Does Convection in Atmospheric Rivers at Genesis Impact Precipitation Production Later in Their Lifecycle?

July 16, 2026

The manuscript entitled “Does convection in atmospheric rivers at genesis impact precipitation production later in their lifecycle?” was recently published in the journal Climate Dynamics. The work is the fruit of a collaboration between the Center for Western Weather and Water Extremes (CW3E) at the Scripps Institution of Oceanography, Columbia University/National Aeronautics and Space Administration Goddard Institute of Space Studies (NASA-GISS), the University of California in Los Angeles (UCLA), and the California Institute of Technology- Jet Propulsion Laboratory (JPL). The paper is led by Catherine Naud (Columbia Univ./NASA-GISS), co-authored by CW3E scientists Rosa Luna Niño and Alexander Gershunov, and JPL colleagues Derek Posselt and Juan Crespo (UCLA-JPL). The work was supported by a NASA Precipitation Measurement Mission grant. The research explores the potential impact of convective activity early in the development phase of Atmospheric Rivers (ARs) on the production of precipitation throughout the ARs lifecycle.

As such, this work belongs to CW3E’s 2025-2029 Strategic Plan priority “Atmospheric Rivers and Extreme Precipitation Research, Prediction, and Applications”. ARs are narrow filaments of water vapor transport originating in the remote oceans that, when making landfall, produce vast amounts of precipitation. While they are regarded as beneficial in semi-arid coastal regions, they can also produce hazardous conditions, including flash flooding and strong winds. Accurately forecasting their arrival, location, and impacts is complicated by their formation occurring in remote regions where few observations are available for assimilation in forecast models. As a result, the accuracy of the forecast relies on the model skills at representing all the mechanisms that contribute to the ARs characteristics. Therefore, it is paramount to use existing observations to examine the processes that significantly contribute to rain production within ARs throughout their lifecycle. One such process, convection, is known to occur within ARs, but what is less clear is how its presence affects the AR rain production at the time it occurs and later in the AR lifecycle.

Using 10 years of NASA Global Precipitation Measurement (GPM) mission estimates of precipitation rates and types, combined with an AR database, the occurrence of convection during early AR development is explored. A compositing approach, centered on the point of maximum Integrated Vapor Transport (IVT) along the ARs’ main axis, is introduced to average the observed characteristics of the ARs. With this maximum IVT centered perspective, IVT is strongest to the west, but precipitation rates are largest to the east and poleward of the composite center (e.g. Figure 1i vs Figure 1j). This is because mean precipitation rates and frequency align with the region of maximum ascent. Precipitation intensity, however, tends to be largest west of the maximum IVT point, and so is convective fraction. ARs tend to have larger precipitation intensity and convective fraction during the developing phase, with mean precipitation rates (including non-precipitating times) and IVT peaking later in life (Figure 1). Selecting ARs based on the convective fraction early in life, ARs with relatively higher amounts of convection tend to have larger IVTs and precipitation rates throughout the AR full lifecycle than ARs with lower rates of convection occurrence (Figure 2). This is consistent with prior work that demonstrated the impact of latent heat release on AR moisture transport. The observations confirm that high convection activity is associated with high IVT ARs and provide a constraint and evaluation tool for numerical experiments.

Figure 1. Max(IVT) AR centered composites through AR age groups of, in color, (a,e,I,m,q) median IVT, (b, f, j,n,r) mean IMERG precipitation rate, (c,g,k,o,s) median GPM precipitation intensity, and (d,h,l,p,t) convective fraction for age 0.0-0.2 (a-d), 0.2-0.4 (e-h), 0.4-0.6 (i-l), 0.6-0.8 (m-p) and 0.8-1.0 (q-t). In each panel the dashed contours represent the footprint fraction of that age group from 5 to 85% in 20% increments. White shading indicates areas with no data points. Figure 5 from Naud et al. 2026.

Figure 2. Max(IVT) AR centered composite differences between high and low convective fraction at genesis ARs for (a-c) ARs of age 0.0-0.2, and (d-f) ARs of age > 0.2, in color, (a,d) median IVT, (b, e) mean IMERG precipitation rate, (c,f) GPM convective fraction. In each panel white shading indicates areas where the difference is less than the sampling uncertainty measure (see section 2.6) or the footprint fraction is less than 5%. Figure 8 from Naud et al. 2026.

Citation:

Naud, C. M., Luna-Niño, R., Posselt, D. J., Crespo, J. A., & Gershunov, A. (2026). Does convection in atmospheric rivers at genesis impact precipitation production later in their lifecycle? Climate Dynamics, 64(6). https://doi.org/10.1007/s00382-026-08241-w