Multisensory data is essential for conducting comprehensive and accurate modal analysis in modern structural identification. However, this data is often contaminated with significant noise, which makes effective denoising crucial for reliable analysis. While traditional Multivariate Empirical Mode Decomposition (MEMD) is a powerful tool, it can introduce mode-mixing due to its sifting operations. This mode-mixing leads to inaccuracies in modal identification and decreases the reliability of the assessment of structural conditions. Addressing this limitation is vital for fully utilizing MEMD in structural health monitoring (SHM) and ensuring precise and high-fidelity analysis outcomes. To structure noise separation properly, singular spectrum analysis (SSA) is integrated with the MEMD to alleviate mode-mixing in the resulting modal responses. Unlike other techniques, SSA is data-adaptive and non-parametric, which means it does not require predefined assumptions about the frequency content of the signal. Furthermore, this decomposition process of SSA is based on SVD, which provides a robust foundation for extracting dominant modal components and filtering out noise with high precision. The effectiveness of the integrated MEMD-SSA technique is verified through comprehensive validation studies, including numerical simulations and a full-scale application using the Lysefjord Bridge dataset. This range of studies demonstrates that the technique performs reliably under various real-world conditions, addressing key challenges in SHM. The approaches employed in this study include detecting low-energy frequencies, mitigating mode-mixing from closely spaced modes, and managing significant measurement noise that can obscure modal characteristics. This noise-robust technique focuses on the accuracy and reliability of SHM, leading to improved safety and maintenance. Application to the Lysefjord Bridge dataset demonstrates the ability of this method to manage complex modal behaviour in a full-scale structure which is subjected to various external loadings.