Atmospheric River Event Frequency Is the Principal Moisture Driver for Radial Growth of High‐Elevation Southern California, <scp>USA</scp> Conifers, 1658–2020  <scp>CE</scp>

ABSTRACT We evaluated the influence of atmospheric river (AR) precipitation and event frequency on radial growth patterns of old‐growth Jeffrey pine ( Pinus jeffreyi ) and sugar pine ( Pinus lambertiana ) trees located in the San Jacinto Mountains of Southern California, USA. Annual (September–August) precipitation derived from landfalling AR events between 32°N and 33°N during 1960–2020 accounts for 35.9% of the total rainfall but is more strongly correlated ( p &lt; 0.05) with standardised totalwood radial growth ( r = 0.49) than either total precipitation ( r = 0.47) or non‐AR precipitation ( r = 0.24). However, the strongest climate/growth association ( r = 0.63) was between AR event frequency and radial growth, indicating that AR event frequency accounts for approximately 40% of the interannual variability in radial growth. We reconstructed AR event frequency from 1658 to 2020 ce to evaluate interannual and interdecadal variability during the past 363 years. Regime shift analysis identified seven phases of AR activity, with above‐average AR frequency from 1860 to 1872, 1873 to 1927, and 2008 to 2020, below‐average frequency from 1752 to 1859 and 1928 to 1979, and neutral conditions from 1658 to 1751 and 1980 to 2007. Years with major reconstruction errors (i.e., reconstructed values &gt; 1 σ different from the actual values) during the 61‐year historic record period were significantly associated with low snowfall accumulations and warm temperatures for overpredicted years, and anomalously high snowfall during underpredicted years. These results suggest that: (1) radial growth variability is more strongly associated with AR event frequency than AR precipitation, (2) sensitivity to AR activity is affected by anomalous snowfall and temperature conditions, (3) standardised growth values can be a more accurate proxy than reconstructed values when AR event frequency direction (i.e., the number of years either above or below the mean), and not AR magnitude, is sought, and, (4) the development of historical AR regimes provides context for comprehending AR events that have occurred since the mid‐20th century.