May 22 - July 30
Verification of Proxy Storm Reports Derived From Ensemble Updraft Helicity
Mallory Row, James Correia, Jr., and Patrick Marsh
What is already known:
- Updraft helicity, a measure of rotation in modeled storms, is a new variable in severe weather forecasting that is produced from convection-allowing models.
- As an extension of Sobash et al. 2011, ensemble data from the Storm Scale Ensemble of Opportunity is used to create proxy storm reports from updraft helicity track’s maxima.
- Clark et al. 2013 found a strong correlation between modeled updraft helicity tracks and observed tornado tracks.
What this study adds:
- The ensemble outperforms any individual members and shows decent skill throughout the year, especially in the springtime.
- Amongst the members, there are two separate groups: one with higher Percent of Detection (POD) but lower Frequency of Hits (FOH), and the other with lower POD but higher FOH. Taking these two groups together in an ensemble mean, there is a compensation effect happening that allows for a more skillful ensemble mean.
- Case studies of an outbreak day and lower end severe weather day show this compensation effect is happening on a smaller time scale.
Abstract:
Convection-allowing models (CAMs) are one of the newest improvements the area of numerical weather prediction (NWP) has seen in the last 10 years. One of the new diagnostic fields these models output is updraft helicity (UH), a measure of rotation in modeled storms. Data collected from Storm Scale Ensemble of Opportunity (SSEO) and its individual members in 2012 is used to create proxy storm reports derived from UH track-like objects. Daily probabilistic forecasts are created from the reports allowing for a direct comparison to the observed for that day. 2x2 contingency tables are constructed daily to gain insight to if UH provides a skillful and reliable probabilistic serve weathers forecast and understand the characteristics of the SSEO and members. Various verification metrics are calculated along with looking at correlation data and probabilistic outlooks to provide a fuller understanding. The SSEO is found to have good skill and reliability throughout the year with especially good skill in the spring time (March to June).