Air quality investigation using functional data analysis methods

A comprehensive analysis of air pollution in six Lithuanian regions reveals no significant differences in particulate matter (PM10) and nitrogen dioxide (NO2) concentrations between the studied groups. The study employs advanced functional data analysis methods, including data smoothing, principal component analysis (PCA), exploratory data analysis, hypothesis testing, and time series analysis, to uncover deep patterns within large datasets. These techniques demonstrate their effectiveness in addressing air quality challenges and enable reliable forecasting for PM10 in regions such as Vilnius Old Town, Vilnius Lazdynai, Šiauliai, and Klaipėda. For NO2, successful forecasting is applicable to Vilnius Old Town, Vilnius Lazdynai, and Šiauliai. This research provides valuable insights into air quality dynamics across Lithuanian regions.

Vitkauskaitė, A., & Salytė, M. (2023). Oro kokybės tyrimas naudojant funkcinės duomenų analizės metodus. Lietuvos Matematikos Rinkinys, 64, 36–53. https://doi.org/10.15388/lmr.2023.33658