era5-heatwaves

ERA5_july-2023_heatwaves.mp4

Analysis and visualisation of ERA5 hourly 2m air temperature for July 2023. This was the warmest month on record and was characterised by multiple, simultaneous heatwaves across the Northern Hemisphere (and winter warm spells in the Southern Hemisphere).

The analysis was done with Python and is documented in the Jupyter Notebook "era5_july-2023_heatwaves.ipynb". Processed temperature data and calculated heatwave metrics are exported to bitmap (PNG) for visualisation with the ClimateArchive engine. The visualisation renders in real time in any web browser using the 3D JavaScript library three.js. The final visualisation is shown at the top.

Input data is too large for GitHub, but can be freely obtained from the sources below:

input data

ERA5 reanalysis

Input data includes:

1. era5_2t_hourly_202307.nc: Hourly data for July 2023.
2. era5_2t_daymean_clim1991-2020_lowpass_07.nc: Daily mean 2m temperature climatology for July based on the 1991-2020 reference period.
3. era5_2t_daymax_clim1991-2020_lowpass_07.nc: Daily maximum 2m temperature climatology for July based on the 1991-2020 reference period.
4. era5_2t_daymin_clim1991-2020_lowpass_07.nc: Daily minimum 2m temperature climatology for July based on the 1991-2020 reference period.

Method used to generate the climatology files:

• The daily mean temperatures are calculated by taking the average of the 24 hourly values for each day, from 00 to 23 UTC.
• The daily max temperatures correspond to the highest hourly value for each day.
• The daily min temperatures correspond to the lowest hourly value for each day.
• A lowpass filter is applied to each daily climatology to remove submonthly variability.

Data can be downloaded from the Copernicus Climate Data Store and is published under a Creative Commons Attribution 4.0 International (CC BY 4.0). More info in the ERA5 documentation.

Gridded Population of the World (GPW), v4

Not essential, only used for a summary metric of the global population affected by the July heatwaves. Data has been used at 15 arc-minute resolution and can be downloaded from SEDAC. The GPW data collection is licensed under the Creative Commons Attribution 4.0 International License.

running the notebook

Easiest way to run the notebook locally is to first download the repo with

git clone https://github.com/sebsteinig/era5-heatwaves

and then install conda (if not installed already). Then create an environment env_name with

conda env create --name env_name --file=environment.yml

using the environment.yml file from this repository to install all necessary python packages. The notebooks can then be run interactively by typing

jupyter lab