Summary
This project is generating long-term, harmonised Earth Observation (EO) datasets of nitrogen dioxide (NO₂) and formaldehyde (HCHO) - key compounds driving the formation of ozone, hydroxyl radicals, and secondary aerosols.
Datasets will be adapted for global and regional models to investigate how human and natural emissions of these compounds will influence the Earth’s atmosphere, air quality, and climate, including trends such as declining NOx in North America and rising pollution in Africa. Using inverse modelling, the project will produce top-down emission inventories for 2005–2024, enhancing our ability to track atmospheric changes and better connect observations, models, and policy for more effective air quality and climate action.
Background
The Earth system is largely regulated by land-atmosphere-biosphere-climate interactions. Among those, the emissions of gaseous and particulate pollutants to the atmosphere by human activities (fossil fuel and biofuel use, agriculture, fires) and natural processes (vegetation emissions, lightning, microbial activity in soils) includes a vast array of complex processes, most of which are not well characterised, let alone quantified. Assessing the magnitude, trends and impacts of pollutant emissions on the Earth system represents a major scientific challenge that calls for the use of global observations complemented by high-quality ground-based measurements and advanced models. The integration of Earth Observation (EO) data in models has demonstrated potential for addressing key questions related to the budget and evolution of atmospheric components at global to regional to local scales. The rapid growth of atmospheric monitoring capabilities in combination with improved models hold the potential to narrow down uncertainties in land-atmosphere processes and rationalise the observed trends.
Nitrogen oxides (NOx) and volatile organic compounds (VOCs) are particularly important in tropospheric chemistry; these compounds control the formation of ozone (O₃) and hydroxyl radicals (OH), which in turn affect the oxidizing capacity of the atmosphere and the lifetime of methane, an important greenhouse gas. NOx and VOCs also contribute to the production of secondary aerosols, which affect air quality, radiation, and cloud properties, creating direct links between air pollution and climate.
HCHO, a high-yield product of VOC oxidation, is a proxy for VOC emissions thanks to its relatively short lifetime and global coverage from spaceborne instruments. While short-term variations in biogenic VOC emissions are reasonably captured by current models, their long-term response to drivers such as climate change, land use dynamics, and rising CO₂ levels remain largely uncertain. Observational constraints are limited, making it difficult to assess whether models reliably capture these processes.
Satellite datasets from the OMI (2005–) and TROPOMI (2018–) sounders offer more than two decades of information, but their interpretation is complicated due to differences in instrument design, retrieval algorithms, spatial resolution, and sensor degradation over time. Using reliable, consistent long-term records of NO₂ and HCHO and advanced models, we aim to improve our ability to assess emission changes such as the sharp declines in NOx across the United States and the rapid increases in many African cities.
Aims and objectives
The project aims to address challenges in quantifying emissions of nitrogen oxides (NOx) and volatile organic compounds (VOCs), key drivers of tropospheric chemistry that influence ozone, hydroxyl radicals (OH), methane lifetime, and secondary aerosols.
Harmonised long-term satellite data records of NO₂ and formaldehyde (HCHO) will be developed, ensuring consistent calibration, cross-sensor comparability, and robust uncertainty estimates. These tailored datasets will be designed for direct use in atmospheric models, enabling improved evaluation of long-term variability and trends in observations, but also emissions.
To bridge observational and modelling communities, PIRAMID will generate super-observations of NO₂ and HCHO, specifically adapted for both global and regional model frameworks. These products will support detailed global simulations with the TM5-MP model, providing a multi-decadal reference for assessing the evolution of reactive gases. Over Africa, detailed forward and inverse modelling will be used to disentangle the contributions of anthropogenic, biogenic, fire, soil, and lightning sources. In North America, the emphasis will be on the impact of declining NOx emissions on VOC oxidation and atmospheric chemistry, and on evaluating whether models capture the observed long-term trends.
Using the MAGRITTE model and inverse modelling techniques, PIRAMID will deliver top-down emission inventories of NOx and VOCs spanning 2005–2024, at high spatial and temporal resolution. These will be compared with bottom-up inventories to evaluate their reliability and identify systematic differences.
Finally, the project will assess the added value of satellite-constrained emissions, determining how much they improve model performance and our understanding of long-term atmospheric composition changes. Through dedicated outreach activities, results will be disseminated to the scientific community, policy stakeholders, and the wider public, providing relevant datasets that enhance our ability to track emission trends and their impacts on air quality and climate.
Project plan
The PIRAMID project is organised into a series of work packages (WP) designed to link satellite data records with advanced atmospheric modelling with the aim of understanding long-term trends in NOx and HCHO observations.
We begin by defining scientific requirements (WP10) through consultations with the modelling and data assimilation communities to ensure the satellite products meet user needs. Harmonised and bias-corrected “super-observations” of NO₂ and HCHO are generated and tailored for use in both global and regional models (WP20). These datasets incorporate detailed uncertainty characterisation and cross-sensor consistency checks.
Spatiotemporal analyses will quantify variability and trends in satellite observations, linking them to changes in climate, land use, and emissions (WP31). WP32 will guide best practices for integrating satellite data into models in close collaboration with ECMWF based on the Copernicus Atmosphere Monitoring Service (CAMS) assimilation system. Global simulations with the TM5-MP chemistry transport model (WP33) provide a multi-decadal reference against which regional experiments can be evaluated, as well as improved a priori profiles for OMI and TROPOMI retrievals.
Regional studies will focus on Africa and North America. Over Africa, long-term model simulations (2005-2024) will be compared with the NO2 and HCHO satellite super-observations. Examination of the model-data biases and their temporal evolution will inform us on the stability of the satellite datasets, and therefore on their suitability for long-term studies and inverse modelling. Simultaneous inverse modelling of NOx and VOCs using the MAGRITTE model will quantify the contributions of anthropogenic, biogenic, fire, soil, and lightning sources. This work will deliver top-down emission estimates at high spatial resolution over 2005-2024, which will then be evaluated against available observations (WP34).
Over North America, we will interpret trends in NO2 and HCHO columns and assess the impact of different sources to the model-column intercomparison. We will provide 2-decades of top-down emissions of NOx and HCHO. Inversions of HCHO columns will be conducted to further investigate the week-to-week variability, driven by variations in temperature and clouds, and assess the added value of weekly inversions in this region (WP35). A dedicated comparison of TM5-MP and MAGRITTE results will assess the value of the top-down, satellite-constrained emissions for both Africa and North America (WP36).
Project activities also include an outreach program (WP40), with dissemination through scientific publications, ESA events, a scientific workshop and educational communication.