In the face of climate change, it is widely agreed that the energy production has to rely on more sustainable and renewable forms of harnessing energy. Offshore wind turbine parks play a crucial role in increasing the share of green energy. This paper explores probabilistic methods of assessing the wind energy potential and potential trends in data collected on Helgoland by considering wind speeds as a Weibull distributed random variable. Further, for forecasting, the monthly expected wind power density is extrapolated by a Gaussian process regression model.
This work was done as part of Data Literacy course at University of Tübingen. This is the repository containing all the relevant figures, source and tex files.
notes
: A directory contains multiple jupyter notebooks used in the analysis. This is where all figures are created.paper
: A directory contains the LaTeX files needed to generate the paper in PDF format. Also, all figures are saved in paper/figutil
: A directory contains python scripts used in the analysis.Makefile
: Used for compiling the tex file to a pdf, deleting the data folder and other tasks. See below for a more in-depth description.
First, setup the a new Python environment using conda. Simply run conda env create -f environment.yml
to setup a new enivornment called dataliteracy
.
Then, run conda activate
and pip install -r requirements.txt
to install all
needed dependencies. For reproducing all the used plots, run the notesbook in
notes/
. Finally, use make pdf
to build the paper's PDF. See below for
other make
commands:
make clean
make pdf
make clean-pdf
The data used in this repository and paper was taken from DWD and can be found here:
https://opendata.dwd.de/climate_environment/CDC/observations_germany/climate/10_minutes/