Airborne Wind Energy Resource Analysis Using the ERA5 Dataset
This repository contains the Python code for analysing the airborne wind energy resource using the ERA5 dataset. The paper has been submitted to Renewable Energy and is available as preprint from https://arxiv.org/abs/1808.07718
An archived version of this code, packaged with the original datasets of the published resource analysis, can be accessed at DOI:10.4121/uuid:646eaf3f-c90b-4f22-89bf-8986804def3c
Step 1: Preparing the Python environment
The code is developed in Python 2.7. It is recommended to use Anaconda for setting up the environment. The following notes are for Linux. We welcome contributions for other operating systems.
Installation using Anaconda
We assume that a version of Anaconda is installed on your machine. Create a new virtual environment using the following command
conda create --name [env_name] python=2.7
replacing [env_name] by a name of your choice. Activate the new environment to use it.
source activate [env_name]
As of Anaconda 4.4,
conda activate [env_name] is the preferred way to activate an environment.
Install the required Python packages by executing the provided bash script install_packages.sh from the user-defined directory [code_dir].
Make sure that the new environment is active before running any of the Python scripts, using either
source activate [env_name] or
conda activate [env_name].
In case you wish to install the Python packages differently, refer to the requirements.txt for the required packages and their versions. In addition, the Climate Data Store API client is required. Install the CDS API using the following pip command.
pip install cdsapi
Step 2: Analysis configuration
In the config.py file, configure the general, downloading, and processing settings. Information on the settings can be found in the script. For more information about the request parameters see the ERA5 catalogue and the ERA5 documentation. It is important to first finish this step before proceeding to downloading the data. E.g. if the target directory path for downloading and reading data files does not exist on your machine the next steps will not be successful.
Step 3: Downloading ERA5 data
The wind resource analysis requires ERA5 wind and geopotential data which can be downloaded using the CDS API. This requires an CDS account and installing the CDS API key. Furthermore, it is required to accept the terms and conditions for using the ERA5 data. The command below can be used to start downloading the wind dataset for the requested year (make sure that the new virtual environment is active). This command should be executed manually for each year of data that you want to download. The size of 1 year of wind data for Western and Central Europe, as used in the paper (coarse grid), is roughly 2 GB. Make sure that there is sufficient disk space available, that you point the parameter era5_data_dir in the config.py file to the correct location and that you download all the years of data as specified in config.py. Note that the downloading is time costly, in the order of magnitude of days. It is important to prevent your machine from going into sleep mode while the downloading. The download script requests 1 month of data at the time. It performs 12 sequential download requests, starting a new download only after the previous one has finished.
python download_wind_data.py 2018
The geopotential data is used in the analysis to determine the surface shape of the earth and to transform the data from barometric altitude to height above ground. Download this data using the following command.
Step 4: Processing the data and plotting the results
The single grid point analysis plots are generated using the command below (make sure that the new virtual environment is active). The location and dataset are set in the main() function. Change the eval_lat and eval_lon parameters to perform the analysis for a different grid point.
Before the wind resource maps can be plotted, the wind resource analysis needs to be performed for the complete grid using the following command. The processing of the dataset as presented in the paper takes easily more than 2 hours. The duration is highly depending on the size of the dataset and the memory of your machine. The script saves the processed data in a new netCDF file.
Subsequently, the wind resource maps can be generated using the following command.
Thank you to Kim Lux for testing the installation and use of this package and for valuable feedback for improvements.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the grant agreement No. 691173 (REACH) and the Marie Sklodowska-Curie grant agreement No 642682 (AWESCO).