Development of UAV-based LiDAR and multispectral measurement techniques for monitoring sediment intrusion in coastal wetlands

• A new measurement protocol for monitoring sediment deposition in wetlands was developed. • A classification accuracy of 90 % was achieved with UAV LiDAR and multispectral data. • Scalable coastal monitoring in GEE was advanced via measurement science. • Sediment deposition at a coarser 1-meter resolution was detected by the Fine2Scale algorithm. • Algorithms and deep learning methods are included in the supplementary materials . This study presents an innovative measurement method utilizing Uncrewed Aerial Vehicle (UAV)-based Light Detection and Ranging (LiDAR) sensor and multispectral imaging to precisely monitor sand deposition in coastal wetlands. The classification scheme developed in this research distinguishes four key coastal land cover classes: New Sand (recently deposited sediments), Old Sand (previously deposited and partially stabilized sediments), Vegetation (coastal flora covering stabilized areas), and Water (including freshwater and brackish zones). These classes were identified based on spectral and elevation characteristics, enabling a high-resolution analysis of sediment transport and deposition processes. An initial classification using only LiDAR data achieved an accuracy of 83 %. By incorporating multispectral imaging and analysing the spectral signatures for each defined class (from multispectral measurements), we improved the classification accuracy to approximately 90 %. The spectral information from multispectral imaging proved crucial for differentiating sediment types, and refining the measurement process through better model tuning. Data aggregation was conducted using Google Earth Engine, enabling robust repetition of measurements, efficient handling of large datasets, and consistent data processing workflows. Our approach demonstrates significant advancements in measurement science by reducing time and costs associated with post-event environmental monitoring while improving data precision and accuracy. The combination of spectral data with LiDAR and leveraging spectral signatures for model tuning enhanced measurement outcomes by effectively propagating and minimizing uncertainties. The measurement framework developed offers a reliable and precise tool for tracking coastal sediment deposition, ensuring the reliability of results.