GloUrbEE - Extract GloUrb metrics and indicators from Google Earth Engine

Installation

• Windows

python -m venv env --prompt glourbee
.\env\Scripts\activate
python -m pip install -U pip
python -m pip install -e .

• Linux

python -m venv env --prompt glourbee
source env/bin/activate
python -m pip install -U pip
python -m pip install -e .

Example usage

The notebook.ipynb file contains example of how to use the GloUrbEE tool.

Start the UI

The GloUrbEE-UI allow you to use the main GloUrbEE package workflow with a fancy user-friendly interface.

With streamlit

• Windows

.\env\Scripts\activate
streamlit run ui/00_🏠_HomePage.py

• Linux

source env/bin/activate
streamlit run ui/00_🏠_HomePage.py

With docker

docker run --expose 8501 ghcr.io/evs-gis/glourbee-ui:latest 

The application should be available at http://localhost:8501

Data extracted

Metrics

metric name	description
AC_AREA	Active Channel area (pixels)
CLOUD_SCORE	Percent of the ZONE covered by clouds (%)
COVERAGE_SCORE	Percent of the ZONE covered by the Landsat image (%)
SCALE	Size of a pixel on the selected imagery dataset (meters)
MEAN_AC_MNDWI	Mean MNDWI in the active channel surface
MEAN_AC_NDVI	Mean NDVI in the active channel surface
MEAN_DRY_MNDWI	Mean MNDWI in the surface which is not water
MEAN_MNDWI	Mean MNDWI of the full ZONE
MEAN_NDVI	Mean NDVI of the full ZONE
MEAN_BSI	Mean BSI (Bare Soil Index) of the full zone
MEAN_VEGETATION_MNDWI	Mean MNDWI in the vegetation surface
MEAN_VEGETATION_NDVI	Mean NDVI in the vegetation surface
MEAN_WATER_MNDWI	Mean MNDWI in the water surface
VEGETATION_AREA	Vegetation area (pixels)
VEGETATION_POLYGONS	Number of vegetation patches inside the ZONE
VEGETATION_POLYGONS_p*	Percentiles of the vegetation patches size (in pixels) inside the ZONE
VEGETATION_PERIMETER	Vegetation surface perimeter (projection unit)
WATER_AREA	Water area (pixels)
WATER_POLYGONS	Number of water patches inside the ZONE
WATER_POLYGONS_p*	Percentiles of the water patches size (in pixels) inside the ZONE
WATER_PERIMETER	Water surface perimeter (projection unit)

How the masks are extracted

To extract the water, active channel and vegetation masks, the following expressions are proposed as default parameters depending of the selected imagery. You can define custom ones using the workflow.startWorflow() watermask_expression, activechannel_expression and vegetation_expression parameters. The available layers for expressions are:BLUE,GREEN,RED,NIR,SWIR1,SWIR2,MNDWI,NDWI,NDVI.

Landsat imagery default masks

watermask_expression = 'MNDWI > 0.0'
activechannel_expression = 'MNDWI > -0.4 && NDVI < 0.2'
vegetation_expression = 'NDVI > 0.15'

Sentinel-2 imagery default masks

watermask_expression = 'NDWI > -0.1'
activechannel_expression = 'NDWI > -0.4 && NDVI < 0.2'
vegetation_expression = 'NDVI > 0.15'

Indicators

indicator name	description
occurrence_p*	The frequency with which water was present (JRC Global Surface Water Mapping)
change_abs_p*	Absolute change in occurrence between two epochs: 1984-1999 vs 2000-2021 (JRC Global Surface Water Mapping)
change_norm_p*	Normalized change in occurrence. (epoch1-epoch2)/(epoch1+epoch2) * 100 (JRC Global Surface Water Mapping)
seasonality_p*	Number of months water is present (JRC Global Surface Water Mapping)
recurrence_p*	The frequency with which water returns from year to year (JRC Global Surface Water Mapping)
max_extent	Surface where water has ever been detected (JRC Global Surface Water Mapping)

Credits

Many thanks to:

• Barbara Belletti for the original concept and ideas
• Khalid for the first Python version
• Louis Rey for huge testing
• Julie Limonet for the UI skeleton
• Leo Helling for the Sentinel-2 integration
• Samuel Dunesme

Citation

Dunesme, S., Belletti, B., Helling, L., & Limonet, J. (2024). EVS-GIS/glourbee. UMR5600 Environnement, Ville, Société. https://doi.org/10.5281/zenodo.11235572