EnsemblES - Using ensemble techniques to capture the accuracy and sensitivity of ecosystem service models

如果要实现联合国可持续发展目标（SDG; https：//sustainabledevelopment.un.org/），了解人与自然之间的相互作用至关重要。这些相互作用的一个重要方面可以归类为“自然对人类的贡献”（称为生态系统服务; ES）。然而，量化这些关系所需的经验ES数据在世界各地都很稀少。利用遥感数据提供方面的最新进展，模型越来越能够在缺乏经验数据的情况下提供可靠的信息。具体而言，ES模型生成估计ES的地图（通常基于土地覆盖和其他驱动变量），因此可以提供对空间分布和异质性ES的理解，以帮助规划和优化土地利用决策。然而，大多数ES建模应用程序依赖于每个ES的单个模型，很少有应用程序针对独立的数据集明确验证ES模型。因此，与ES模型（以及支撑它们的数据集）的每次应用相关的不确定性在很大程度上仍然是未知的。这是一个特殊的问题，因为当地尺度验证的结果可能无法转移到新的地点或ES模式输出最广泛使用的区域和国家尺度。EnsemblES试图通过以下方式解决这些问题：1）调查ES模式输入敏感性，在模式模拟开始时改变初始条件; 2）将多个ES模型的输出进行组合（从多个初始条件）使用各种技术转化为模型的“集合”，包括当单个模型性能的数据不可用时;以及3）针对独立数据验证这些模型集合，突出a）ES集合的准确性，以及B）集合的变化系数是否是集合不确定性的良好预测器。

If the United Nations sustainable development goals (SDGs; https://sustainabledevelopment.un.org/) are to be achieved, it is vital to understand the interactions between people and nature. A significant aspect of these interactions can be classed as 'nature's contributions to people' (termed ecosystem services; ES). However, the empirical ES data needed to quantify these relationships are sparse in all parts of the World. Using recent advances in data availability from remote sensing, models are increasingly able to provide credible information where empirical data are lacking. Specifically, ES models produce maps of estimated ES (typically based on land cover and other driving variables) and so can provide the understanding of the spatial distribution and heterogeneity ES required to aid planning and optimisation of land use decisions. However, most ES modelling applications rely on a single model for each ES and few applications explicitly validate ES models against independent datasets. As a consequence, the uncertainties associated with each application of ES models (and the datasets that underpin) them remain largely unknown. This is a particular issue as the results of local-scale validation are likely not to be transferable to new locations or to the regional and national scales at which ES model outputs are most widely used.EnsemblES seeks to address these issues by: 1) investigating ES model input sensitivity, varying initial conditions at the start of model simulations; 2) combining the outputs of multiple ES models (from multiple initial conditions) into 'ensembles' of models using a variety of techniques including when data on individual model performance is vs is not available; and 3) validating these model ensembles against independent data, highlighting a) the accuracy of ES ensembles, and b) whether coefficients of variation of the ensemble is a good predictor of ensemble uncertainty.