Clustering Future Scenarios Based on Predicted Range Maps

TL;DR

This paper introduces a spectral clustering method for future biodiversity scenarios based on predicted species range maps, emphasizing ecological impact over climate model differences to better interpret climate change effects.

Contribution

It presents a novel two-step, interpretable clustering approach that incorporates ecological data to distinguish meaningful future biodiversity trajectories under climate change.

Findings

Clustering based on ecological impact is mainly driven by warming levels.

Clustering based on climate features is mainly driven by climate models.

Incorporating ecological information improves understanding of biodiversity loss.

Abstract

Predictions of biodiversity trajectories under climate change are crucial in order to act effectively in maintaining the diversity of species. In many ecological applications, future predictions are made under various global warming scenarios as described by a range of different climate models. The outputs of these various predictions call for a reliable interpretation. We propose a interpretable and flexible two step methodology to measure the similarity between predicted species range maps and cluster the future scenario predictions utilizing a spectral clustering technique. We find that clustering based on ecological impact (predicted species range maps) is mainly driven by the amount of warming. We contrast this with clustering based only on predicted climate features, which is driven mainly by climate models. The differences between these clusterings illustrate that it is crucial to incorporate ecological information to understand the relevant differences between climate models. The findings of this work can be used to better synthesize forecasts of biodiversity loss under the wide spectrum of results that emerge when considering potential future biodiversity loss.