WISER: Which Ecosystem Service Models Best Capture the Needs of the Rural Poor?

人们普遍认为，贫困农村社区的生计往往高度依赖生态系统服务（ES），尤其是在困难或危机时期作为安全网。然而，理解和管理这些流向穷人的利益的一个主要挑战是缺乏关于穷人对生态系统服务的供应、需求和使用的数据，特别是在对生态系统的依赖往往最高的发展中国家。最近的研究表明，与常用的全球代理（例如利益转移）相关的错误可能很大，因此会造成政策制定或管理干预（例如不正当补贴）的混乱或更糟糕的误导。鉴于这些问题，最近对集成建模平台的改进（在某些情况下基于基于桌面流程的模型）旨在提供当前和未来 ES 分布的改进和动态地图，这对基于 ES 的扶贫干预和政策有很大帮助。虽然这些下一代基于流程的模型似乎在基于 ES 的扶贫工作中发挥着作用，但提供政策相关信息所需的复杂程度和数据需求却知之甚少。例如，尚不清楚即使是最复杂的基于过程的生物物理模型在基于全球可用数据集时是否能够为区域或地方规模的政策决策提供足够准确的信息。同样，也没有尝试量化综合建模平台中受益人分类的必要程度，以提供与最贫困人口相关的管理自然资产的信息。此类分析对于确保下一代模型尽可能高效地产生有用且可信的结果至关重要，也就是说，在数据收集和定制模型开发方面的投资最少。我们将评估当前一系列不同复杂程度的建模方法的有效性，以在撒哈拉以南非洲的多个空间尺度上绘制至少六种生态系统服务——作物生产、储存的碳、水的可用性、非木材林产品（NTFP）、放牧资源和授粉。我们将根据两个广泛的指标来评估模型性能：模型数据要求和决策的有用性。首先，我们将评估每个建模层的数据要求，使用数据可用性、空间分辨率和不确定性对所需输入的强度进行评分。那些具有密集数据需求的模型将获得较差的评分。其次，我们将使用统计二元判别器测试来评估模型在决策过程中的有用性。我们将使用相同的方法，通过将生物物理模型的输出与社会经济指标和模型进行比较（也使用二元判别器测试）来评估受益人考虑对决策的影响。我们在这个项目中的目标是确定需要应用于以有助于扶贫的分辨率绘制 ES 的建模的复杂程度。该项目的研究结果将使决策者能够： 1) 最好地利用现有的 ES 模型，为国家和地区的土地利用/覆盖变化政策提供信息，支持 ES 管理并促进这些服务受益人之间的平等和正义； 2) 设定优先顺序，确定稀缺资源应投入到哪些领域，以改善 ES 的有效管理。因此，WISER 可以通过评估撒哈拉以南非洲地区政策制定者可用的工具来帮助改善该地区约 4 亿贫困人口的生活。

It is widely acknowledged that poor rural communities are frequently highly dependent on ecosystem services (ES) for their livelihoods, especially as a safety net in times of hardship or crisis. However, a major challenge to the understanding and management of these benefit flows to the poor is a lack of data on the supply, demand and use of ecosystem services by the poor, particularly in the developing world where dependence on ES is often highest. Recent work suggests that errors associated with the commonly used global proxies (eg. benefits transfer) are likely to be substantial and therefore confuse or worse, misdirect, policy formulation or management interventions (e.g. perverse subsidies). Given these issues, recent improvements in integrated modelling platforms - in some cases founded on desktop process-based models - which aim to provide improved and dynamic maps of current and future distributions of ES have much to offer ES-based poverty alleviation interventions and policy.While these next generation process-based models appear to have a role to play in ES-based poverty alleviation efforts, the level of sophistication and data needs that is required to deliver policy relevant information is poorly understood. It is, for example, unclear whether even the most sophisticated process-based biophysical model is able to provide sufficiently accurate information for regional- or local-scale policy decision making when based on globally available datasets. Similarly, there has been no attempt to quantify the degree to which disaggregation of beneficiaries is necessary within integrated modelling platforms to provide information on managing natural assets that is relevant to the poorest people. Such analyses are vital to ensure that next generation models produce useful and credible results as efficiently as possible - that is, with a minimum investment in data collection and bespoke model development.We will evaluate the effectiveness of a range of current modelling approaches of varying degrees of complexity for mapping at least six ecosystem services - crop production, stored carbon, water availability, non-timber forest products (NTFPs), grazing resources, and pollination - at multiple spatial scales across sub-Saharan Africa. We will assess model performance based on two broad metrics: model data requirements and the usefulness to decision-making. Firstly, we will evaluate the data requirements of each modelling tier, using data availability, spatial resolution and uncertainty to score in the intensity of the required inputs. Those models with intensive data requirements will be scored poorly. Secondly, we will evaluate the usefulness of the model in a decision-making process using statistical binary discriminator tests. We will use the same approach to evaluate the impact of consideration of beneficiaries on decision making by comparing the biophysical model outputs with both socioeconomic measures and models also using binary discriminator tests. Our goal in this project is to ascertain the degree of complexity of modelling that needs to be applied to map ES at resolutions that are useful for poverty alleviation. The findings of this project will enable decision makers to: 1) best use existing ES models to inform national and regional land use/cover change policies supporting ES management and promoting equality and justice amongst the beneficiaries of these services; and 2) set priorities determining where scarce resources should be invested to improve effective management of ES. Thus, WISER may help improve the lives of the approximately 400 million people living in poverty in sub-Saharan Africa by evaluating the tools available to policy makers in this region.