Sources/Usage: Public Domain. View Media Details

The field of satellite remote sensing has grown considerably since the launch of Landsat 1 in 1972. Despite the increasing reliance on satellite-derived shorelines in coastal monitoring, the absence of benchmarking studies has hindered efforts to precisely evaluate the accuracy of shoreline mapping algorithms. To begin addressing this shortfall, the research team developed a robust framework to compare the performance of five established algorithms using publicly available satellite imagery from Landsat and Sentinel-2.

The study focused on four sandy beaches with a range of wave exposures and wave climates, providing a detailed assessment of algorithm performance under varying conditions. By comparing the results against long-term, in-situ beach surveys, the researchers were able to gauge the accuracy and precision of the shoreline mapping algorithms.

The findings of the study revealed that all five algorithms demonstrated horizontal accuracy on the order of 10 meters at microtidal sites. However, as tidal ranges increased, particularly in high-energy environments with complex foreshore morphology, the accuracy deteriorated significantly. For instance, at the Truc Vert beach in France, which is characterized by complex low tide morphology, the accuracy exceeded 20 meters.

"Our study highlights the critical need for standardized benchmarking in shoreline monitoring,” said Kilian Vos, a researcher at the University of New South Wales, Sydney, and lead author of the study. “By evaluating the performance of existing algorithms across a range of coastal settings, we can identify areas for improvement and ensure the reliability of satellite-derived shoreline data."

One of the key objectives of the newly introduced benchmarking framework is to foster collaboration and facilitate the exchange of knowledge among researchers in the field. By providing an open-source platform for evaluation, the framework aims to encourage innovation and the development of more accurate shoreline mapping algorithms in the future.

"This collaborative approach to benchmarking not only enhances the reproducibility of methods but also serves as a stepping stone for testing and validating new developments in shoreline monitoring," said Dan Buscombe, a USGS contractor at the Pacific Coastal and Marine Science Center and a co-author of the study. “It also encourages people to share their data, by signaling to the community that shared resources such as data and code can be independently evaluated and benchmarked in a systematic and fair manner.”

This study provides an important resource for coastal managers responsible for monitoring sandy coastlines, informing them of the accuracy and reliability of satellite-derived shoreline mapping methods and how they can best make use of them. As coastal areas face growing threats from sea-level rise and extreme weather events, the ability to accurately monitor shoreline changes is increasingly important. 
 

Read the study, Benchmarking satellite-derived shoreline mapping algorithms, in Communications Earth & Environment.