Consensus and Disagreement in Atmospheric River (AR) Detection: ARTMIP Global Catalogues

Most atmospheric river detection tools (DTs) agree when detecting frequency of occurrence for strong and moderate ARs.

The Science

From a climatological perspective, the average frequency of atmospheric river (AR) occurrence among the catalogues and its associated standard deviation indicate significant disagreement that warrants caution. However, the frequency of catalogue majority-consensus AR conditions suggests a high degree of agreement in extratropical areas. The ensemble of methods shows that AR conditions robustly occur along five tracks over the extratropical oceans, with some extension into extratropical continents. Outside these regions, AR detections are highly method dependent and should be treated with care. Two case studies of AR conditions over the North Pacific and southern Indian Ocean in November 2006 illustrate when and where methods tend to agree or disagree. The main sources of discrepancy come from marginal features—that is, regions of modest integrated water vapor transport (IVT) that may or may not be filamentary in structure and which tend to be associated with weak, decaying areas, or those areas where the IVT gradients are shallow. Strong ARs, however, are robustly detected by the algorithms, and there is even general consensus on AR shape and inland penetration when IVT gradients are sufficiently sharp.

The Impact

With a wide range of AR detection tools (DTs) represented in the literature, quantifying where methods agree and disagree, globally, is valuable for interpreting AR metrics and their context. With ARTMIP, the research team find that most methods tend to agree quantitatively for frequency of occurrence when detecting moderate and stronger features. Disagreement typically occurs for regions with slightly enhanced vapor transport, or weak ARs.


Atmospheric rivers constitute an important mechanism for IVT, but research on their characteristics and impacts has relied on a diverse assortment of detection methodologies, complicating comparisons. The AR Tracking Method Intercomparison Project (ARTMIP) provides a platform for comparing such methodologies, but analysis of ARTMIP catalogues has heretofore focused primarily on specific regions. This study investigates ARs as detected by an ensemble of algorithms with global coverage. The researchers find that the frequency of occurrence of the majority-consensus ARs produces a robust distribution, featuring five hot spots over the extratropical oceans, against which they compare individual algorithm results. They further explore the underlying similarities and differences via two case studies of AR evolution. The dominant source of disagreement between detection methodologies globally consists of detections (or lack thereof) of weak features, and the algorithms otherwise tend to agree remarkably well on the footprints of ARs.

Principal Investigator(s)

Christine A. Shields
National Center for Atmospheric Research

Gerald A. Meehl
National Center for Atmospheric Research


U.S. Department of Energy (DOE) support was provided by the Regional and Global Model Analysis (RGMA) component of the Earth and Environmental System Modeling Program of the Earth and Environmental Systems Sciences Division of the Office of Biological & Environmental Research (BER), within the DOE Office of Science, via National Science Foundation Interagency Agreement No. 1844590.


Lora, J.M., Shields, C.A., Rutz, J.J. “Consensus and disagreement in atmospheric river detection: ARTMIP global catalogues.” Geophysical Research Letters 47(20), e2020GL089302 (2020). [DOI:10.1029/2020GL089302]