Ecological security assessment for environmental restoration and management in SW China

Regional ecological security diagnosis is the most basic prerequisite for environmental management and restoration. However, the traditional single method is greatly affected by human and difficult to quickly identify. Thus, This paper combined principal component analysis and unsupervised k-means clustering algorithm to provide a fast and efficient method for  safety evaluation of environmental management. Basing on this method, the clustering result was defined as five different levels of ecological safety zone combining with the principle of place name + landform + danger degree. The results showed that there were more than half of the study area were unsafe (56.3%), what was even more surprising was that there were many high-value of NDVI and NPP in these areas. The partition result was finally verified in order to ensure the accuracy of the partition, and it had been proved that the ecological environment is not necessarily safe where in the areas with good vegetation coverage. This article provided a new technical reference for the evaluation of ecological areas.