Dockerfile based on Sen2cor 2.9
Build
docker build -t ewoc_s2c .
Run
s2c_plan: Process an entire plan
docker run -ti --rm --env-file /home/ewoc_user/env.dev -v /local_folder/work/:/work ewoc_s2c:0.8.4 s2c --verbose v s2c_plan -p /work/21HTC_plan.json --data_source aws --production_id <some_id>
In order to read the json plan you will need to mount a volume to the container
s2c_id: Process one S2 L1C product
docker run -ti --rm --env-file /home/ewoc_user/env.dev ewoc_s2c:0.8.4 s2c --verbose v s2c_id -p S2B_MSIL1C_20190822T105629_N0208_R094_T30SWF_20190822T131655 --data_source creodias --production_id <some_id>
This command will run sen2cor on input L1C S2 and convert the result into ewoc format. The level 1 data is automatically downloaded from the data provider using dataship
The --env-file
is used to environment variables to the ewoc_s2c
container in order to upload the ARD result to s3 bucket.
Options
- The
--data_source
parameter can accept any value compatible withewoc_dag
(aws, creodias, ...) - The
--only_scl
parameter will constraint the sen2cor processing to the Scene Classification map
Sen2cor aux data:
- DEM: srtm tiles are automatically downloaded by
ewoc_dag
from aws public or private S3 buckets - Not used ESA CCI package can be added to this container in order to improve the atmospheric correction. This package is quite heavy for GH (5Gb), so the archive need to be downloaded in the container. Be aware that the docker image size will be around 8 Gb. To use the ESA CCI package, you'll need to download it from here, add it to this folder and uncomment the corresponding lines in the Dockerfile.