On the role of patterns in understanding the functioning of soil-vegetation-atmosphere systems

•Patterns may play an important role in improving prediction of processes.•Data-driven approaches provide new insights on information contained in patterns.•Use of pattern information in models may reduce uncertainty. In this paper, we review the role of patterns to improve our understanding of water, mass and energy exchange processes in soil-vegetationatmosphere systems. We explore the main mechanisms that lead to the formation of patterns in these systems and discuss different approaches to characterizing and quantifying patterns. Specific attention is given to the use of data-driven methods to detect patterns in spatial and temporal data that do not make assumptions about underlying statistical properties of patterns. These methods include e.g. decomposition methods, binning based methods, unsupervised classification and temporal stability analysis. We then analyze the value of considering patterns in evaluating model performance, reducing uncertainty in prediction of states and fluxes, as well as for upscaling and downscaling. Finally, we present ways forward to make better use of patterns in the description of flow and transport processes in soil-vegetationatmosphere systems.