Atmospheric rivers (ARs) are extreme weather events that can alleviate drought and cause billions of dollars in flood damage. Because they transport latent energy towards the poles, they are crucial to maintaining the climate system’s energy balance. While ARs are characterized by long, narrow filaments of water vapor, there is no first-principles definition of ARs grounded in geophysical fluid mechanics. Therefore, AR identification is currently performed by a large array of expert-defined, threshold-based algorithms. The variety of algorithms has introduced uncertainty in the projected future behavior of ARs (O’Brien et. al. 2020) and their resulting floods and droughts. Using the dynamics of the large-scale atmospheric circulation, we propose a physics-based test to constrain future AR projections.

A key property of ARs in boreal winter is that they dominate the meridional transport of latent energy. The magnitude and latitudinal variation of this transport can be rigorously derived from the equations of atmospheric dynamics. We assess whether the aggregate of ARs detected by a given algorithm satisfies this physical property. We conduct this test on AR detections in three-hourly MERRA2 reanalysis and CMIP6 historical and future simulations. We find that AR-induced latent energy transport varies significantly across the ensemble of AR detection algorithms. Using a hierarchy of energy balance models and CMIP6 simulations, we explore the implications of this variation on future changes in poleward energy transport and AR-induced droughts and floods.