Resumen
This study focuses on the analysis, implementation and integration of techniques and methods, also based on mathematical algorithms and artificial intelligence (AI), to acquire knowledge of some phenomena that produce pollution with an impact on environmental health, and which start from illicit practices that occur in urban areas. In many urban areas (or agroecosystems), the practice of illegal waste disposing by commercial activities, by abandoning it in the countryside rather than spending economic resources to ensure correct disposal, is widespread. This causes an accumulation of waste in these areas (which can also be protected natural areas), which are then also set on fire to reduce their volume. Obviously, the repercussions of such actions are many. The burning of waste releases contaminants into the environment such as dioxins, polychlorinated biphenyls and furans, and deposits other elements on the soil, such as heavy metals, which, by leaching and percolating, contaminate water resources such as rivers and aquifers. The main objective is the design and implementation of monitoring programs against specific illicit activities that take into account territorial peculiarities. This advanced approach leverages AI and GIS environments to interpret environmental states, providing an understanding of ongoing phenomena. The methodology used is based on the implementation of mathematical and AI algorithms, integrated into a GIS environment to address even large-scale environmental issues, improving the spatial and temporal precision of the analyses and allowing the customization of monitoring programs in urban and peri-urban environments based on territorial characteristics. The results of the application of the methodology show the percentages of the different types of waste found in the agroecosystems of the study area and the degree of concentration, allowing the identification of similar areas with greater criticality. Subsequently, through network and nearest neighbour analysis, it is possible to start targeted checks.