中文摘要：

区域生态安全研究对实现区域可持续发展具有重要战略意义。以价值量为基础，采用部分生态估算模型为例，区分出生态系统服务和损失，分析在具体区域使用价值量比值形式表述生态安全程度的可行性和合理性。研究表明：①生态系统服务价值和生态风险损失的时间波动规律不一致，在陕北地区生态恢复背景下生态系统安全程度并未出现直线上升，人类活动和气候变化在地理一生态过程中的交互作用复杂。②使用SOM神经网络基于生态安全度将研究区划分为4个子区域，并进行优化调控讨论。结果说明基于价值量的计算方法可以降低指标权重打分过程产生的不确定性，时空精度高，在实践中可以应用。由于生态估算模型本身还有较大改进空间，因此不确定性仍然存在，多尺度、多方法相互对照评价是后续研究的目标。

英文摘要：

Research on regional ecological security is of great significance in realizing regional sustainable development. However, there are rare studies which can not only synchronously guarantee temporal and spatial precisions but also involve enough evaluation indices at present. Based on the replacement of values, this study adopts a part of the ecological estimation models as examples to distinguish the ecosystem services and eco-risks. Then it analyses the feasibility and rationality for using the ratio of value to express the ecological security degree in a specific area. And the partitions are considered to be more comprehensive than directly zoning by ecological security degree. The analysis of the specific examples in the research shows that： （1） the ecosystem services value and eco-risk loss value are not consistent in the time variation. Under the background of ecological restoration in northern Shaanxi, the tendency of ecological system security degree on the graph is not presented as a straight line. The result shows that human activities and climate changes have complex interactions in the geographical- ecological process. （2） The result of evaluation is focused on the description of ecological security degree with the view of human development. So it is suitable to analyze the ecological processes which are influenced by human activities. The calculation method in this study can reduce the uncertainty produced by marking the weight on the indices, as well as provide a high temporal and spatial precision. By using a variety of year scales or other remote sensing image data of high resolution to guarantee the accuracy, and applying specific models to concrete the composite index, this research contributes to describe the specific spatial pattern and time process reflected by the index, while it still guarantees the geography or ecology meaning at the same time. However, as there is still much room for improving ecological estimation models, the uncertainty has not been eliminated. Further studies o