Updating HAILCAST Hail Size Predictions in NSHARP

HAILCAST is a 1-D coupled cloud and hail growth model that uses a representative environmental profile to provide an ensemble prediction of the maximum size of hail reaching the surface, assuming storms develop. The version of HAILCAST described in Jewell and Brimelow (2009, hereafter "JB09 HAILCAST") has been widely used in analysis of observed and forecast soundings for the past decade as part of the National Center Sounding and Hodograph Analysis and Research Program (NSHARP) at the Storm Prediction Center (SPC). Adams-Selin and Ziegler (2016) and Adams-Selin et al. (2019) modified several aspects of the hail growth model (e.g. density in different growth regimes, liquid water shedding, parameterization of hail advection across the updraft) and adapted HAILCAST to run on gridded output from convection-allowing models rather than on individual soundings.

In light of favorable test results, the most recent version of HAILCAST (hereafter "AS19 HAILCAST") is being converted to use on single soundings in NSHARP as well. The utility of the updates to the hail growth model is evaluated by running both JB09 HAILCAST and AS19 HAILCAST on a large set of objective-analysis-derived soundings associated with deep moist convection. Because of the uncertainties inherent in hail size verification, the ability of both models to predict instances of "no hail," "non-severe hail," "severe hail," and "significant severe hail" are tested against (1) archived hail reports from the SPC severe weather database and (2) gridded Maximum Estimated Size of Hail (MESH) products for the period from 2011 to 2017.