In this work, Ambient Noise Tomography (ANT) is applied for a 3D subsurface reconstruction in terms of shear-wave velocity (Vs) beneath one of the main emissive vents of the Nirano Mud Volcanoes site (northern Italy). The adopted workflow is based on short ambient noise records (~1 h) acquired with a small seismic array composed of vertical geophones and implements an automated procedure to capture seismic heterogeneities in the subsoil volume. This procedure can be of general interest when the use of more invasive procedures is hampered by the logistic or economic limitations. In the case here considered a low shear-wave velocity anomaly is identified right below the emissive vent extending vertically from the surface up to approximately 30 m depth. A quantitative model is proposed to explain the correspondence of the low velocity anomaly and emitting vents in terms of increased porosity and saturation of hosting sediments, where deep seated fluids and gas may accumulate close to the surface. This may increase the possible hazard relative to potentially damaging eruptions in the case of a sudden release of the gas stored in this shallow reservoir.