Exploring the Use of Ensembles for Real Time Diagnosis of the Severe Storm Environment

Almost 20 years ago, the Storm Prediction Center (SPC) developed a system to diagnose the mesoscale state of the atmosphere for use in forecasting severe convective weather. This system, known as the SPC Mesoanalysis (https://www.spc.noaa.gov/exper/mesoanalysis/), uses three-dimensional gridded 1-h forecast fields from the Rapid Refresh (RAP) (the Rapid Update Cycle prior to 2012) that are merged with a Barnes objective analysis of current surface observations (termed the SFCOA) to derive severe weather forecast fields. To this day, forecasters rely heavily on this system for guidance on issuing short-term outlooks, mesoscale convective discussions, and convective watches.

Advances in computing power, and extensive improvements in data assimilation/numerical weather prediction systems, allow for real time systems that could significantly improve our ability to diagnose the mesoscale state of the atmosphere. One such system is the subject of this presentation and is based on the experimental High Resolution Rapid Refresh Ensemble (HRRRE). Like in the SPC Mesoanalysis, 1-h forecasts from the HRRRE are used for the background atmospheric state above the surface. However, unlike the SPC Mesoanalysis, current surface observations (METAR, marine, and from select Mesonets) are incorporated into the analysis through an Ensemble Adjustment Kalman Filter (EAKF) encoded in the Data Assimilation Research Testbed system. This technique has the advantage of spreading information in the surface observations to grid points above the surface through the ensemble covariances. In the SPC Mesoanalysis, the SFCOA simply replaces surface fields in the 1-h RAP forecasts. The EAKF technique should result in analyses with more consistent and realistic vertical structures that are then used to derive severe weather forecast fields. Another potential advantage of this system is the use of analysis uncertainty information inherent in the ensemble fields.

This experimental analysis system was run in spring 2018 alongside the current SPC Mesoanalysis system for comparison. Early indications show patterns of severe weather diagnostic fields (CAPE, storm-relative environmental helicity, etc.) often vary significantly compared to the current SPC Mesoanalysis for reasons yet to be identified. Verification of the ensemble analysis system will use independent radiosonde and ACARS observations and will be presented at the conference. Furthermore, differences between the ensemble analysis system and the 3D Real Time Mesoscale Analysis (RTMA) in development within NOAA will be discussed. The 3D-RTMA is being designed to closely fit observations for more local applications and will be a deterministic system in the foreseeable future, whereas the ensemble analysis system is designed to depict the meso-b and larger scale state of the atmosphere (and its uncertainty) akin to the spatial scales that are represented in SPC forecast products