Integrated satellite and drone-based multispectral analysis for 40-Year shoreline reconstruction on the Southern Latium Coast

Coastal zones are widely recognized as among the most dynamic and sensitive geomorphological systems, particularly in response to weather and climate conditions. Coastal erosion and deposition alternate cyclically, influenced by fluvial sediment transport, wave and tidal regimes, sea level rise, tectonics, coastal flooding, and anthropogenic pressures. With approximately 2.15 billion people residing in coastal areas - and projections indicating significant population growth in these zones - understanding shoreline morphodynamics is essential for cost-effective and sustainable management strategies.

Within the framework of Italy’s National Recovery and Resilience Plan (PNRR), funded by Next Generation EU, the extended partnership RETURN (multi-Risk sciEnce for resilienT commUnities undeR a changiNg climate) aims to strengthen research on environmental, natural, and anthropogenic risks associated with climate change. Specifically, the Diagonal Spoke (DS) 8, Science underpinning climate services for risk mitigation and adaptation, focuses on developing innovative models to forecast atmospheric, hydrological, and marine impact-oriented indicators, alongside assessing their uncertainties. In this context, shoreline position and morphology emerge as critical indicators for assessing the impacts of climate change on coastal regions. Italian coastline spans approximately 7,500 km, of which 943 km are currently eroding, and 970 km are prograding (ISPRA, 2023), based on comparisons of shorelines between 2006 and 2020. Reconstructing coastal dynamics in specific study areas is therefore pivotal for effective land management and provides a valuable tool for government agencies and stakeholders.

The southern coastal area of the Latium region (Central Italy) represents an ideal case study for investigating shoreline morphodynamics, with a coastline approximately 30 km long. This study utilizes multispectral and multi-mission satellite imagery from Landsat 4, 5, 8, and Sentinel-2, offering an unparalleled dataset for reconstructing coastal changes. The primary objectives of the research are: i) annual reconstruction of the instantaneous waterline, and ii) identification of erosional and depositional sectors with quantified rates. Using the Normalized Difference Water Index (NDWI), 40 instantaneous shorelines were reconstructed for the summer season from 1984 to 2024. The application of the Digital Shoreline Analysis System (DSAS) developed by the USGS revealed maximum shoreline regression rates of approximately 1 m/yr (1.07 m/yr and 1.2 m/yr, respectively LRR and WLR). Additionally, in winter 2024/2025 drone survey, conducted using a Matrice 350 RTK equipped with a multispectral MicaSense RedEdge-P camera, were integrated into the methodology to provide high-resolution and spatially detailed data on shoreline position and morphology, enhancing the accuracy of the reconstructed coastal dynamics and complementing the satellite-based analyses. Finally, the accuracy of the reconstructed shorelines was validated by comparing satellite-derived shorelines from 1998, 2005, and 2019 with ISPRA’s orthophoto-derived shorelines. The results demonstrate strong agreement, with RMSE of 14.44 m, 12.60 m, and 5.83 (1998, 2005 and 2019, respectively), falling within the uncertainty range of Landsat and Sentinel imagery. This study highlights the potential of multi-sensor remote sensing surveys and geospatial techniques in monitoring coastal dynamics, providing critical insights for climate adaptation and risk mitigation strategies in coastal regions.