Circulation‐type classification derived on a climatic basis to study air quality dynamics over the Iberian Peninsula

Abstract

ABSTRACTA circulation‐type classification is derived for the 1983–2012 climatic period to characterize air quality dynamics over the Iberian Peninsula (IP). Sensitivity tests to automatic classification techniques and other factors affecting the classification (number of patterns, temporal and spatial resolution, domain size, etc.) are performed to objectivize the set‐up that maximizes its quality. The ERA‐Interim reanalysis and the cost733class classification software are used. The identified circulation types (CTs) are described in terms of frequency, persistence, transitions, and location of isobaric systems. The temporal stability of the classification is evaluated following a cross‐validation process that compares the results of the climatic and yearly classifications, leading to the identification of a representative year. Moreover, a representative day for each CT is identified using an objective score that minimizes the absolute value difference of the daily grid with respect to the average CT grid. The reference set‐up uses a cluster‐based technique (Ck‐means) on a daily mean sea‐level pressure database. Six CTs are identified, which are consistent with synoptic patterns found in the literature. Furthermore, the CTs of the climatic period are temporally stable showing similar characteristics in terms of frequency and location of high‐ and low‐pressure systems as those of the representative year (2012). As a first application of the CT classification, 1 h‐maximum NO2 concentration maps from the CALIOPE Air Quality Forecasting System is analysed for the representative days. Synoptic circulation controls the origin and strength of advection explaining transport and NO2 background concentration over the IP. However, there is a strong spatial heterogeneity: in central, northern, and southern IP; NO2 concentrations are controlled by the synoptic circulation, whereas in Spanish Mediterranean coastal areas, a combination of synoptic and mesoscale dynamics explains the NO2 concentration patterns.