Atmospheric rivers (AR) have been noted to play a substantial role in extreme precipitation in several locations globally and studies show that this is more prevalent in a warming climate. Over the northwest and western coasts of North America, these filaments of water vapour transport move inland and cause substantial amounts of precipitation over the region. Due to the effects of landfalling ARs, various researchers have developed AR detection tools (ARDTs), which identify ARs by implementing heuristic, quantitative algorithms for defining ARs. Although results from these ARDTs are broadly similar with respect to strong ARs, there are notable differences associated with moderate and weak ARs. We hypothesise that there are distinct synoptic weather patterns that lead these ARDTs to detect different frequencies and intensities of ARs. In this study, we examine landfalling ARs in the western U.S. during (a) ‘consensus times’ in which AR are detected by all ARDTs (b) ‘non-consensus times’ where all ARDTs are in disagreement on ARs detected. We further compare the composite atmospheric fields (eg: IVT, upper level potential vorticity, mean sea level pressure, etc.) for these consensus and non-consensus times for the various algorithms. If our hypothesis is true, then we expect that the various non-consensus composites will have significantly different patterns from the consensus composites as well as between the ARDTs themselves. We present preliminary results from this analysis.