Stationarity of High‐ and Low‐Flows Under Climate Change and Human Interventions Across Global Catchments

Abstract

The assumption of stationarity is fundamental for predicting future hydrologic changes based on historical data. Here we present the first global‐scale, observation‐based assessment of long‐term stationarity in annual streamflow extremes (i.e., maximum and minimum monthly streamflow, or Qmax and Qmin). Observational evidence from 11,069 catchments worldwide reveal that Qmax and Qmin series remain stationary in approximately 93% and 67% of catchments exclusively influenced by climate change, respectively, indicating that climate change alone has not disrupted stationarity in annual streamflow extremes. In contrast, these proportions decrease to 76% and 44% for catchments subjected to direct human interventions, with the prevalence of non‐stationary Qmax and Qmin series generally increasing in tandem with the degree of human interventions. These findings provide valuable quantitative insights into the extent of hydrologic extreme alterations caused by human activities and emphasize the need for adaptive measures to mitigate direct human impacts on the hydrological system.