Detection of Atmospheric Rivers: An Algorithm for Global Climatology and Model Evaluation Studies

Abstract:

Atmospheric rivers (ARs) are narrow, elongated, synoptic jets of water vapor that play important roles in the global water cycle and regional weather and hydrology. Previous studies have developed techniques for the identification of ARs based on intensity and/or geometry thresholds indicative of AR conditions. Such techniques have facilitated the investigation of ARs on local to regional scales. Recent advancement in the understanding of AR’s global signatures and impacts (including those in less explored areas such as Greenland and Antarctica), and the need for understanding the representation of key AR characteristics in global weather/climate models motivate the development and evaluation of AR detection techniques suitable for global climatological and model evaluation studies. In this work, an objective AR detection algorithm is developed based on thresholding global, 6-hourly fields of integrated water vapor transport (IVT) derived from ERA-Interim reanalysis. Long, narrow filaments of enhanced IVT are detected by applying a set of intensity and geometry criteria, along with other considerations. Key output of the algorithm includes the AR shape boundary, main axis, location of landfalls, and a tabulated list of the basic statistics such as length, width, and mean IVT strength/direction of each detected AR. Sensitivity of detection is examined for selected parameters, and the result is evaluated and compared with an independent database of landfalling ARs in the west coast of North America based on satellite images of integrated water vapor (Neiman et al. 2008). Global distribution of key AR characteristics, and examples of their modulation by climate variability, will be presented.