Abstract
Despite its great scientific and technological importance, wall-boundedturbulence is an unresolved problem in classical physics that requires newperspectives to be tackled. One of the key strategies has been to studyinteractions among the energy-containing coherent structures in the flow. Suchinteractions are explored in this study for the first time using an explainabledeep-learning method. The instantaneous velocity field obtained from aturbulent channel flow simulation is used to predict the velocity field in timethrough a U-net architecture. Based on the predicted flow, we assess theimportance of each structure for this prediction using the game-theoreticalgorithm of SHapley Additive exPlanations (SHAP). This work provides resultsin agreement with previous observations in the literature and extends them byrevealing that the most important structures in the flow are not necessarilythe ones with the highest contribution to the Reynolds shear stress. We alsoapply the method to an experimental database, where we can identify completelynew structures based on their importance score. This framework has thepotential to shed light on numerous fundamental phenomena of wall-boundedturbulence, including novel strategies for flow control.