Analysis of LINET lightning dataset

Published: 23 August 2016| Version 5 | DOI: 10.17632/2p25kmm23j.5
Petar Sarajcev


This is a supplementary material for the following research paper: P. Sarajcev, J. Vaslji, D. Jakus: "Monte–Carlo analysis of wind farm lightning-surge transients aided by LINET lightning-detection network data", Renewable Energy, 99, 2016, pp. 501-513. A dataset of lightning information for a particular region of cca 400 km2 of Croatian mainland, for a calendar year of 2014, has been obtained from the German company nowcast GmbH, which operates the European LINET lightning detection network. This dataset is proprietary and has been obtained for the research purposes only. It can't be shared with third parties due to licensing restrictions (contact nowcast GmbH for more information). This is a Jupyter (IPython) Notebook which contains Python source code, developed for the more-less general analysis of LINET datasets, and applied to the obtained dataset mentioned above. Notebook features seasonal, monthly, weekly, daily, and diurnal analysis of lightning activity. It also features bivariate kernel density estimation of geographical lightning density distribution, wind turbine effective height calculation, wind farm lightning incidence analysis from LINET data, kernel density estimation of lightning amplitudes probability distributions, and more. A simple attempt at lightning flash-cell identification using (density based) clustering techniques is attempted, along with its tracking and nowcasting, where nowcasting is attempted by means of weighted least squares linear regression. Jupyter Notebook has been rendered in HTML as well, and can be viewed without installing Python.



Sveuciliste u Splitu Fakultet elektrotehnike strojarstva i brodogradnje


Source Code