The rapid development of approaches to automatically detect and extract shorelines from satellite imagery is revolutionizing the way we study coastal systems, offering deeper insights into their dynamic responses and adaptive capacities. This study evaluates the performance of the CoastSat toolkit in detecting shoreline retreat and recovery at Praia de Faro beach, South Portugal, following storm Emma in March 2018. The mesotidal, low-tide-terrace beach experienced significant erosion, with an average retreat of 20 m. PlanetScope imagery is used to assess shoreline detection accuracy, due to its higher temporal coverage compared to other open-access satellite constellations. This includes understanding the correspondence between the detected waterline and two water levels: the tide level and the instantaneous water level (combined effect of wave runup and tide). Linear water level corrections are applied to horizontally translate the detected water levels to a reference elevation that enables temporal comparisons. Results suggest that beach morphology strongly influences optimal correction approaches due to the type of beach response to storms, such as profile rotation. Under non-storm conditions, detected waterlines generally represent astronomic tide levels. While tide-level-based corrections improve accuracy during storms, runup-based corrections consistently outperform under non-storm conditions. Consequently, applying corrections that account for slope changes, runup levels during non-storm conditions, and tide level under storms enables the detection of storm impacts and recovery with an accuracy of approximately 6 m in Praia de Faro. These findings highlight the importance of tailoring detection methodologies to the morphological state of the beach as it evolves over time.