Reconciling the Attribution of Changes in Streamflow Extremes From a Hydroclimate Perspective

With increasing impacts of climate change and human interventions on the hydrological cycle, attributions of changes in streamflow and its extremes were widely carried out to provide guidelines for adaptation. However, most of them neglected either human-induced climate or hydrological change. Here we propose a reconciled framework to attribute influences from anthropogenic and natural climate change, land cover change and human water withdrawal. A set of 45 years high-resolution (0.01 degrees) land surface hydrological simulations were conducted over a semiarid river basin in North China, forced by changing climate and/or land cover scenarios. There are discernible anthropogenic climate change fingerprints on the changes in annual mean streamflow and wet extremes (ratio of wet months), while land cover change and water withdrawal are more important for the change in dry extremes (ratio of dry months). The impact of human-induced hydroclimate change on streamflow extremes could be underestimated by up to 21-59% or 12-43% based on hydrological or climate attribution alone, suggesting the necessity of the reconciled framework for distinguishing different anthropogenic factors. Plain Language Summary The risk of hydrological extremes is increasing with intensified influences from anthropogenic factors, including human-induced warming, land cover change, and human water consumption, etc. Therefore, attribution of changes in hydrological extremes is critical for a sustainable water resources management. However, most previous studies underestimated human influence by neglecting either anthropogenic climate change, land cover change, or human water use. Here we present a reconciled framework for the attribution of changes in streamflow and its extremes from a hydroclimate perspective. By carrying out a set of high-resolution land surface hydrological simulations driven by CMIP5 climate model scenarios and satellite-observed land cover change data, we found that the impact of human-induced hydroclimate change on streamflow extremes could be underestimated by up to 21-59% or 12-43% based on hydrological (land cover change and water withdrawal) or climate (anthropogenic climate change) attribution alone. Given that anthropogenic factors have irreversible impacts on hydrological system, we call for an integrated hydroclimate attribution for the hydrological extremes, which will provide more reliable guidelines for adaptive managements of water resources and water-hazards in a changing environment.