Atmospheric rivers (ARs) are critical to the western US water cycle often acting as both drought busters and flood generators. An outstanding scientific and decision-relevant question is how landfalling AR characteristics will respond to different global warming levels and how those changes might influence water resource management. To meet this need, we investigate coastal western US AR characteristic changes across several stabilized global warming scenarios and identify how shifting AR conditions alter water management outcomes and, more specifically, their annual flood damage potential. To do this, we combine a new AR detection workflow (which leverages TempestExtremes) with an ensemble of uniform, high-resolution (0.23 degree x 0.31 degree) Community Earth System Model simulations (CAM5.3-FV, CLM4.0, and AMIP protocols) to estimate a world that might have been (+0 degree Celsius since industrial revolution [IR]), a world that corresponds to present day warming (+0.85 degree Celsius since IR), and several future worlds (+1.5, +2, and +3 degrees Celsius since IR). We show that the number of water management relevant landfalling ARs increase with every degree of warming (19/year to 24/year), a subtle, but important, uptick in the number of flood inducing ARs occurs (2% to 8% of ARs), and, ultimately, for every degree of warming annual average flood damages associated with ARs increase by ~$1 billion. We also highlight the importance of model ensembles in assessing AR related impacts and how even with ~60 simulated years return period frequencies of the most hazardous ARs may not yet convergence (e.g., southern California).