AEROSTG-AIR

The current research/data set deals with environmental data related to the region around the city of Manaus, Brazil, located in the Amazon Basin central region. The research problem encompasses the answers to the questions: :  Is it possible to obtain the spatial distribution of air pollutant concentrations in high resolution based on a set of ground based, airborne and remote sensing observations? The developed high-resolution dataset could be used in future studies to assess relationships between fauna distribution and atmospheric conditions, for example.

To answer these questions, an initial data set was generated, taking into account the main pollutants with sources from NASA/Copernicus satellites, the Amazon Tall Tower Observatory (ATTO) experiment, and the GOAmazon project, along with other local data, such as the position of thermal electric power stations. The timeline of the data collection and treatment considered the period of 2012 to 2016. The output dataset took the time interval between 2014 and 2015 due to the overlap of all data sources.

It is worth noting that the input data was heterogeneous with respect to temporal and spatial resolutions, measurement height, and geographical locations. The current dataset was generated in a grid of 40x40 geographical points, with a grid size of aprox. 5 km, covering a surface area within the boundaries of Latitude -2.00 and -4.25 (South) and Longitude -59.00 and -61.25 (West), with a time resolution of 30 minutes. To produce this dataset, the machine learning (ML) Random Forest (RF) algorithm was selected and used, among other ML algorithms, to fill in the data into this new grid. The features considered were: data from NASA satellites and from the two projects above, covering the pollutants: ozone, NOx, aerosol number concentration (N), isoprene, CO, and acetonitrile. Moreover, distances and bearings to Manaus and 11 thermal power stations nearby, night and day time periods, dry and wet seasons, were also considered to compose the features.

The interpolation produced by the Random Forest algorithm focused the pollutants: acetonitrile, ozone, NOx, Particulate N, isoprene, and CO, which are related to bioclimatic research commonly. In total, a data set was created with 51,148,800 rows, with 44 columns, 38 for the features and 6 for the targets.

Investigators: Andre Luis Ferreira Marques, Felipe Valencia de Almeida, Victor Madureira Ferrari, Renato Okabayashi Miyaji, Lucas de Oliveira Bauer, Márcio José Teixeira, Pedro Luiz Pizzigatti Corrêa, Luciana Varanda Rizzo, Alan James Peixoto Calheiros, Edson Toshimi Midorikawa, Paulo Eduardo Artaxo Netto, Giri Prakash

• Aerosol concentration