cyclonic-wind-impacts

This repository contains companion code for our study of future changes in European windstorm characteristics under different climate scenarios. If you use the code, please cite the paper:

Little, A.S., Priestley, M.D.K. & Catto, J.L. Future increased risk from extratropical windstorms in northern Europe. Nat Commun 14, 4434 (2023). https://doi.org/10.1038/s41467-023-40102-6

Installation

The code was originally used in an isolated Python 3.8 environment. Besides numpy and pandas, the packages required include:

• xarray: Handles multi-dimensional (e.g., spatio-temporal) arrays
• b2sdk: Provides access to data in B2 cloud storage
• netCDF4: Enables xarray to read NetCDF files
• tqdm: Displays progress bar during analysis

Assuming conda is installed, an environment can be setup in a local version of the repository as follows:

conda create --name=cyclonic-wind-impacts python=3.8
conda activate cyclonic-wind-impacts
conda install --file requirements.txt

The code has only been tested on machines running MacOS 12 (Monterey) and MacOS 13 (Ventura).

Data

We used three sets of input data in our study, all stored on a private encrypted BackBlaze bucket owned by the main author. These must be downloaded to a local repository prior to analysis by running python get_data.py. The ID/key pair needed to access the API are hardcoded into this file; there is no security risk as the full contents of the bucket are non-sensitive and open-access.

Name	Source	File type	Total size
Wind field	ERA5/CMIP6 data nodes (re-gridded)	NetCDF	~650 MB
Tracks	Derived using objective feature tracking	Text	~554 MB
Population	NASA SEDAC (re-gridded)	NetCDF	~35 KB

Full details of the sources and processing steps used to generate these input data are outlined in the companion paper. Notably, the objective feature tracking code belongs to Kevin Hodges and is available here.

Note: There is a daily data bandwidth limit associated with the BackBlaze bucket, as it is operated on a free plan. This unfortunately means that downloads via the API are limited to 1 request per day.

Analysis

The full analysis comprises two main steps, each of which applies a series of purpose-built functions to all individual cyclone tracks in the selected experiment/model combinations:

• Wind footprints: Generate cyclonic wind speed maps
• Storm severity index: Calculate MET-SSI and SOC-SSI maps

One or both steps can be executed as follows (must be in this order):

python generate_footprints.py
python generate_ssis.py

All cyclone maps produced are saved to the outputs/ directory, which is structured similarly to that of data/. Examples can be seen in figures accompanying the main text of the companion paper.

Several key parameters underpinning the model can be configured in params.py. This allows the following tasks to be achieved readily:

• Explore the effects on model outputs when perturbing a single variable
• Reduce the extent of the domain bounding box and/or experiment/model combinations, for shorter computation times

License

This code is licensed under the GNU General Public License v3.0.

Author

Alex Little

Acknowledgements

Thanks to the co-authors for assisting with code structure and development throughout the project:

• Matthew Priestley
• Jennifer Catto