Evidence and Controls of the Acceleration of the Hydrological Cycle Over Land

Abstract

AbstractInvestigating modifications in the hydrological cycle is essential to understand the impacts of climate change on ecosystems. This study assesses the change in the velocity of the water cycle over land at the global scale, whereas previous studies have mostly focused on changes in the atmospheric water cycle. The hydrological acceleration is quantified by a decrease in average residence time (RT) of water in the first meter of soil. The soil water RT is shown to be sensitive to the soil texture and seasonality of hydroclimatic variables. Despite substantial local variability, most of the RTs are in the range of 50–300 days. The global mean soil water RT declined at a rate of −2.30 and −0.36 days decade−1 (−1.6 to 1.0 days decade−1 the range of nine models) from 2001 to 2020 as measured by reanalysis and CMIP6 simulations for the historical scenario, respectively, which corresponds to −6.8 and −1.1 days °C−1 when expressed per degree of global warming over land. This acceleration is projected to continue at a rate of −1.35 days decade−1 (−3.4 to 0.0 days decade−1 the range of nine models) or −2.2 days °C−1 during the period 2015–2100 under the most extreme emission scenario: SSP 585. Changes in precipitation dominantly drive the acceleration of the terrestrial water cycle compared to changes in evapotranspiration. Rising temperatures and increasing carbon dioxide have opposite effects on the speed of the terrestrial water cycle with compensatory roles keeping RT relatively unchanged in the absence of PR trends.