The role of gap-filling observational data in air quality data fusion methods: a case study with CALIOPE PM2.5

Abstract

Reliable air quality data are vital for informed decision-making, enabling evidence-based mitigation strategies to improve public health and sustainability. Data-fusion methods combining physicsbased air quality models with observational data provide reliable results with full spatial coverage. This study quantifies the impact of imputing missing observational data in these data-fusion methods. We focus on PM2.5 for the Catalonia region during 2019, for which data availability is strongly limited. We first present straightforward gap-filling methodologies, such as linear interpolation and persistence. We then compare these techniques with a state-of-the-art artificial intelligence gapfilling method based on the Gradient Boosting Machine algorithm trained with several years of data (2019, 2021, 2022). To assess gap-filling methodologies, we generate random gaps of varying characteristics identifying the optimal technique for each gap size and availability. Finally, we study how these methods affect the data fusion process applied to the mesoscale air quality model CALIOPE. The PM2.5 output of this system has a horizontal spatial resolution of 1 km x 1 km on a daily scale. The data fusion method uses universal kriging, a geostatistical technique based on a regression model and the spatial correlation between the model and observational data. Data fusion results significantly improve from the raw model estimations, with +24 % and +61 % for the r-value, not using gap-filling of observational data and using it, respectively. Notably, the method’s effectiveness depends on the availability of observations, performing better with GBM-filled data.