| 摘要: | 桃園藻礁為臺灣沿岸珍貴且具代表性的潮間帶生態地形，兼具地質、景觀與生態價值。然而，藻礁環境長期受到覆沙作用之影響，其礁體分布範圍與表層藻類生長狀態容易隨環境條件而變動。傳統藻礁監測方式多仰賴人工現地調查，調查工作常受限於潮汐時段與天候條件，亦難以同時掌握大範圍且具高空間解析度之變化特徵。
為提升潮間帶藻礁監測之效率與空間解析度，本研究結合無人機（Unmanned Aerial Vehicle, UAV）與多光譜遙測技術，採用P4 Multispectral（P4M）無人機，於桃園大潭 G1 區域進行兩個時期之多光譜影像航拍與分析。研究流程包含影像蒐集、影像前處理、人工圈選藻礁之可視裸露分布範圍、植生指數計算與資料分析等階段，並利用 Pix4Dmapper 影像處理軟體進行幾何與相對輻射校正，建立具地理參考之多波段正射影像（GeoTIFF）。後續於 ArcGIS 地理資訊軟體中計算 NDVI、NDRE、SAVI、OSAVI、MSAVI、NDWI 與 EVI 指數，以比較藻礁與覆沙區域之光譜差異特性。
研究結果顯示，多光譜影像能有效呈現藻礁與覆沙區域之光譜差異，其中紅邊（Red Edge）與近紅外（NIR）波段對藻礁表層反射特徵具較高敏感性；NDVI 與 NDRE 指數於藻礁分布分析中表現相對穩定，適合用於區域尺度之相對比較。
本研究驗證 UAV 多光譜影像應用於潮間帶藻礁監測之可行性，並建立一套具實務可行性之分析流程，可作為未來藻礁長期監測與沿岸生態保育之技術應用參考。;The Taoyuan algal reefs are a valuable and representative intertidal geomorphic feature along the coast of Taiwan, possessing important geological, landscape, and ecological significance. However, the algal reef environment has long been affected by sand burial processes, causing the spatial distribution of reef bodies and the growth conditions of surface algae to vary with environmental conditions. Traditional monitoring of algal reefs relies mainly on field surveys, which are constrained by tidal conditions and weather, making it difficult to simultaneously capture large-area changes with high spatial resolution.
To improve the efficiency and spatial resolution of intertidal algal reef monitoring, this study integrates unmanned aerial vehicle (UAV) technology with multispectral remote sensing. A P4 Multispectral (P4M) UAV was used to acquire multispectral imagery over the G1 area of Datan, Taoyuan, during two survey periods. The research workflow includes image acquisition, image preprocessing, manual delineation of visually exposed algal reef areas, vegetation index calculation, and data analysis. Geometric correction and relative radiometric correction were performed using Pix4Dmapper to generate georeferenced multispectral orthomosaic images (GeoTIFF). Subsequently, vegetation and water-related indices—including NDVI, NDRE, SAVI, OSAVI, MSAVI, NDWI, and EVI—were calculated in ArcGIS to compare the spectral characteristics of algal reef and sand-covered areas.
The results indicate that multispectral imagery can effectively reveal spectral differences between algal reefs and sand-covered substrates. Among the spectral bands, the red-edge (Red Edge) and near-infrared (NIR) bands exhibit higher sensitivity to the surface reflectance characteristics of algal reefs. In addition, the NDVI and NDRE indices show relatively stable performance in analyzing the spatial distribution of algal reefs and are suitable for regional-scale comparative analysis.
This study demonstrates the feasibility of applying UAV-based multispectral imagery for monitoring intertidal algal reef environments and establishes a practical analysis workflow. The proposed approach can serve as a technical reference for future long-term algal reef monitoring and coastal ecological conservation applications. |
