Air quality parameters outliers detection using functional data analysis in the Langreo urban area (Northern Spain)

Polluted air of cities is a harmful factor to health that may eventually cause respiratory problems and cardiovascular disease. The monitoring and control of pollutants is an essential activity in order to protect the environment and the health by minimizing pollution levels through the detection of contaminants. Contaminants are emissions of substances to the atmosphere (mainly gases and particulate matter) whose values are greater than the limits allowed by the environmental legislation (they are anomalous values). Thus they are considered as vector samples where each component represents the gas concentration value in the air. In this sense, a model based on functional analysis has been implemented for the outliers detection in air quality samples in this research work. This model transforms the vectorial sample by creating a new functional sample in order to determine functional outliers by adjusting the concept of depth to the functional event. This method has been compared to classical outliers analysis from a vectorial point of view, emphasizing the power of use of such functional techniques over the traditional ones. The main aim of this research work is to compare the results corresponding to the classical and the functional methods and to obtain the most appropriate methodology to analyze this type of dataset in order to reach a better solution for the air quality control.