31UGP"] is specified for example.

The following figure shows a zoomed view on observation 7 with Sentinel-2 THEIA data.

NOTE : If the initial vector file contains points, their locations are preserved and their id_pixel attribute is set to 1.