海岛生态系统承载力不确定性分析——以庙岛群岛南部岛群为例

从人-自然耦合的复合生态系统角度，考虑社会经济发展与生态系统之间的交互作用，借鉴指标体系概念框架，提出了包含海岛基础承载力（FCCI）和海岛现实承载力（RCCI）两个综合指数的海岛生态系统承载力（EBCCI）评估方法，并以北方典型海岛庙岛群岛为例，评估了庙岛群岛南部岛群生态系统承载力。为提高评估结果的准确性，提出了参数导致评估结果不确定性的计算方案，给出了不同置信度下不确定性度量方法。结果显示，在95%的置信水平上，庙岛群岛南部岛群RCCI得分区间为[0.6908，0.7348]，不确定度为4.4%，FCCI得分区间为[0.7949，0.8417]，不确定度为4.68%，社会支持能力（SSCI）得分区间为[0.8612，0.8863]，不确定度为2.51%，社会经济发展较好地支持了基础生态功能的实现。通过引入对模型参数的不确定性分析，提出了一种新的承载力估计方法，用区间代替单一值来衡量生态系统承载力，使结果更加准确可信。

Uncertainty analysis on ecosystem-based carrying capacity of island: A case study in the Southern Island Group of Miaodao Archipelago

Using the conceptual framework of indicator system for reference, a method was proposed in this paper to evaluate the ecosystem-based carrying capacity of island (EBCCI) from the perspective of human-nature coupled ecosystem.The fundamental and realized carrying capacities of island (FCCI and RCCI) were integrated to evaluate EBCCI by emphasizing the interdependences between natural system and social system.A case study was conducted in the Southern Island Group of Miaodao Archipelago, the typical archipelago of the north of China located in the intersection of the Yellow Sea and the Bohai Sea.To improve the accuracy of evaluation, an uncertainty analytical method based on varying model parameters was proposed, and the measurement of the uncertainty level was also proposed corresponding to different confidence levels.Results showed that the uncertainty level of the RCCI was 4.4%, of FCCI was 4.68%, and of SSCI was 2.51%, and the value range of RCCI was 0.6908 to 0.7348 out of 1, of FCCI was 0.7949 to 0.8417 out of 1, and of SSCI was 0.8612 to 0.8863 out of 1, respectively, both with confidence levels of 95%, which showed that the socioeconomic development supported the realization of fundamental carrying capacity well in Southern Island Group of Miaodao Archipelago.A brand new carrying capacity evaluation method was proposed in this paper by introducing the uncertainty analysis of model parameters.A value range of carrying capacity instead of a single value was given through the uncertainty analysis based on the model parameters, which made the results more credible.