研究目的
To evaluate a long-term wetland conservation strategy by analyzing the changing mangrove distribution in the Zhanjiang Mangrove Forest National Nature Reserve using remote sensing techniques.
研究成果
Mangrove forests in the Zhanjiang wetlands increased by at least 23.7% from 1991 to 2011, with variable rates of change (decrease of 7% from 1991-2000 and increase of 34.3% from 2000-2011). Losses are attributed to deforestation, conversion to agriculture, and erosion, while gains are due to conservation efforts like mangrove plantations and increased awareness. The high turnover among land cover types may result from tidal fluctuations. Satellite methods are affected by tidal cycles, indicating a need for airborne remote sensing for more accurate local monitoring.
研究不足
Satellite-based approaches tend to underestimate mangrove distributions due to effects of local tidal cycles. Limited availability of cloud-free images over the 20-year period restricted data selection. Tidal level fluctuations at image acquisition times may have caused misclassification and imprecise area estimates. Airborne remote sensing is suggested as a more efficient alternative for local-scale monitoring.
1：Experimental Design and Method Selection:
Used remote sensing techniques with multi-temporal satellite data (Landsat TM, ETM+, HJ-1A) to map and detect changes in mangrove forests from 1991 to
2：Applied unsupervised classification (ISODATA algorithm) in ENVI 0 software for image processing, including calibration, atmospheric corrections, and classification. Sample Selection and Data Sources:
20 Study area is the Zhanjiang Mangrove Forest National Nature Reserve in China. Data sources include one Landsat TM image from 1991, one Landsat ETM+ image from 2000, one HJ-1A image from 2011, one digital aerial photograph from 2011, 1:10,000-scale topographical maps, tide level records, and in-situ investigations from 2009 to 2012 with 434 ground control points.
3：List of Experimental Equipment and Materials:
Satellite images (Landsat TM, ETM+, HJ-1A), digital aerial photograph, topographical maps, GPS for field investigations, digital camera, ENVI
4：0 software with FLAASH module for atmospheric corrections. Experimental Procedures and Operational Workflow:
Pre-processed images with georeferencing, atmospheric corrections, and resampling to 30-m resolution. Conducted unsupervised classification to create land cover maps. Performed accuracy assessment using high-resolution aerial photography and error matrices. Analyzed changes in mangrove area over time.
5：Data Analysis Methods:
Used error matrices to calculate overall accuracy and Kappa coefficient for classification accuracy. Quantified changes in mangrove coverage and turnover rates between land cover types.
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
