研究目的
To explore the abilities of airborne multispectral images, spaceborne hyperspectral images, and LiDAR data in spatial distribution extraction and aboveground biomass (AB) estimation of Suaeda salsa (S. salsa) for mapping its spatial distribution and AB.
研究成果
The fusion of hyperspectral and LiDAR data provided the highest accuracies for classifying wetland vegetation and estimating aboveground biomass of S. salsa. Multispectral images alone offered a cost-effective alternative with good accuracy. The study demonstrates the effectiveness of combining spectral and structural features for remote sensing applications in coastal wetlands, providing a reference for future monitoring and ecosystem management.
研究不足
The study may be affected by atmospheric, illumination, and land surface conditions. LiDAR data showed limited sensitivity for short and sparse wetland vegetation like S. salsa, and the acquisition of high-resolution remote sensing data, especially hyperspectral images, is costly and challenging. The methods may not be directly applicable to other vegetation types without adaptation.
1：Experimental Design and Method Selection:
The study involved comparative experiments with four combinations of remote sensing data: multispectral images, hyperspectral images, multispectral + LiDAR data, and hyperspectral + LiDAR data. Methods included deriving spectral features (e.g., vegetation indices), structural features (e.g., canopy height model from LiDAR), using support vector machine for classification, and multivariable linear regression for AB estimation.
2：Sample Selection and Data Sources:
Field-collected data included 21 AB samples of S. salsa and over 100 category samples of wetland vegetation species (S. salsa, Spartina alterniflora, Phragmites australis) from Dafeng District, Yancheng City, China, acquired in October 2014. Remote sensing data included airborne hyperspectral and LiDAR data from a Cessna 208B aircraft, and spaceborne multispectral images from Pleiades-1 satellite.
3：Remote sensing data included airborne hyperspectral and LiDAR data from a Cessna 208B aircraft, and spaceborne multispectral images from Pleiades-1 satellite. List of Experimental Equipment and Materials:
3. List of Experimental Equipment and Materials: Equipment included a hyperspectral scanner (SPECIM AISA EAGLE II), LiDAR system (RIEGL LMS Q-680i), GPS unit (Trimble Pro XXR), high-precision balance, steel tape, and software such as ENVI 5.3.1, SPSS 17, and CloudCompare.
4：1, SPSS 17, and CloudCompare. Experimental Procedures and Operational Workflow:
4. Experimental Procedures and Operational Workflow: Preprocessing of hyperspectral and LiDAR data (radiometric calibration, geometric correction, atmospheric correction, point cloud filtering), calculation of vegetation indices and canopy height model, classification using support vector machine, correlation analysis, and development of AB estimation models with cross-validation.
5：Data Analysis Methods:
Statistical analysis using Pearson correlation coefficient, normal distribution tests, accuracy assessment with error matrices, Kappa coefficient, overall accuracy, R2, RMSE, estimation error, RPD, and SSR for model evaluation.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容-
LiDAR System
LMS Q-680i
RIEGL
Collecting LiDAR data with laser wavelength 1550 nm, pulse repetition frequency 200 kHz, spot diameter ≤0.6 m, and sounding density ≥1.5 points/m2.
-
Hyperspectral Scanner
AISA EAGLE II
SPECIM
Acquiring hyperspectral images with spectral range 400-970 nm, 64 bands, spatial resolution 0.8 m, and spectral resolution 10 nm.
-
GPS Unit
Pro XXR
Trimble
Positioning field-collected samples with high precision.
-
Software
ENVI 5.3.1
Exelis Visual Information Solutions Corporation
Used for atmospheric correction, image processing, and classification tasks.
-
Software
SPSS 17
International Business Machines Corporation
Used for statistical analysis, including Pearson correlation coefficient calculations.
-
Software
CloudCompare
Used for point cloud filtering and processing LiDAR data.
-
Satellite
Pleiades-1
Acquiring multispectral images with panchromatic band (0.5 m resolution) and four multispectral bands (blue, green, red, near-infrared at 2 m resolution).
-
Aircraft
Cessna 208B
Platform for airborne data acquisition, flying at speed 240 km/h and altitude 1200 m.
-
High-precision Balance
Weighing wet weight of aboveground biomass samples.
-
Steel Tape
Measuring average height of vegetation samples.
-
登录查看剩余8件设备及参数对照表
查看全部
