干旱区生态环境知识资源中心

Ecological vulnerability evaluation has important real significance and scientific value. In this Study, under the support of Remote Sensing and Geographic Information System, we use TM images, distribution map of sand desertification and soil salinization, and related geographic information, and adopt a combined landscape pattern and ecosystem sensitivity approach to access the ecological vulnerability of Duerbete County. We consider the following five factors to develop the model: (1) reciprocal of fractal dimension (FD), (2) isolation (FI), (3) fragmentation (FN), (4) sensitivity of sand desertification (SD), and (5) sensitivity of soil salinization (SA). Then we build the evaluation model and calculate the vulnerability of landscape type of Duerbete. Through Kriging interpolation, we get the regional eco-environment vulnerability of whole county. Then we evaluate this cropping-pastoral interlacing region-Duerbete County. The conclusions are: (1) The vulnerability of all landscape types is in the following decreasing order: grassland > cropland > unused area > water area > construction area > wattenmeer > reed bed > woodland > paddy field; (2) There are significant positive relationships between VI and FN, VI and SD, SD and FN, SA and FN. This suggests that FN and SD have considerable impact on the eco-environmental vulnerability; (3) Willi the combination of FN, SD and SA the regional eco-environment vulnerability can be evaluated well. The result is reasonable and can support ecological construction.