Impact of In-Situ Observations from Three TORUS Cases on Model Analyses and Forecasts of Severe Convection
Presenter Abstract
The quality of forecasts of severe deep convection in high-resolution numerical weather prediction is dependent on the representativeness of the model initial conditions. This representativeness may be negatively impacted by limited observation availability and by the presence of mesoscale heterogeneities not detected by conventional observations assimilated into convection-allowing models. By assimilating storm-scale observations from the Targeted Observation by Radars and Uncrewed Aircraft Systems (UAS) of Supercells (TORUS) campaign into an ensemble styled after the Warn-on-Forecast System, this study aims to investigate if data from field work platforms can improve the quality of initial conditions in the ensemble, potentially leading to improved forecast quality. Using ensemble Kalman filter assimilation procedures from the Data Assimilation Research Testbed, this work extends a previous study conducted to assimilate data from the 8 June 2019 TORUS IOP by considering three additional cases from TORUS. Data from mobile mesonets, UAS, and radiosondes are assimilated from two cases during TORUS-2019 (17 May 2019 and 28 May 2019) and one case during TORUS-LItE (Left-flank Intensive Experiment; 26 May 2023). Limited conventional observation availability and the presence of mesoscale heterogeneities were the two main factors that guided the selection of the three cases from TORUS. Data-denial experiments are performed for each of the three cases across three layers of the atmosphere (surface, planetary boundary layer, and the free atmosphere). In addition, a UAS-focused data-denial experiment was conducted to evaluate the specific impact of that platform. Results from all three cases show examples of improved probabilities of convection around target supercells in ensembles assimilating TORUS observations when compared to control ensembles (which included conventional observations only). Similar to previous work assimilating data on 8 June 2019, one platform or one layer did not appear to be most important across the three cases, with forecast improvements from UAS and radiosonde data apparent on 17 May and 28 May and forecast improvements from mobile mesonet data apparent on 26 May.
Presentations
Presented in Session 9: Field Campaigns
Impact of In-Situ Observations from Three TORUS Cases on Model Analyses and Forecasts of Severe Convection
The quality of forecasts of severe deep convection in high-resolution numerical weather prediction is dependent on the representativeness of the model initial conditions. This representativeness may be negatively impacted by limited observation availability and by the presence of mesoscale heterogeneities not detected by conventional observations assimilated into convection-allowing models. By assimilating storm-scale observations from the Targeted Observation by Radars and Uncrewed Aircraft Systems (UAS) of Supercells (TORUS) campaign into an ensemble styled after the Warn-on-Forecast System, this study aims to investigate if data from field work platforms can improve the quality of initial conditions in the ensemble, potentially leading to improved forecast quality. Using ensemble Kalman filter assimilation procedures from the Data Assimilation Research Testbed, this work extends a previous study conducted to assimilate data from the 8 June 2019 TORUS IOP by considering three additional cases from TORUS. Data from mobile mesonets, UAS, and radiosondes are assimilated from two cases during TORUS-2019 (17 May 2019 and 28 May 2019) and one case during TORUS-LItE (Left-flank Intensive Experiment; 26 May 2023). Limited conventional observation availability and the presence of mesoscale heterogeneities were the two main factors that guided the selection of the three cases from TORUS. Data-denial experiments are performed for each of the three cases across three layers of the atmosphere (surface, planetary boundary layer, and the free atmosphere). In addition, a UAS-focused data-denial experiment was conducted to evaluate the specific impact of that platform. Results from all three cases show examples of improved probabilities of convection around target supercells in ensembles assimilating TORUS observations when compared to control ensembles (which included conventional observations only). Similar to previous work assimilating data on 8 June 2019, one platform or one layer did not appear to be most important across the three cases, with forecast improvements from UAS and radiosonde data apparent on 17 May and 28 May and forecast improvements from mobile mesonet data apparent on 26 May.