Example FloodMapper Deployment

This directory contains the minimum set of notebooks and instructions to deploy FloodMapper on a new event. The example here is flooding in the Boulia LGA during March 2023.

1. Define the gridded AoI to map

• 01_Define_AoI_Grid.ipynb

2. Query and visualise available data

• 02_Query_Available_Data.ipynb

3. Download data and create maps

Based on the quick-look analysis in the previous notebook, the following download and mapping commands should be run in a terminal:

# Change to the scripts directory
cd ../scripts

# Download data by specifying a list of LGA names
python 01_download_images.py \
    --lga-names Boulia \
    --session-code boulia_test \
    --flood-start-date 2023-03-12 \
    --flood-end-date 2023-03-25 \
    --ref-start-date 2022-12-10 \
    --ref-end-date 2022-12-16 \
    --path-env-file ../.env

# Start the inference process
python 02_run_inference.py \
    --session-code boulia_test \
    --path-env-file ../.env \
    --collection-name both \
    --distinguish-flood-traces

# Aggregate and merge the predictions into a final flood map
python 03_run_postprocessing.py \
    --session-code boulia_test \
    --path-env-file ../.env

After the postprocessingscript has completed, the final maps will be available on the GCP bucket under the operational/boulia_test folder.

4. Monitor download and mapping progress

Progress of the download and mapping tasks can be monitored using the following notebooks.

• 03_Monitor_Downloads.ipynb

• 04_Monitor_Mapping.ipynb

5. Visualise and validate the flood-extent map

The following notebook loads the floodmap and plots it over the satellite data.

• 05_Validate_Map.ipynb