Correspondence of Storm Attributes in NWP and Observations: a Look at Rotation Strength and Environmental Characteristics

Updraft helicity (UH) serves as an effective proxy for supercells, and forms the basis for probabilistic forecasts of multiple severe convective hazards. However, establishing a relationship between UH and observed storm rotation has proven difficult, in part due to a lack of a large database characterizing storm rotation. This work seeks to answer the question, ‘what can an ensemble of simulated storm characteristics tell us about observed storms?’ by using the 4 km National Severe Storms Laboratory (NSSL)-WRF ensemble and an extensive hand-analyzed dataset established by the Storm Prediction Center (SPC). Through comparison of these datasets, the relationships between high-resolution modeled and observed storm characteristics can be explored via examination of ensemble-generated values within a certain space-time window around the report occurrence. In particular, the relationships between UH and rotational velocities are characterized, as well as relationships between the model-produced and analyzed environmental parameters such as the Significant Tornado Parameter (STP). Additionally, the relationship between the analyzed STP and rotational velocity and the model-derived STP and UH are compared, to determine if the ensemble is reflecting storm characteristics that would be expected in a given environment (i.e. stronger rotation in an environment with higher STP). Comparisons are drawn by locating the closest model grid point to a given report, and then searching the ensemble within various space and time neighborhoods.

The model/observational relationships are also explored through the lens of the hazards they produce: the tornadic and non-tornadic storms are examined separately to determine whether or not the strength of various model parameters discerns between observed tornadic and non-tornadic storms. UH integrated over different layers is shown to have a statistically different relationship with rotational velocity in tornadic storms than in non-tornadic storms, across multiple space and time windows. Ongoing work at the SPC establishes probabilities of a tornado given a right-moving supercell with a certain value of rotational velocity; if information about the rotational velocity can be taken from the ensemble members’ UH, an unconditional model-generated tornado probability with an observational basis is possible.