Air pollution and its impact on human health is one of the most important problems facing contemporary society. The main objective of the project is in fact the development and testing of new methodologies for the provision of advanced satellite products for the assessment of air quality from space. In particular, the project aims to capitalise on the information contained in the hyperspectral data acquired by PRISMA to obtain qualitative and quantitative information on the atmospheric particulate load and its type (chemical composition) in urban areas, thus discriminating ‘polluting’ particulate of anthropogenic origin from that of natural origin. This information is key to understanding the impact of the chemical composition of atmospheric particulate matter on environment and health.
Neural network algorithms, or more generally artificial intelligence algorithms, have enormous potential in satellite remote sensing because they are able to identify relationships in the data between input (radiometric measurements) and output (chemical-physical quantities) that are also very weak and strongly non-linear. On the other hand, these mathematical models should not be used, as sometimes happens, with a black-box approach, i.e. without being aware of the physical mechanisms that determine their characteristic parameters.
For this reason, their design must be accompanied by the use of chemical-physical models that can guide the choice of architectures and the correct execution of training phases. To this end, therefore, the proposed project will use a series of state-of-the-art chemical-physical models and databases in the sector. In particular, simulations of aerosol fields with chemical speciation provided by the European Copernicus Atmospheric Monitorng Service (CAMS) will be collected and archived. Downstream of the aerosol field simulation, a post-processing software package (FlexAOD) will be applied to calculate the optical properties of particulate matter from individual simulated aerosol profiles. Finally, a radiative transfer model (e.g. Libradtran or 6S) will use the optical parameters of the aerosol constituents obtained earlier to simulate the radiances and reflectances measured by PRISMA at the top of the atmosphere and corresponding to the specific ‘scenes’ generated by the modelling simulations in the previous steps. It should be noted that the database generated will have global coverage, ensuring the applicability of the algorithms that will be developed outside the test areas envisaged in the project.
The types of aerosols to be identified and quantified in the project’s algorithms are: inorganic particulate matter, organic particulate matter, elemental carbon, mineral dust, and sea salt. This preliminary list may be revised and enriched during implementation. The chemical species listed correspond to those normally simulated in chemistry and transport models and/or estimated through chemical analyses of particulate samples. A marked difference with currently available satellite algorithms lies in the type of aerosol chemical species that will be classified. In the products available for the various satellite platforms, aerosols are in fact usually divided into predefined ‘mixtures’ of the aforementioned chemical species (e.g. continental, marine, desert, urban aerosol, etc.), making direct comparison with models and estimates from in-situ samples difficult.
campioni in-situ.
TEAM MEMBERS
1. University of Rome ‘Tor Vergata’ Department of Civil Engineering and Computer Engineering (DICII)
The Earth Observation group of the University of Rome ‘Tor Vergata’ was founded in 1982 by Prof. Domenico Solimini, one of the founders of the same University and a national pioneer of teaching and research in the field of Earth Observation, as well as the person responsible for numerous international scientific projects. The group, which currently consists of 3 full professors, 2 research fellows, 8 PhD students and a laboratory technician, has been working since the beginning on remote sensing of the atmosphere, especially in the field of microwave radiometry, and has increasingly developed, over time, a particular expertise in inversion problems, not only with microwave data, but also with data from the visible and infrared band. From the end of the 1990s, the group began in particular to develop methodologies for the inversion of remote sensing data based on neural networks.
In the meantime, in 2006, the group founded GEO-K, the first spin-off of the Tor Vergata University, which uses knowledge-based techniques to extract application information from earth observation data. On the research front in this period, neural techniques also began to be considered for processing hypespectral data, in particular those from the CHRIS PROBA satellite mission, and from the AHS and AVIRIS platforms. The results obtained are published in the most important international journals of the sector and presented at various conferences, 7 papers being presented at the ESA Hyperspectral Workshop 2010 plus 2 at the Workshop on Hyperspectral Image and Signal Processing (WHISPERS) in the same year.
Concerning collaborations with ESA, the group currently supports Phi-Lab, the research facility that ESA recently set up to develop artificial intelligence-based technologies in OT applications. The OT laboratory at ‘Tor Vergata’ was also one of the first members of the Copernicus Academy (https://www.copernicus.eu/it/opportunita/il-settore-dellistruzione/copernicus-academy). These activities are complemented by research in other OT sectors, such as microwave electromagnetic modelling, radar reflectometry, oil spill detection at sea from SAR images, SAR interferometry applications, automatic change detection algorithms, and technology transfer. In particular, the group recently concluded a project developed under the Horizon 2020 programme for the in-house realisation of a Fablab dedicated to space applications with a funding received of EUR 260000. In general, the group boasts a significant number of participations, also in a leadership role, in international scientific projects and over 800 papers in journals, books or proceedings of international conferences.
The CETEMPS (Centre of Excellence for Telesensing of Environment and Model Prediction of Severe events, http://cetemps.aquila.infn.it) was set up at the University of L’Aquila in 2001 as a result of a selection by the MIUR for Centres of Excellence, which financed its activities for three years. In the years that followed, CETEMPS managed to finance itself thanks to close collaboration with national public bodies such as the Civil Protection Department, the Abruzzo Region, the Italian Space Agency (ASI) and international bodies such as the European Space Agency (ESA) and the European Science Foundation (ESF).
CETEMPS comprises around 40 people dedicated to intense scientific activity, which has led to numerous publications in leading journals. The centre’s scientific activities are originally mainly focused on meteorological and hydrological observations and forecasting. More recently, CETEMPS’s interests have expanded to include climate modelling and the measurement and modelling of atmospheric composition. The inauguration of the latter activity is largely due to the contribution of the scientific coordinator of the subcontracting unit and extends from the global to the regional and local scales. The activities relevant to the present proposal, in particular, have been supported by funding from 3 Italian Space Agency projects in the years 2006-2015 and 2 national projects in the years 2015-2017.
The interests of the research line in Environmental Modelling, of which PRIMARY is the scientific coordinator, focus on the study of Atmospheric Chemistry and Physics, in particular the understanding of the processes that control the atmospheric balance of chemical species that play an important role in both global climate change and regional air quality. The main tools of investigation are: (1) Eulerian-type chemistry and transport models (CTM), both at global and regional scales; (2) remotely sensed satellite and in-situ observations from ground-based and airborne instrumentation. The excellent experience is also gained through direct collaboration with international research groups of reference, and includes full mastery of the IT tools necessary for the production and analysis of results. CETEMPS uses and directly contributes to the development of three international community models such as the chemistry-transport models CHIMERE and GEOS-Chem and the coupled weather-chemistry-radiation model WRF/Chem.
3. CNR – ISAC
The Institute of Atmospheric Sciences and Climate (ISAC) is CNR’s main institute for research in atmospheric sciences. The scientific activities carried out by the Institute aim at an integrated understanding of processes involving the atmosphere, terrestrial and non-atmospheric, through a multidisciplinary approach that combines scientific and technological capabilities in the fields of meteorology, climatology, atmospheric dynamics, chemical composition, and earth observation. ISAC is organised into 7 territorial units, 7 permanent observatories, which also include 3 Global Stations of the World Meteorological Organisation’s Global Atmosphere Watch Programme and 2 atmospheric supersites in Rome (CIRAS) and in the Po Valley (San Pietro Capofiume). ISAC is internationally recognised through its collaborations with numerous European laboratories and research centres worldwide. An important part of ISAC, most of which can be traced back to one of the three maroareas into which the institute is structured (CAFCA, Atmospheric Composition, Climate Forcing and Air Quality), has been conducting cutting-edge research on atmospheric particulate matter and its impacts on climate, environment and health for decades, at the basis of which there is always a marked focus on the experimental part. In particular, within the ISAC, the study theme of the research group participating in the PRIMARY proposal is the physical-chemical characterisation of atmospheric aerosol, from the local to the regional scale, through the use of active and passive remote sensing observations, from the ground and from space coupled with in situ multi-sensor measurements and numerical modelling.
4. CNR – IIA
The CNR-IIA Institute on Atmospheric Pollution is divided into four sites including the Main Site located in the Rome-1 Research Area in Monterotondo (Rome) and three Secondary Sites. The institute, in support of large public and private sectors, provides high consulting services in the field of air pollution in urban and industrial areas. In particular on the emission, transformation, transport, deposition and circulation of air and environmental pollutants in urban and industrial areas and supports the development of observational and legislative strategies to assess the impact of air and environmental pollution on various ecosystems and at different spatial scales. It promotes the development of international systems and standards for environmental data sharing and the interoperability of systems and infrastructures. The Institute is engaged in the development of integrated environmental analysis systems that involve the integration of observational systems with numerical air quality models in order to develop socio-economic analyses useful for the implementation of European directives and international treaties at both national and international scales. The main research lines related to the PRIMARY project are:
Integration of space and ground observational components (ground-based and in situ) for the study of the atmosphere and air quality.
Acquisition of in-situ data for the calibration and validation of remotely sensed data and derived products (core and downstream services).
Development of algorithms for atmospheric correction of images and for studying radiative transfer in the atmosphere.
Air quality monitoring in urban and rural settings through the use of ground-based and satellite OT data.
5. SERCO
Serco is an international service company that combines technical and commercial know-how with a deep interest in public service. It supports, with its more than 45,000 employees, governments, international organisations, agencies and companies in more than 35 countries around the world, who are looking for qualified partners with a solid track record of providing effective solutions to ensure the delivery of services at high quality standards. Serco has gained experience and expertise in the various IT sectors, covering all market areas where Serco operates, following a single policy of Quality. Our Serco IT Services are provided to Public Administrations, International Organisations, Telecommunications Companies, Organisations operating in the aerospace, urban security and remote sensing sectors.
The IT services we offer cover almost all of our customers’ needs and are divided into:
– Systems Administration Services
– Help Desk services
– Desktop/Laptop/Printers/Hardware support
– Systems Integration Services
– Office Automation Support
– Development Services
– LAN and WAN Network Administration
– Security Services
Serco Italia SpA has been providing services in the Remote Sensing sector to the European Space Agency (ESA) for more than 25 years.
RESULTS
Publications in scientific journals
(Published) Ana Carolina Amarillo, Gabriele Curci, Davide De Santis, Cristiana Bassani, Francesca Barnaba, Samuel Rémy, Luca Di Liberto, Christopher R. Oxford, Eli Windwer, Fabio Del Frate; Validation of aerosol chemical composition and optical properties provided by Copernicus Atmosphere Monitoring Service (CAMS) using ground-based global data, Atmospheric Environment, Volume 334, 2024, 120683, ISSN 1352-2310, https://doi.org/10.1016/j.atmosenv.2024.120683.
Publications in conference proceedings
D. De Santis et al, ‘Air Quality Monitoring at Urban Scale using Prism Hyperspectral Data: The “Primary” Project’, IGARSS 2024 – 2024 IEEE International Geoscience and Remote Sensing Symposium, Athens.
Sarathchandrakumar T. Sasidharan, Davide De Santis et al, ‘AI Feature Extraction for Prism Hyperspectral Data’, IGARSS 2024 – 2024 IEEE International Geoscience and Remote Sensing Symposium, Athens.
De Santis, D., Sasidharan, S. T., Di Giacomo, M., Bencivenni, G., Del Frate, F., Curci, G., Amarillo, A. C., Barnaba, F., Di Liberto, L., Pasqualini, F., Bassani, C., Scifoni, S., Casadio, S., Cofano, A., Cardaci, M., and Licciardi, G.: AI and physical models for air quality monitoring at urban scale with PRISMA hyperspectral data, EGU General Assembly 2024, Vienna, Austria, 14-19 Apr 2024, EGU24-19760, https://doi.org/10.5194/egusphere-egu24-19760, 2024.
D. De Santis et al, ‘The “Primary” Project: Air Quality Monitoring at Urban Scale with Prisma Hyperspectral Data,’ IGARSS 2023 – 2023 IEEE International Geoscience and Remote Sensing Symposium, Pasadena, CA, USA, 2023, pp. 2576-2579, https://doi.org/10.1109/IGARSS52108.2023.10282391 .