little picture - burned area across Europe

Background on this little picture

burned area across Europe 2001-2020 – a streamgraph

Monitoring wildfires from space is crucial for understanding their impact on climate, including the release of greenhouse gases & aerosols that influence the Earth system. This data driven Little Picture illustrates satellite Burned Area data for European countries between 2001-2020. Each country has its own stream. The width of each stream represents the burned area for that year.

https://climate.esa.int/en/little-pictures-gallery/Burned-area-across-Europe/

Data Sources

The CLIP uses the following datasets:

• ESA Fire CCI: MODIS Fire_cci Burned Area Pixel product, version 5.1, Cate dataset identifier: esacci.FIRE.mon.L4.BA.MODIS.Terra.MODIS_TERRA.v5-1.grid.
• cate.ds.data.countries/countries-110m.geojson country polygons provided as public domain from Natural Earth.

Data Preparation

Description

Data preparation comprises the following steps:

• Opening the Fire CCI dataset esacci.FIRE.mon.L4.BA.MODIS.Terra.MODIS_TERRA.v5-1.grid.
• Opening the countries GeoJSON dataset cate.ds.data.countries/countries-110m.geojson
• Select European countries Features from countries dataset.
• Use European countries to spatially subset the Fire dataset.
• Generate annual sums of burned areas of the Fire dataset.
• Group the annual Fire dataset by country codes.
• Generate the sums of burned areas for each country.
• Put burned area sums in a data frame and save a CSV files data/ba-europe-countries.csv and the cumulative sums in data/ba-europe-countries-cumsums.csv. The units are square meters.

One-time script setup

If you do not have a Cate account yet, please visit Cate App, select Cate Cloud Service and register. Using the Cate credentials, login to the Cate JupyterLab. From the JupyterLab's Launcher, open a Terminal window and type

cd ~/work
git clone https://github.com/littlepictures/pytrevalabs.git
git clone https://github.com/littlepictures/clip_05_burning_lands.git

Note you can also use the JupyterLab's Git extension (left side bar) to clone the repos.

Running the script

In JupyterLab's File Browser navigate to /clip_05_burning_lands/scripts/dataprep and open the Notebook extract-burned-areas.ipynb and run it.

To stay in sync with the repos, open a Terminal window and

cd ~/work/pytrevalabs
git pull --rebase
cd ~/work/clip_05_burning_lands
git pull --rebase

Note you can also use the JupyterLab's Git extension (left side bar) to update the repos.

Data Visualisations

• The data is loaded from a CSV file, 'ba-europe-countries.csv', into a pandas DataFrame.

• The DataFrame is then preprocessed by "melting" it, transforming it from a wide format to a long format. This is done to make the data suitable for creating a streamgraph.

• Further, certain countries (Europe, Ukraine, and Belarus) are filtered out from the DataFrame.

• The processed data is visualized using Altair, a declarative statistical visualization library for Python.

• A color scale is defined for all the countries included in the analysis.

• The streamgraph is created, where the x-axis represents the burned area, the y-axis represents the years, and the color represents the different countries.

• The streamgraph is then configured and displayed, showing the trend of burned areas in various European countries over the years.

• A Bauhaus-style poster is created by adding title, logo, and detail texts to the streamgraph. The poster is then configured and displayed.

CREDITS & LICENSE

• Idea by: ESA Climate Office
• Processing Scripts by: Brockmann Consult
• Visualisation by: Ubilabs

The code in this repository is published under CC BY-SA 4.0 license