- Development of a high-precision drone-based LiDAR and multispectral data fusion workflow for coastal erosion monitoring and shoreline change detection.

Project Abstract

This study presents a high-precision drone-based LiDAR and multispectral data fusion workflow designed for coastal erosion monitoring and shoreline change detection, addressing the critical need for timely, accurate, and cost-effective coastal topography assessments. The methodology integrates lightweight UAV-acquired LiDAR point clouds with high-resolution multispectral imagery to produce temporally consistent 3D coastal surfaces and derived geomorphometric metrics. A robust data acquisition protocol was implemented to maximize point density and spectral fidelity while mitigating common offshore challenges such as wind-induced motion, wave action, and atmospheric scattering. The LiDAR subsystem utilizes time-of-flight measurements to generate dense vertical layers of nearshore topography, including bathymetric penetration where water clarity permits, complemented by multispectral data for semantic labeling of land cover, vegetation, sand, and wetland features. The fusion workflow employs a multi-sensor co-registration strategy informed by ground control networks, feature-based alignment, and error propagation analysis to achieve sub-decimeter vertical accuracy and centimeter-level horizontal precision across repeat surveys. Core processing steps encompass point cloud normalization, coastal DEM/DSM generation, shoreline extraction via hyperspectral indices and edge-detection algorithms, and change analysis through multi-temporal comparison, voxel-based aggregation, and uncertainty estimation. An advanced feature extraction module targets shoreline position, dune morphology, beach toe, dune crest, and nearshore bathymetry where feasible, while incorporating spectral signatures to differentiate sediment textures, moisture regimes, and anthropogenic alterations. To enhance temporal consistency, the workflow includes radiometric calibration, photogrammetric refinement of spectral data, and machine learning-assisted classification for automated delineation of coastal habitats and geomorphic units. The study also develops a comparative framework for error budgeting, sensitivity analysis, and scalability across different coastal environments, ensuring the approach remains transferable to diverse littoral zones with varying tidal regimes and sediment dynamics. Validation experiments were conducted across multiple study sites with disparate coastal typologies, including sandy beaches, dune systems, and cliffed coasts, under varying tidal states and weather conditions. Ground-truth data, in situ tide gauge records, sediment sampling, and high-precision GNSS benchmarks were used to quantify vertical and horizontal accuracies, shoreline positioning bias, and change detection performance over monthly to seasonal intervals. Results demonstrate that the integrated LiDAR–multispectral fusion workflow significantly improves shoreline delineation accuracy, reduces uncertain shoreline migration estimates, and enhances the detection of subtle morphological changes such as dune migration, toe retreat, and cliff face recession. The generated datasets enable stakeholders, including coastal engineers, urban planners, and conservation agencies, to perform risk assessments, shoreline management planning, and climate resilience research with enhanced confidence. The research also discusses limitations related to sensor penetration, atmospheric variability, and computational demands, and proposes avenues for future enhancements, including real-time processing pipelines, autonomous flight planning optimizations, and adaptive algorithms to cope with rapid coastal dynamics.

Project Overview

What This Project Is About

---

A practical project that uses drones to collect two kinds of data about coastlines: precise ground measurements (LiDAR) and camera imagery that captures color and light (multispectral). The goal is to combine these data to track how coastlines change over time, such as erosion or sediment buildup, in a way that is accurate yet affordable for field work.

---

---

The Problem It Addresses

---

Coastal areas change shape due to waves, storms, and human activity. Traditional surveys can be slow, costly, or unsafe. This project aims to provide a faster, safer, and cheaper method to monitor shoreline movement with high accuracy, helping communities plan for erosion, flooding, and habitat protection.

---

---

Objectives of the Project

---

1. Develop a workflow that fuses LiDAR point clouds with multispectral drone imagery.
2. Create a method to convert data into precise coastal elevation and land-cover maps.
3. Quantify shoreline change over time with clear metrics.
4. Assess data quality and error sources to ensure reliable results.
5. Produce a simple user guide for field teams.

---

---

What You Will Do Step by Step

---

1. Review relevant literature and select suitable sensors and flight plans.
2. Plan and conduct drone flights to collect LiDAR and multispectral data over a chosen coast.
3. Process LiDAR data to generate elevation models and extract shoreline features.
4. Process multispectral images to classify land and water and detect vegetation.
5. Integrate datasets to create fused coastal maps and change indicators.
6. Validate results using ground checks or reference data.
7. Analyze trends and produce visualizations and a technical report.

---

---

Expected Outcome

Deliverables include a tested data fusion workflow, coastal change maps, a report detailing accuracy and limitations, and practical guidelines for field deployment and data interpretation.