Remote sensing monitoring of wind-preventing and sand-fixing effects of coastal protection forests: A case study in Haitan Island, Fujian, China

Forests play an important role in protecting the coastal environment by slowing down the wind speed and fixing the soil. With the aid of Remote Sensing technology, the monitoring and analysis of the wind-preventing and sand-fixing effects of the coastal protection forests can provide useful and near real-time information for better forest construction and management in such areas. One major issue in previous studies was the somewhat subjective selection of the indicators when establishing the models for evaluating the wind-preventing and sand-fixing effects of the protection forests, which made the research results less convincing and less reliable. Moreover, the analysis of the relationships between the indicators and effects of the evaluation models was also seldom reported. To overcome the aforementioned weakness of previous research, this present paper, taking Haitan Island, China, as study area, presented a new and detailed procedure for indicator selection for coastal protection forests' wind-preventing and sand-fixing effects evaluation and monitoring. Data used for the present study included: one Landsat 7 ETM+ image and one Landsat 8 OLI image, a DEM image with a spatial resolution of 5 meters and the meteorological data of the study area. According to a comprehensive literature survey, six most applicable indicators for the study area, namely, Climate Erosion Force (CE), Bare Sand Area Ratio (BS), Soil Adjust Vegetation Index (SAVI), Surface Roughness (SR), Soil Moisture (SM), and Soil Erodibility (SE) were selected. The Principal Component Analysis (PCA) and the Stepwise Regression Analysis (SRA) were applied to determine the weights for the six different indicators and to construct the evaluation model. By using the established evaluation model, the wind-preventing and sand-fixing effects of the coastal protection forests of the study area in 2002 and 2013 were computed, compared and analyzed.