GNGTS 2024 - Atti del 42° Convegno Nazionale

Session 3.2 GNGTS 2024 Conclusions In this work, we presented a machine-learning based approach to upgrade leachate identfcaton in municipal solid waste landfll trough geoelectrical data. This approach supplies a promising method to reduce the residual ambiguites arising from ERT or TDIP standalone applicatons. We demonstrated two advantages: i) integratng ERT and IP data through clustering analysis can be efectve for mapping the leachate accumulaton zones with more accuracy compared to a traditonal approach; ii) the membership secton derived by using sof clustering can provide a quanttatve validaton of the fnal landfll clustered model, in additon to the available direct informaton. These development over the traditonal approach represents an important step to improve landfll’s drainage operatons, partcularly during the maintenance of MSW landflls. Future work will be focused on applicaton of this method to other landfll sites in order to validate the parameters selected during the cluster analysis, especially with regard to the optmal number of clusters. References Assamoi B., Lawryshyn Y.; 2012: The environmental comparison of landflling vs. incineraton of MSW accountng for waste diversion. Waste management, 32(5), 1019-1030. Bezdek J. C.; 2013: Patern recogniton with fuzzy objectve functon algorithms. Springer Science & Business Media. De Donno G., Cardarelli E.; 2017: VEMI: a fexible interface for 3D tomographic inversion of tme‐ and frequency‐domain electrical data in EIDORS. Near Surface Geophysics, 15(1), 43-58. Oldenburg D. W., Li Y.; 1994: Inversion of induced polarizaton data. Geophysics, 59(9), 1327-1341. Piegari E., De Donno G., Melegari D. & Paolet, V.; 2023: A machine learning-based approach for mapping leachate contaminaton using geoelectrical methods. Waste Management, 157, 121-129. Rousseeuw P. J.; 1987: Silhouetes: a graphical aid to the interpretaton and validaton of cluster analysis. Journal of computatonal and applied mathematcs, 20, 53-65. Soupios P., Ntarlagiannis D.; 2017: Characterizaton and monitoring of solid waste disposal sites using geophysical methods: Current applicatons and novel trends. Modelling trends in solid and hazardous waste management, 75-103. Corresponding author: davide.melegari@uniroma1.it