Related Experiment Video
Updated: Dec 27, 2025
Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation
Published on: February 9, 2024
Estimating reed loss caused by Locusta migratoria manilensis using UAV-based hyperspectral data
Peilin Song1, Xiaomei Zheng2, Yingying Li2
1Institute of Applied Remote Sensing and Information Technology, Zhejiang University, Hangzhou 310058, China; Key Laboratory of Water Cycle and Related Land Surface Processes, Institute of Geographic Sciences and Natural Resources Research, The Chinese Academy of Sciences, Beijing 100101, China; Key Laboratory of Agricultural Remote Sensing and Information Systems, Zhejiang University, Hangzhou 310058, China.
Unmanned aerial vehicle (UAV) remote sensing effectively quantifies locust damage to vegetation. Vegetation indices like NDVI, MSAVI, and GNDVI show high sensitivity for estimating reed loss, outperforming red edge parameters.
Area of Science:
- Agricultural remote sensing
- Ecological monitoring
- Pest management
Background:
- Locusta migratoria manilensis causes significant vegetation and crop damage.
- Traditional satellite remote sensing has limitations in accurately assessing locust-induced vegetation loss due to insufficient temporal-spatial resolution.
Purpose of the Study:
- To develop and evaluate quantitative loss estimation models for locust damage using unmanned aerial vehicle (UAV) hyperspectral remote sensing.
- To compare the effectiveness of various vegetation indices and red edge parameters in estimating vegetation loss caused by locusts.
Main Methods:
- A simulated Locusta migratoria manilensis damage experiment was conducted on Phragmites australis (reed) canopies.
- Hyperspectral imagery was acquired using a UAV over experimental plots with varying locust densities and damage durations.
- Loss estimation models were constructed using vegetation indices (RVI, NDVI, SAVI, MSAVI, GNDVI, IPVI) and red edge parameters (Dr, SD r).
Main Results:
- Vegetation indices, specifically NDVI, MSAVI, and GNDVI, demonstrated higher sensitivity and lower estimation errors (RMSE 8.8–9.1 g/m²) for dry weight loss of reed green leaves.
- Red edge parameters (Dr, SD r) showed inferior performance with higher RMSEs (27.5 g/m² and 26.1 g/m²) compared to vegetation indices.
Conclusions:
- UAV-based hyperspectral remote sensing provides an efficient and quantitative method for assessing locust damage.
- Selected vegetation indices are superior to red edge parameters for estimating locust-induced vegetation loss.
- The developed methodology holds potential for future application with satellite remote sensing for large-scale locust damage monitoring.
