Development of a Toolbox to Compare Atmospheric Composition Datasets: Long-term trends in urban NO2 concentrations in Spain derived from CAMS reanalysis and GOME-2 data

Abstract

Satellite and model atmospheric composition data are stored in different platforms, using heterogeneous file formats, varying spatiotemporal resolutions and noncompatible metadata. Comparing these datasets is not a trivial task, but required in data assimilation, validation and mutual coverage studies. This thesis investigates the prevailing methods used to compare sensor observations with data from the forecast and reanalysis system developed by the Copernicus Atmospheric Monitoring Service (CAMS). These are implemented in the development of the first prototype of the Atmospheric Datasets Comparison (ADC) Toolbox. This toolbox, which is the core part of the project, contains a set of tools that facilitate the file interoperability, binning and regridding, computation of levels pressure, conversion of units, application of the averaging kernels, datasets merge, geostatistical comparison and trend analysis. The contribution of this work is twofold: a toolbox is developed to merge and compare atmospheric composition datasets systematic and automatically for any region and time, and its applicability is shown in a case study, where the NO2 emissions in Spain in the last decade are analyzed using satellite and model data.