In the last two decades, several devastating wave-induced events such as the 2004 Indian Ocean tsunami, 2011 Tohoku tsunami and the 2018 Palu tsunami, demonstrated the need for reliable information on the magnitude and recurrence interval of such events. Sediment archives are a major source of data on extreme palaeoenvironmental events. However, despite a considerable progress in tsunami research, the multi-proxy characterization of tsunami deposits for reliable identification of changes in sediment provenance remains a major scientific challenge because it provides insights into source materials and their environmental context at the time of the event. The increasing reliance of such studies and the need for robust identification and characterization of sediment deposits, motivated the research of reliable, time-saving analytical tools and procedures. Here we compare four rapid (1sec – 2min), non-destructive and accurate spectroscopic technics for semi-quantitative mineralogical, geochemical and bio-chemical characterization of sediments. Portable visible spectrocolorimetry, benchtop mid infrared spectroscopy (Fourier Transform attenuated total reflectance and diffuse reflectance) and portable X-Ray fluorescence were used to characterize and discriminate high energy wave-induced sediment deposits. Sediment deposits were collected from 7 ka year record of high-energy deposits recorded in a coastal cave along the north-western coast of Sumatra (Indonesia). Compared to more conventional technics such as atomic absorption spectroscopy, inductive coupled plasma mass spectroscopy, scanning electron microscopy, visible (VIS) and infrared spectroscopy (MIR) analytical approaches used in this study are faster and less costly. Compared to conventional X-Ray diffraction, MIR spectroscopy is not subject to analytical biases introduced by the differential settling or preferred orientations of minerals, allows identifications of amorphous phases and only requires small samples (> 20 mg) and brief measurement times (~1 min). Due to its wide range of wavenumbers (VIS, 400-700 nm; MIR, 4000-450 cm-1), spectra are highly discriminant and allow strong source to sink determination using multivariate statistics and machine learning. The results show strong discrimination between the subsequent high-energy events, suggesting significant environmental differences of source materials and their palaeoenvironmental (inland, offshore) and geomorphological context (lagoon-mangrove, dune, beach etc.). The advantages, the limits and the wider significance of the methods are discussed.