Given the strong impact of air quality on health, environment, and economy, Morocco has implemented an air quality network to assess air pollutants including PM10 (particulate matter with a diameter less than 10 μm). This network which is composed of 29 fixed measurement stations is spatially limited and does not provide sufficient time resolution. The scarcity of measured air quality data led to seek an optimal alternative source to conduct related data-based studies. This represents the primary objective of this paper. PM10 concentrations of global Copernicus Atmosphere Monitoring Service Reanalysis (CAMSRA) data (4D Variational analysis "4v" and analysis "an"), as well as regional CAMSRA data, were examined against the average daily PM10 concentrations collected from six fixed Moroccan air quality measurement stations in 2016 (i.e., observation data). The verification is carried out by studying and analyzing seasonal, extreme, and annual values. The study shows a strong seasonal dependence with a positive bias in winter and a negative bias during summer. For the study of extreme values, global CAMSRA "an" and "4v" data record significant bias of approximately 184 and 161 μg/m3, respectively. However, the annual analysis shows that the CAMSRA global "an" data have the smallest average bias (20.008 μg/m3) and hence has the closest representation of observation data. We conclude that the CAMSRA global analysis data could be used to compute climatology, study trends, evaluate models, benchmark other reanalysis, or serve as boundary conditions for regional models for past periods.
Keywords: Air quality data; CAMSRA; Data assimilation; PM10.