A Model Integrating Satellite‐Derived Shoreline Observations for Predicting Fine‐Scale Shoreline Response to Waves and Sea‐Level Rise Across Large Coastal Regions

Abstract

Satellite‐derived shoreline observations combined with dynamic shoreline models enable fine‐scale predictions of coastal change across large spatiotemporal scales. Here, we present a satellite‐data‐assimilated, “littoral‐cell”‐based, ensemble Kalman‐filter shoreline model to predict coastal change and uncertainty due to waves, sea‐level rise (SLR), and other natural and anthropogenic processes. We apply the developed ensemble model to the entire California coastline (approximately 1,760 km), much of which is sparsely monitored with traditional survey methods (e.g., Lidar/GPS). Water‐level‐corrected, satellite‐derived shoreline observations (obtained from the CoastSat toolbox) offer a nearly unbiased representation of in situ surveyed shorelines (e.g., mean sea‐level elevation contours) at Ocean Beach, San Francisco. We demonstrate that model calibration with satellite observations during a 20‐year hindcast period (1995–2015) provides nearly equivalent model forecast accuracy during a validation period (2015–2020) compared to model calibration with monthly in situ observations at Ocean Beach. When comparing model‐predicted shoreline positions to satellite‐derived observations, the model achieves an accuracy of <10 m RMSE for nearly half of the entire California coastline for the validation period. The calibrated/validated model is then applied for multi‐decadal simulations of shoreline change due to projected wave and sea‐level conditions, while holding the model parameters fixed. By 2100, the model estimates that 24%–75% of California's beaches may become completely eroded due to SLR scenarios of 1.0–3.0 m, respectively. The satellite‐data‐assimilated modeling system presented here is generally applicable to a variety of coastal settings around the world owing to the global coverage of satellite imagery.

Can Satellite‐Derived Beach Images Resolve the Responses to Human Activities?

“…The greater the water‐level fluctuation is, the flatter the beach is, and the larger the random error is. This explains why the random position errors for the high‐water line at Lingbei Beach are comparable to those in other steep beaches and are slightly better than those at beaches with moderate slopes (Castelle et al., 2021; Hagenaars et al., 2018; Pardo‐Pascual et al., 2018; Vitousek et al., 2023; Vos et al., 2019, 2023; Zhang et al., 2021). The same principle applies when explaining why the random position errors for the low‐water line at Lingbei Beach are relatively large.…”

“…(2021), and Vitousek et al. (2023), while the systematic position error of the low‐water line (corresponding to a gentle low‐tide beach with a slope of 1°) is relatively large. The large positive (seaward‐biased) position errors of the low‐water lines occur because the satellite‐derived waterlines are not the actual waterlines, as light can penetrate seawater and cause the satellite‐derived waterlines to be biased in a seaward direction relative to the actual waterlines.…”

“…(2021) and Vitousek et al. (2023) also found that wave runup is critical for accurately estimating waterline positions. Vos et al.…”

“…(2021) and Vitousek et al. (2023) used the program shared by Vos et al. (2019) to recognize the waterlines of Truc Vert Beach and Ocean Beach, respectively, and the random position errors of these waterlines were 12.8–14 m. Castelle et al.…”

“…Vos et al (2020) also used satellite-derived waterline positions to estimate beach slopes. Castelle et al (2021) and Vitousek et al (2023) used the program shared by Vos et al (2019) to recognize the waterlines of Truc Vert Beach and Ocean Beach, respectively, and the random position errors of these waterlines were 12.8-14 m. Castelle et al (2021) and Vitousek et al (2023) also found that wave runup is critical for accurately estimating waterline positions. Vos et al (2023) evaluated the reliability of five shoreline identification algorithms using measured data from four beaches and reported that these algorithms produced shoreline positions with random errors ranging from 6.9 to 48.3 m. In particular, in Truc Vert Beach, the results were generally unsatisfactory, with random errors ranging from 20.1 to 48.3 m. Zhang et al (2021Zhang et al ( , 2022 proposed the use of a transect-focused approach that integrates automatic and manual subpixel shoreline recognition methods to directly identify the positions of multiple shorelines on transects.…”