Modeling priority areas for restoration of water-related ecosystem services under epistemic uncertainty: a case study in the Atlantic Forest, Brazil

Hydrological models are crucial tools in planning the restoration of water-related ecosystem services because they help target priority areas for efficient resource allocation. However, when planning the expansion of Nature-based Solutions and Payments for Ecosystem Services, there are four key requirements that need to be taken into account. These are (1) the principle of additionality, which states that restoration policies must seek additional gains in terms of ecosystem services; (2) the representation of multiple runoff mechanisms, which can be fundamentally different in nature; (3) the calculation of farm-scale spatial outputs, which allows for the examination of the impacts of management practices at the level of individual farms; and (4) the estimation of epistemic uncertainty, which is the uncertainty that arises due to a lack of knowledge and information.

While addressing these requirements is important for making future planning more effective it can also be challenging. To address this challenge, this paper presents a comprehensive modeling framework that integrates these requirements in a way that allows for an improved selection of top priority areas with farm-scale spatial resolution, and a deeper understanding of how epistemic modeling uncertainty affects the results. This is particularly important when it comes to evaluating the risks of overestimating water-related ecosystem services benefits.

The modeling approach that we propose, called PLANS, uses the design of TOPMODEL to simulate both saturation-excess and infiltration-excess runoff at the farm-scale resolution. It also employs a novel saturation index based on a combination of the Height Above the Nearest Drainage (HAND) and Topographical Wetness Index (TWI) terrain descriptors. To estimate output epistemic uncertainty, we apply the Generalized Likelihood Uncertainty Estimation (GLUE) method, aided by an evolutionary algorithm. We demonstrate the effectiveness of the PLANS model in a case study watershed in the Atlantic Forest biome of Brazil. Our results show that uncertainty can significantly impact the definition of priorities, with a 97% ranking change. We also find that simulated topographic effects can outweigh local effects of land cover and soil type. By better evaluating uncertainty, we demonstrate that the cost of the restoration program in the study case could potentially be reduced by up to 27%, making it more cost-effective.

Overall, our modeling approach offers a promising way to address the challenges of planning the expansion of Nature-Based Solutions in watersheds and deploying programs of Payments for Ecosystem Services. It allows for the improved selection of top priority areas and a deeper understanding of the impacts of epistemic uncertainty on the outputs. By taking these considerations into account, society can make more informed decisions about how to allocate resources and design restoration programs that are both effective and efficient.