Ensemble Sensitivity Analysis of Multiple Great Plains Convective Events

Through the use of an ensemble of numerical weather forecasts, it is possible to identify atmospheric features at analysis or earlier forecast times to which a specific forecast aspect (response function) is sensitive. The examination of such sensitivity, known as ensemble sensitivity analysis (ESA), is quantified by dividing the covariance between the response function and some initial state field by the initial or earlier forecast state field variance (which is statistically equivalent to calculating the slope of the least-squares fit line between the variables). With the advent of modern computing capabilities, it is now possible to run a sufficiently large ensemble of forecasts at convection-allowing grid spacings and subsequently calculate sensitivities of many response functions to many state variables. This allows for the identification of specific flow features (e.g., a shortwave trough in 500 hPa heights) that are most influential on the occurrence and magnitude of certain characteristics of deep, moist convection (e.g. max updraft helicity within some specified area). This work utilizes a 50 member Ensemble Kalman Filter (EnKF) configuration using the WRF-ARW model with 4km grid spacing across portions of the U.S. Great Plains to assess the ensemble sensitivity for a number of independent convective events. The Data Assimilation Research Testbed Ensemble Adjustment Filter (DART-EAKF) is used to assimilate mesocale-resolving observations into the model prior state. A main goal of this work is to use ESA to understand the variability of the flow features that most strongly control the characteristics of the identified severe convective events (coverage, severity, storm mode).