Thesis title: Advanced Geophysical Imaging of Landfills with ERT/TDIP Data: Full-Waveform Processing, Stochastic Inversion and Clustering Analysis
Municipal solid waste landfills are complex systems in which leachate accumulation
can generate both environmental and geotechnical hazards. In this context, Electrical
Resistivity Tomography (ERT) and Time-Domain Induced Polarization (TDIP) provide
complementary sensitivity to pore-fluid conductivity and interfacial polarization
processes, thus offering significant potential for non-invasive landfill monitoring.
However, the practical use of these methods is still limited by three major issues:
the poor quality of TDIP data, the lack of rigorous uncertainty quantification in
local deterministic inversions, and the difficulty of integrating multiple geoelectrical
outputs into a single, operationally meaningful interpretation.
This thesis addresses these limitations through three complementary research
lines. The first focuses on the improvement of TDIP data quality by means of fullwaveform
acquisition and advanced signal processing specifically designed for 100%
duty-cycle measurements, enabling a full-decay forward modelling and inversion. The
second investigates uncertainty analysis through a global inversion framework for joint
ERT/TDIP data, with the aim of moving beyond single deterministic models toward
ensembles of plausible solutions and associated statistical indicators. The third
develops machine-learning clustering approaches for the integrated interpretation of
multiparameter tomographic models, progressively moving from hard clustering to
target-oriented fuzzy mapping.
Within the broader framework of environmental engineering, these developments
are relevant because they contribute to the identification, delineation, and monitoring
of leachate-related targets in complex landfill systems, thus supporting more informed
site characterization and management decisions. Rather than proposing a single
unified workflow from the outset, the thesis develops three distinct but converging
methodological directions, provides both advances in applied geophysics and practical
tools for more reliable landfill monitoring and management.