A global sensitivity analysis of a lake ecosystem model (GOTM-FABM-PCLake) was undertaken to test the impacts of lake morphology on parameter sensitivity in three different lakes. The analysis was facilitated by the Parallel Sensitivity and Auto-Calibration tool (parsac) and included a screening step with the density-based Borgonovo's method followed by in-depth analysis with both Borgonovo's and the variance-based Sobol’ methods. The Borgonovo's method proved efficient in ranking the most influential parameters and its results were corroborated by the Sobol’ method. For total phosphorus and total nitrogen, parameters related to the benthic-pelagic coupling and phytoplankton, were particularly important for the shallower lakes, whereas the most important parameters for total nitrogen were related mainly to the benthic-pelagic coupling in the deepest lake. For chlorophyll a, phytoplankton and zooplankton parameters were most influential. We conclude that lake morphology shapes the parameter sensitivity of lake ecosystem models.

进行了湖泊生态系统模型（GOTM-FABM-PCLake）的全局敏感性分析，以测试湖泊形态对三个不同湖泊参数敏感性的影响。平行灵敏度和自动校准工具（parsac）促进了分析，包括使用基于密度的Borgonovo方法进行筛选步骤，然后使用Borgonovo和基于方差的Sobol方法进行深入分析。事实证明，Borgonovo方法可以有效地对最有影响力的参数进行排名，其结果得到了Sobol'方法的证实。对于总磷和总氮而言，与底栖-上浮耦合和浮游植物有关的参数对于较浅的湖泊尤为重要，而总氮最重要的参数主要与最深湖底栖-浮游耦合有关。叶绿素a，浮游植物和浮游动物参数的影响最大。我们得出结论，湖泊形态影响着湖泊生态系统模型的参数敏感性。