In this report we present and discuss the data and model fusion methodology used to integrate three information data sources (i.e. Earth Observation (EO), ground-based information and atmospheric modelling) to derive the pollution loading at the ground level at very high spatial resolution. The report provides in addition main results obtained in deriving the estimation of the tropospheric fine particulate loading, namely PM₁₀, for two application case studies: the region of Western Macedonia (Greece) and the metropolitan area of Athens regarding two satellite sensors reflecting different seasonal periods.

The information fusion methodology illustrated includes a tiered approach to processing and generating environmental information:

1. Development of a physical chemistry model and of an optical signal processing model for fusing ground-based measurement data of atmospheric pollution and meteorological conditions with satellite-derived estimates of air quality indicators, such as optical depth of aerosol.
2. Development of a decision theory-based model for fusing the output from the data fusion model with the results of numerical atmospheric pollution modelling accounting for both the dispersion / transport and chemical transformation processes of airborne chemicals in the atmosphere.

The report is organized as follows:

The first part provides a thorough explanation of the methodology usable to integrate different data sources to obtain pollution concentration levels at the ground. The detailed information on each component of the computational methodology is provided in the respective sub-chapter, starting from the AOT calculator to the two different schemes of data fusion techniques developed.

In the second part some concrete results obtained through the application of the developed methodology are provided along with their quantitative assessment in different application sites.

File Attachments: 

WP2.1- Technical report on Data Fusion Methodology FINAL.doc