Seismic source models, including moment tensors, point forces, and finite-fault representations, allow us to interpret earthquake mechanics, volcanic unrest, and induced seismicity across a wide range of magnitudes and scales. Despite major advances in data availability and methodology, the reliability of inferred source parameters and associated uncertainties remains inconsistently reported and often poorly constrained. This session explores how recent progress in multi-observable source estimation, using dense seismic arrays, distributed acoustic sensing (DAS), and geodetic data such as GNSS and InSAR, can support more robust and reproducible source descriptions. We welcome contributions that improve or introduce source characterization methodologies, particularly for uncertainty quantification from point-source to finite-fault models, integration of emerging datasets, improved stress drop and source resolvability at small magnitudes, and innovative applied or computational approaches, including machine learning and probabilistic methods. We invite studies of source processes of any type or scale, including anthropogenic, tectonic, volcanic, and environmental phenomena.