Perspectives of using a satellite imagery data for prediction of heavy metals concentration
Dubna, JINR, email@example.com, firstname.lastname@example.org, email@example.com
Air pollution has a significant negative impact on the various components of ecosystems, human health, and ultimately, cause significant economic damage. Increased ratification of the Protocols of the Convention on Long-range Transboundary Air Pollution (LRTAP) is identified as a high priority in the new long-term strategy of the Convention. Full implementation of air pollution abatement policies is particularly desirable for countries of Eastern Europe, the Caucasus and Central Asia (EECCA) as well as South-Eastern Europe (SEE). Nowadays the UNECE International Cooperative Program (ICP) Vegetation is realized in 39 countries of Europe and Asia. The goal of this program is to identify the main polluted areas, to produce regional maps and to model the long-range transboundary pollution. The Data Management System (DMS) of the UNECE ICP Vegetation consists of a set of inter-connected services and tools deployed and hosted at the Joint Institute of Nuclear Research (JINR). The DMS elements are to facilitate IT-aspects of all biological monitoring stages starting from a choice of collection places and parameters of samples description and finishing with generation of pollution maps of a particular area or state-of-environment forecast in the long term.
One of the perspective prediction methods is to use data that we can get from satellite images together with sampling data from DMS to learn NN and then use only data from satellite images to predict concentration. There are open programs like LandSat, MODIS, Sentinel 2 that provide free access to their data. One can search their database and find necessary images. Special software such as ENVI or ERDAS can be used to process images. It is not useful because images are of gigabyte size, and we should have few of them to cover the region; some software exists to search through image archives, but the functionality of these programs is rather poor, and they often work with only one satellite data source; it is almost impossible to automate the process, and even if we could, the consumption of resources would be too huge. Alternatively, Google Earth Engine platform can be used. It combines a multi-petabyte catalog of satellite imagery and geospatial datasets with planetary-scale analysis capabilities and makes it available for scientists, researchers, and developers to detect changes, map trends, and quantify differences on the Earth's surface. The platform has advanced mechanisms to search, process and analyze satellite data.
Possibilities of prediction of heavy metals concentration by a special neural network are considered. Sources for neural network learning are satellite imagery from Google Earth Engine platform and environmental monitoring data from UNECE ICP Vegetation Program.