Tracking Atmospheric Rivers Globally: Spatial Distributions and Temporal Evolution of Life Cycle Characteristics

The Tracking Atmospheric Rivers Globally as Elongated Targets (tARget) algorithm is further developed to Version 3, adding the capability to track atmospheric river (AR) life cycles along with other refinement. The results indicate AR genesis is more frequent toward the western boundaries of midlatitude ocean basins and nearby upstream land areas (~1 month−1) compared to the eastern boundaries (~0.5 month−1) and least frequent in tropical and polar areas (reaching toward 0). AR termination is more frequent toward the northeastern sectors of North Pacific/Atlantic and adjacent downstream land areas and in the Southern Ocean near Antarctica (~1 month−1) compared to the adjacent ocean sectors (~0.5 month−1) and least frequent in tropical areas and interior Antarctica where AR genesis is similarly infrequent. ARs tend to be longer‐lived when the genesis (termination) occurs toward the western (eastern) boundaries of midlatitude ocean basins and adjacent land areas (maximum lifetime >72 hr) compared to the opposite side of the ocean basins (24–72 hr) and when terminated at high latitudes. AR travel speed is higher in the midlatitude ocean basins and strongly steered by the zonal wind around 650 hPa. AR tracks are nearly linear in most cases, with the overall travel direction closely correlated with the direction of integrated water vapor transport (r = 0.69) although being more zonal than the latter. Temporally, ARs tend to be longer/stronger around the middle of the life cycle. Seasonal variations in AR life cycle characteristics are also examined. The handling of AR separations/mergers contributes the largest sensitivity in tracking result compared to selection/resolution of input data. Key Points A technique is introduced for spatiotemporally tracking atmospheric rivers (ARs) from genesis to termination around the globe Sensitivity of AR tracking results is examined with respect to various factors, especially, how AR separations/mergers are handled AR life cycles are quantitatively characterized, including their spatial distributions, temporal evolution, and seasonal dependence