Development of Single Column Local Ensemble Transform Kalman Filter with JMA-NHM-based 4DVAR data assimilation system using TRMM PR

The Japan Meteorological Agency (JMA) has been operating a meso-scale model (MSM), and the initial condition of the MSM is given by a JMA-NHM-based 4DVAR data assimilation system, called JNoVA. The accuracy of hydrometeors profiles in the initial condition provided by the JNoVA is vitally important for the MSM. Accordingly, we have been developing assimilation techniques for the improvement of hydrometeors using reflectivity observation of the Tropical Rainfall Measuring Mission (TRMM) precipitation radar (PR). However, the operational JNoVA does not include the hydrometeors control variable and the Tangent Linear and Adjoint (TL/AD) model of the cloud microphysics, because the minimization of a cost function is interfered by non-linearity of that. Thus we have been developing a Single Columns Local Ensemble Transform Kalman Filter (SC-LETKF) to support the reflectivity assimilation of the JNoVA. The purpose of the SC-LETKF is to update the atmospheric profile around the observation, and the optimized atmospheric columns are inputted to outer-loop in the JNoVA by incremental analysis updates approach. Moreover, since the ensemble perturbations of the SC-LETKF are provided by the set of single-column profiles in the first guess, the SC-LETKF does not need the ensemble forecasts. In this presentation we will show that this technique for assimilation of vertical distribution of reflectivity from the TRMM PR provides the improvement of the meso-scale forecast.