生态系统是人类赖以生存和发展的基础。全面综合的评价生态系统安全对保障社会经济发展和生态文明建设具有重要意义。本研究选取京津冀生态屏障区的滦河和潮河流域作为研究对象，构建了SWAT与InVEST模型耦合方法，评价了子流域2010年和2015生态安全。为了充分描述生态现状，将土地利用各类型所占比例、水源涵养能力、面源污染负荷和生境质量作为影响生态安全的主要指标，分别分析了各指标空间分布与变化特征。在此基础上使用PSR模型构建了流域生态安全评价指标体系，通过主成分分析与经验估测相结合的方法对评价指标权重进行动态赋值，分析了生态安全分布特征与各层指标值特征。主要结论如下：

（1）研究区超过70%的土地利用类型为林地和草地，分布集中且具有林草交错特征。2005至2015年间，研究区生态环境得到了有效的保护，各土地利用类型变化率较低，土地利用较为稳定。总体上看，研究区景观格局破碎度、空间异质性以及景观优势度有上升的趋势。景观格局趋向于不规则和不均匀。

（2）2005、2010和2015年，研究区水源涵养量分别为16.8亿m3、45.79亿m3和17.05亿m3。水源涵养量能力最强的土地利用类型是草地与林地。从空间分布上来看，研究区氨氮负荷支流区域低，干流区域高。对于总磷来说，2005和2010年总磷负荷自东向西北部递减，然而2015年总磷负荷集中在流域的北部和干流区域。

（3）构建SWAT与InVEST模型耦合方法用于生境质量模拟分析。2010年至2015年间，生境退化度、生境质量和生境稀缺性三个指数数值在空间分布的变化看出，生态结构和功能不稳定，生态环境易遭破坏，生态系统更脆弱，生态安全保障有所降低。面源污染物在水体中的迁移使得空间影响范围增大，面源污染负荷和空间位置不确定性增加，对生境的影响明显增大。

（4）生态安全评分结果可得，研究区生态系统整体稳定，但生态安全状态空间差异较大。2010和2015年研究区总体生态安全等级均为敏感级，其中响应层水源涵养量值由0.14增加至0.29，生境稀缺性值由0.80下降至0.50。水源涵养量值增加说明加大水源涵养能力建设可提高生态服务价值，保障流域生态安全。生境稀缺性值降低说明生态环境受人类活动的影响逐渐变弱，生态结构趋于稳定，生态环境状态趋于安全水平。

    The ecosystem is the basis for human survival and development. Comprehensive assessing the ecosystem safety is essential for ensuring the social and economic development and achieving the ecological civilization goals. The Luanhe and Chaohe river basin which recognized as the important ecological barrier area for Beijing-Tianjin-Hebei region was selected as our study catchment. SWAT and InVEST model coupling method was built. ecological security was evaluated of 2010 and 2015 in sub-basin scale. In order to fully describe the current ecological status, the land use types proportion, water conservation, the non-point source pollution and habitat quality were considered as indices which have important impacts on ecological security. The tempo-spatial characteristics changes for each index were analyzed, respectively. Basin ecological safety evaluation index system was established by PSR model. Through the combination of principal component analysis and experience estimation, the evaluation indices weights were dynamically assigned. Furthermore, spatial characteristics of ecological security and characteristics of index values of each layer were analyzed. The main conclusions are as follows:

    (1) The main land use types are forest and grassland in the study catchment, which covers over 70% of the total area. The distribution feature for forest and grassland area concentrated and interlaced. During the study period, the restricted environmental protection regulation limited the human activities in the study catchment, and results in the low land use change rate. The fragmentation, heterogeneity and landscape dominance of the landscape pattern were increased in this period. Moreover, the distribution of landscape patterns tended to be irregular and uneven.

    (2) In 2005, 2010 and 2015, the water conservation was 1.68 billion m3, 4.579 billion m3 and 1.705 billion m3, respectively. The grasslands and forest present the highest water conservation ability. Generally, the ammonia nitrogen load was low in tributary area and high in mainstream area. In 2005 and 2010, the total phosphorus load was gradually decreased from east to northwest, while, the total phosphorus load was relatively high at northern and mainstream in 2015.

    (3) The coupling method of SWAT and InVEST model which used for analyzing the habitat quality of the study area was built up. From 2010 to 2015, the spatial changes of the three index values, habitat degradation, habitat quality and habitat scarcity, showed that the ecological structure and functions were unstable, the ecological environment and the ecosystem was more vulnerable, and the ecological security guarantee was reduced. The migration of non-point source pollutants in the water body increased the spatial affecting range. The more non-point source pollution load increased, the more habitat was significantly increased.

    (4) The results of ecological security evaluation indicate that the overall ecosystem of the study area was stable, however, there is largely spatial difference in ecological security status. In 2010 and 2015, the ecological security level was identified as sensitive. Specifically, the water conservation value of the response layer increased from 0.14 to 0.29, and the habitat scarcity value decreased from 0.80 to 0.50. The change of water conservation indicated that increasing the capacity of water conservation will increase the ecological services and will ensure the ecological security. Similarly, the change of habitat scarcity indicated that the influence of human activities has gradually weakened. The ecological structure had turned to be stable and the status of the ecological environment had migrated to a safer level.