Skip to main content
U.S. flag

An official website of the United States government

Publication Date
3 December 2020

Quantifying Uncertainty in the Detection of Atmospheric Rivers

Summary of the 3rd Atmospheric River Tracking Method Intercomparison Project Workshop.
Print / PDF
Powerpoint Slide

The Berkeley Lab CASCADE SFA has created a high-performance-computing tool for quantifying uncertainty in the detection of atmospheric rivers (ARs): the Toolkit for Extreme Climate Analysis Bayesian AR Detector – TECA-BARD v1.0.1.


This is the first time that a statistical machine learning technique has been used to train a method (TECA-BARD v1.0.1) to simultaneously (a) emulate the behavior of experts when identifying ARs, and (b) quantify the uncertainty associated with the expert’s subjective opinions. This allows research on AR research that explicitly quantifies a major source of uncertainty. It builds on a high-performance computing (HPC) code (TECA) that allows AR analyses to be done quickly and efficiently on HPC machines like those at the National Energy Research Supercomputing Center.


The Berkeley Lab CASCADE SFA developed a novel atmospheric river (AR) detection, with the goal of ‘training’ it to emulate how an atmospheric science expert would detect ARs.  To accomplish this, eight members of the team counted ARs in meteorological maps.  This dataset of expert AR counts was used within a statistical machine learning (Bayesian) framework to determine optimal parameter settings for the novel AR detection algorithm.  This machine learning process resulted in a set of 128 separate AR detectors designed to emulate each of the 8 experts, resulting in a total of 1,024 separate AR detectors.  These 1,024 detectors were incorporated in the Toolkit for Extreme Climate Analysis (TECA) as the TECA Bayesian AR Detector (TECA-BARD v1.0.1). The team used TECA-BARD v1.0.1 to investigate the question “How does El Niño affect the number of ARs?”; TECA-BARD v1.0.1 provided 1,024 different answers.  Differences indicate that the answer to the question (whether there are more or fewer ARs during El Niño) depends on which expert the AR detector was trained on, which led the authors of the paper to call for more research to constrain our definition of ARs.

TECA-BARD v1.0.1 is publicly available to the scientific research community as part of the TECA software suite.

Point of Contact
William D. Collins
Lawrence Berkeley National Laboratory
Funding Program Area(s)