Automatized outlier screening applied to the AirBase 2001-2010 NO2 and PM10 time series : evaluation of background stations from AT, BE, BG, CH, CZ, DE, ES, FI, FR, GB, HU, IT, MK, NL, PL, PT, RO, and SK

In this report, we present the further development of a screening tool for the automatic recognition of anomalies in air quality monitoring data, which considers both attribute values and spatio-temporal relationships. The screening tool was originally designed for the evaluation of PM10 data. Special focus was directed to the necessary program extensions for extending the toolset to another pollutant (NO2). An application example of identification of anomalies in the AirBase 2001-to-2010 time series of PM10 and NO2 datasets is demonstrated. Furthermore, an extensive catalogue of all 1964 (PM10) and 2272 (NO2) time series plots of daily abnormal values for the years 2001 to 2010 for all background stations of 18 selected countries (AT, BE, BG, CH, CZ, DE, ES, FI, FR, GB, HU, IT, MK, NL, PL, PT, RO, and SK) is provided in the appendixes. The implemented method can be of interest as the basis of a data quality screening system, as well as a data preparation step when comparing measurement data to modelling results. Furthermore, it can also provide a simple solution towards the investigation of the consistency of station classification in AirBase.