Exploring the Utility of Assimilating Observations from a Mesoscale Network of StickNet Platforms During VORTEX-SE with the High Resolution Rapid Refresh Ensemble

The physical science component of the Verification of the Origins of Rotation in Tornadoes EXperiment – SouthEast (VORTEX-SE) broadly aims to diagnose how environmental factors unique to the southeast U.S. contribute to the formation, intensification, structure, and climatology of tornadoes. Surface heterogeneities resulting from varying land-surface characteristics and terrain elevation changes were targeted as important features to observe and investigate. In order to properly sample the near-surface environment, a mobile network of observing platforms was needed. Texas Tech University contributed 24 StickNet (Weiss and Schroeder 2008) observing platforms during the 2016 and 2017 field phases, with eight probes being used for rapid deployments in advance of developing severe storms. The remaining sixteen platforms were positioned in a mesoscale 4 x 4 array (i.e., StesoNet) centered over Hunstville, AL with approximately 40-km spacing between stations. The platforms operated at 10 Hz, observing u and v wind components, temperature, moisture, and pressure.

This unique and high-resolution dataset, positioned in a relatively data-sparse region near the surface, can be further exploited through data assimilation. Experimental versions of the High Resolution Rapid Refresh Ensemble (HRRRE) run at the Global Systems Division of NOAA ESRL are being utilized to investigate the utility of StickNet observations in surface data assimilation towards improving severe thunderstorm forecasting in the Southeast. Retrospective simulations are begun with hourly-cycled ensemble Kalman filter data assimilation at 0000 UTC on the day of the event, followed by 36-member ensemble forecasts initialized at 1200, 1500, and 1800 UTC. Conventional and radar reflectivity observations are included in the assimilation, along with 1-min averaged bias-corrected StickNet observations of temperature and specific humidity valid at the top of the hour. Ensemble spread is generated through initial-condition and boundary-condition perturbations. Simulations from selected intensive observing periods (IOPs) with and without StickNet observations, characterizing a spectrum of convective modes, forecast successes, and forecast failures, will be presented to examine if the mesoscale network can improve forecasts.