Abstract Atmospheric River Reconnaissance (AR Recon) collects targeted observations during major winter weather events associated with heavy precipitation and flooding on the US West Coast. Ten WC‐130J weather reconnaissance aircraft equipped with Global Navigation Satellite System airborne radio occultation (ARO) technology provide additional data on the broader AR environment, specifically as the AR approaches the coastline. This study evaluates the impact of assimilating ARO bending angle observations using the Model for Prediction Across Scales—Atmosphere—Joint Effort for Data assimilation Integration (MPAS‐JEDI). Experiments use the Local Ensemble Transform Kalman Filter and Three‐Dimensional Ensemble‐Variational algorithms. The control experiment assimilates only global conventional data, while the other experiments incorporate additional ARO observations. The results show that ARO assimilation corrects moisture, temperature, and wind fields, and reduces the error in the integrated vapor transport forecast at landfall. The changes mitigate overestimated precipitation in mountainous regions of Washington and Idaho.