Occurrence Frequency of Global Atmospheric River (AR) Events: A Data Fusion Analysis of 12 Identification Data Sets

The atmospheric river (AR) is a long, narrow, and transient corridor of strong horizontal water vapor transport. Various AR detection methods have been proposed, which have introduced significant uncertainty to the identified AR characteristics. This study has designed a data fusion algorithm to merge 12 data sets of different global and regional AR identification algorithms published by the Atmospheric River Tracking Method Intercomparison Project (ARTMIP) covering the period from 1980 to 2016. It aims to conduct frequency statistics to further research the global distribution, interannual variation, the trends in poleward shifting, and the impacted factors of AR occurrence. The quantitative results indicate an overall increasing trend in interannual variation, with a more pronounced growth trend observed in the oceanic region between 40 and 60oS. Additionally, the study identifies a poleward shift in the peak latitude of AR occurrence frequency, with speeds of 0.589 degrees and 0.769 degrees per decade in the Northern and Southern Hemispheres, respectively. This shift may be associated with the tropical poleward expansion. Upon examining the relationship between AR frequency and sea surface temperature (SST) as well as zonal wind, the study finds that distinct dominant factors influence AR in different regions. AR events near the 30 degrees N/S ocean are influenced more significantly by zonal wind than by SST. These findings shed light on the global characteristics of AR occurrences and provide insights into the factors governing their variability across different areas. Researchers employ an array of tools to detect a weather phenomenon event known as the atmospheric river (AR), which is a long, narrow corridor in the atmosphere that transports water vapor originating from the tropics. Each tool employs a distinct methodology for identifying these atmospheric rivers, which can lead to varying results. This diversity makes it challenging for us to select a single methodology to study AR on a global scale and over long periods. To address this problem, we have developed an innovative algorithm. This algorithm merges the results from different tools to create a more consistent and comprehensive data set of atmospheric rivers. Utilizing this novel data set, we investigated how atmospheric rivers have evolved over time and across different locations. Our research indicates a poleward shift of AR positions throughout the research period. Moreover, we also examined the sea surface temperature and wind patterns, and their connections to the frequency of atmospheric rivers. These insights could provide insightful implications for how climate changes might influence these important weather phenomena in the future. Proposed a data fusion algorithm merging 12 Atmospheric River Detection Tools data sets based on their identification results conflicts The poleward shifts of the AR frequency peak and tropical expansion have a moderate correlation during the research period AR events near the 30 degrees N/S ocean are influenced more significantly by zonal wind than by sea surface temperature