Abstract: This paper re-examines the causal impact of the EU Emissions Trading System previously estimated using a Difference-in-Differences approach (Dechezleprêtre et al., 2023) by applying Double Machine Learning (Chernozhukov et al., 2018) to address high-dimensional confounders and potential model misspecification. We re-estimate this impact on carbon dioxide emissions and the economic performance of regulated firms, measured in terms of growth and employment. Using historical data from 2012–2025 across firms in regulated industries, we implement a block k-fold cross-fitting procedure combined with machine learning techniques to reduce overfitting and improve causal inference.

Mayer, N. (2025). "Quantifying Renewable Resource Uncertainty: A Nonparametric and Copula-Based Approach''. Working paper.

Abstract: Renewable energy sources are inherently uncertain due to the stochastic nature of natural resources used as inputs for its production. In this paper, we propose a methodology for uncertainty quantification that combines kernel density estimation for marginal distributions with a copula-based model for multivariate joint dependence. Using 23 years of hourly data, our results show that commonly used parametric approximations fail to accurately represent the empirical density distributions of meteorological variables. With Joe–Clayton copulas, we identify strong seasonal variability in joint tail dependence, with summer having the most pronounced joint extremes between wind speed and sun irradiance. Additionally, meteorological-driven uncertainty is highest during winter daytime and summer night-time. These findings contribute to the assessment of compound weather risks and joint risks associated with renewable energy production surplus and scarcity.

Mayer, N. (2025). “Copula-Based Analysis of Meteorological Interdependencies.” Working paper.

Abstract: In this study, we analyze the dependencies among four key meteorological variables—total precipitation, sun irradiance, air temperature, and wind speed in France using copula-based methods. We apply nonparametric multivariate vine copula to construct the joint probability and estimate the dependence structure among these variables. Our vine structure incorporate time-varying aspect to account for nonstationary condition.

Bachelor of Engineering (Aerospace)'s final year project (2017). This is equivalent of bachelor thesis. I worked with Prof. Fei-Bin Hsiao (Department of Aeronautics and Astronautics, NCKU, Taiwan) during my bachelor degree. The work is on design and optimization of a small turbine blade based (i.e., wind turbine for household uses) on Blade Element Momentum Theory. I used Ansys, MATLAB, and SolidWorks