Modelarea concentrațiilor de particule în suspensie pe platforma de monitorizare a calității aerului eALERT-TRACOM utilizând metoda 4DEnVar [Articol]

Abstract

We illustrate the methodology and robustness of the four-dimensional ensemble-variational (4DEnVar) data assimilation technique. Spatial comparison is provided to evaluate how the assimilation process adjusts the model state toward a true distribution; that is, such spatial comparison is presented between the true PM2.5 concentration values, the background state, and the analysis obtained from the 4DEnVar-like data assimilation experiment. Also, the Pearson correlation quantifies how similarly the particulate matter (PM) concentrations vary over the period of 3-28 February 2025, while diurnal averages of PM1, PM2.5, and PM10 profiles reveal typical daily patterns in particle concentrations, characterized by higher concentrations especially during morning hours and lower levels in the latter afternoon, which can be mainly attributed to traffic-related emissions.