2The ability of metallic nanostructures to confine light at sub-diffraction volumes in their near-field allows the local enhancement of inherently weak phenomena such as Raman scattering [1], infrared absorption [2] and higher harmonic generation [3]. The decay of these optical excitations, given the large absorption cross-sections and fast electronic relaxation processes, make nanostructured conducting materials efficient local transducers of far-field electromagnetic radiation into mechanical energy. In addition, the strong optical modulation provided by the launched coherent acoustic modes in these nanostructures allow their exploitation as exquisitely sensitive mechanical probes of their near environment. The efficient generation of acoustic waves by nanostructured transducers and the strong self-modulation provided by the launched coherent phononic modes have enabled their application as, for instance, light-source modulators [4], photoacoustic amplifiers [5] and mass sensors [6] [7].The spectrally narrow acoustic modes obtained in these nanostructures are defined by the resonator's constitution and multiple boundary conditions such as size, shape, composition, their substrate and embedding media. This parameter space has been systematically explored where resonances were tuned by changes in adhesion layer thickness [8], mechanical constraints [9], by positioning resonators over trenches [10] and even by mode-interference from a delayed two-pump excitation scheme [11]. Equally importantly, the damping of acoustic vibrations, which defines the spectral linewidth of modes, is a prominent aspect in the application of these systems, and has also received considerable attention, being successfully modelled for infinite isotropic environments [10,[12][13][14][15][16]. However the multiple decay mechanisms in environments without spherical or cylindrical symmetry, such as when particles lie on a substrate, remain poorly understood [17]. In these cases, to determine the damping and quality factors of resonances current practice, [18], is to use empirical fitting of the decaying modulated time-domain signals.Among different contributions to the measured effective damping of a particular phononic mode, the radiation of acoustic waves to the embedding matrix or the substrate, is qualitatively assessed via the mismatch in acoustic impedance (Z) between the vibrating object and its environment. In the longitudinal plane wave limit the impedance is given by Z = ρ j c Lj (being ρ j the j medium density and c Lj the corresponding longitudinal wave speed). However, such a qualitative analysis may be misleading as Z is a mode-dependent parameter, and for which low-damping can be obtained even for perfect impedance match [16]. Accordingly, theoretical calculations have systematically predicted shorter acoustic damping times for particles in solid matrixes and longer damping times for liquid environments when compared to experimental values [18]. Nevertheless, the damping through the coupling of nanostructures to the substrate ...