Spatial and Temporal Variability of the Mondego River Plume: A Satellite Based Approach

The dynamics of river plumes in coastal environments play a key role in sediment transport, water quality, and ecosystem functioning. This study aims to characterize the spatial and temporal variability of the Mondego River plume (Figueira da Foz, Portugal) through satellite observations, assessing how different methodological approaches influence plume detection and interpretation.

High resolution satellite data, including Sentinel-2 and Sentinel-3 were used as the primary source in this study. Environmental datasets from various sources (SNIRH, Copernicus, IPMA) were used to support image interpretation. The analysis covers the period from 2019 to 2025, including a representative set of cloud free satellite images. Different atmospheric correction approaches, such as Sen2Cor (ESA L2A) and Dark Spectrum Fitting (ACOLITE), were applied to evaluate the impact of atmospheric pre-processing on plume detection and analysis. Several plume detection techniques were tested, including reflectance and turbidity thresholds, spectral distance and unsupervised clustering methods. This analysis was conducted in two phases: an initial phase following well-documented methodologies successfully applied in previous studies in literature, followed by an exploratory approach in which case-specific tests were designed to improve plume detection. 

Results reveal pronounced seasonal and event driven variability in plume extent and dispersion, consistent with variations in river discharge, tides and wind forcing. Differences between atmospheric correction methods lead to variations in plume detectability and estimated spatial extent, highlighting the sensitivity of satellite derived products. Clustering based approaches capture plume morphology and spatial continuity, while threshold based methods provide rapid and easily interpretable estimates of plume extent. 

Overall, this study demonstrates how satellite based plume characterization strongly depends on methodological approaches and pre-processing strategies, providing a reproducible framework for monitoring plume dynamics. This work is relevant for coastal management, while also underlining limitations related to atmospheric conditions and ancillary data quality.