| Due to the arid climate, water deficiency and soil erosion, northern China is confronted with several severe challenges, among which land sandification is the main cause for slowing the local economy development and the sand storm frequently raiding Beijing, capital of People's Republic of China. As the first step for sandy land control, sandy land monitoring can provide effective technical support for the decision and sandy land management project. Satellite remote sensing monitoring technique is characterized for its rich information and direct,macro and exact vision, and therefore is an effective tool for sandy land monitoring. Duolun county, a typical farming-grazing transitional region located in Inner Mongolia, is chosen to be the typical research area and careful field investigation and spectral measurement were taken there. Multi-temporal satellite imagery was selected on which integrated sandy land monitoring system was based.1. Sandy land monitoringFor the sandy land information extraction, this study chose appropriate imagery which is able to display maximum spectral difference. Based on elaborate field spectral data and field investigation data, an index was introduced for the sandy land information extraction, and results show the accuracy reaches up to 90.06 percent.2. Sandy land level assessment Unlike the traditional assessment which chose vegetation indicator to represent the sandy land condition, this thesis attempted to integrate the dynamic vegetation indicator with fixed indicators (e.g. soil texture)to assess the current sandy land level. Afterwards, three indicator images are hierarchized and then the composite image were used to make the sandy land level assessment by means of PCA method.3. Automatic inversion of sandy land typeBased on the characteristic of different sandy land type, two ecological models are creatively employed to realize the automatic inversion of different sandy land type.
|
PDF Full Text Request