The prediction of drift and dispersion of contaminants and biological organisms is a key issue for the sustainable management of coastal environments. However, numerical ocean circulation models are limited by their spatial resolution and struggle to explicitly represent small-scale structures that strongly influence transport processes. In this context, relative dispersion, which measures the separation over time of particles that were initially close together, is a valuable tool for linking unresolved dynamics to observed processes. This project therefore aims to characterize the statistical parameters associated with relative dispersion in the oceanic surface layer, based on the Tracer Release Experiment (TREX) 2020 and 2021 field campaigns conducted in the Lower St. Lawrence Estuary. These campaigns combine passive tracers (rhodamine), surface drifters, ADCP current profiles, and drone-based aerial imagery to estimate turbulent horizontal diffusivity in the surface mixed-layer. To highlight the role of strain in dispersion analysis, three approaches were compared at spatial scales of a few hundred meters: those of Okubo (1971) and Romero et al. (2019), which do not account for background strain, and that of Sundermeyer et al. (2001), which explicitly includes it. The results show that the diffusivity coefficients derived from rhodamine are systematically higher than those obtained from drifters, due to the presence of a density front. Okubo’s empirical law $\ K \propto \ell^{1.15}$ is well followed by rhodamine within the first 180 meters, unlike the drifter data. Furthermore, the methods of Okubo and Romero overestimate rhodamine diffusivity by average factors of 2.6 and 3.1, respectively. These findings emphasize both the limitations of classical approaches, which tend to neglect strain effect, and the marked divergences between passive tracers and surface drifters.