Introduction

Dune ecosystems are dynamic landscapes shaped by geological, human, and climatic factors. Utilizing Unmanned Aerial Vehicles (UAV), this study investigates the Manila Dunes in Humboldt County, California, focusing on how vegetative stabilization, social trails, and invasive species influence dune movements. Through advanced remote sensing techniques, the research aims to shed light on the interactions between human activity, vegetation density, and dune dynamics, offering insights for more effective coastal management and conservation strategies.

Methods

This research employed UAVs equipped with high-resolution cameras to capture detailed imagery of the Manila Dunes, covering a 22.5-acre plot. The imagery underwent a photogrammetry process to create an orthomosaic image, providing a detailed overview of the area's topography and vegetation. The study utilized object-based feature extraction and pixel-based supervised classification methods to analyze the dune vegetation, focusing on distinguishing invasive from native species. Additionally, the analysis included monitoring dune movement through comparison of UAV-derived digital surface models across different years, aiming to assess the impact of trails and vegetation on dune stability.

Results

The findings reveal that established trails contribute to reducing dune movement, contrasting with social trails, which exhibit more local movements. The research successfully identified areas of erosion and deposition within the dunes, highlighting the stabilizing effect of vegetative cover against dune movements. Furthermore, the study compared two classification methods to map dune vegetation, finding that object-based feature extraction offered a more accurate identification of invasive and native species compared to pixel-based classification.

Discussion

The study emphasizes the complexity of coastal dune ecosystems, where human activities, such as the creation of social trails, can significantly impact dune dynamics and vegetation distribution. The results underscore the importance of integrating remote sensing technologies and machine learning methods in environmental management practices. By providing a baseline of dune movement and vegetation distribution, the research supports the development of informed conservation strategies, aiming to preserve these dynamic landscapes against the challenges posed by invasive species and human disturbances.