The Dependence of Background Error Statistics Characteristics in Radar Data Assimilation for Prediction of a Heavy Rainfall Event.

Abstract:

The Indian subcontinent experiences an increased frequency and magnitude of extreme rainfall events. Knowledge of future extreme precipitation in near real time is crucial for implementing adaptation and mitigation measures. The state of art solutions for precipitation forecasts are solely based on the numerical weather prediction (NWP) models. Yet, forecasting of extreme convective weather events using NWP models remain elusive. Improper representation of storm structure in the initial condition remains one of the major reasons for reduced forecast skill of convective events. Assimilating Doppler weather radar (DWR) observations using variational system offers potential for improving initial state of the NWP model. However, there is a caveat to the application of this technique, as variational assimilation systems are sensitive to the error statistics used for weighting the short term forecasts (background). Moreover, the effect of background error statistics (BES) structure on radar data assimilation remains unclear. Based on a heavy convective event, we have investigated the role of momentum control variables in improving the analysis skill of both four and three dimensional variational assimilation system (3DVAR) [1,2]. Our results suggest that proper choice of control variables significantly increases wind and moisture convergence, thereby improving the precipitation forecast skill. The study also highlights the need for an ensemble based stationary BES in the 3DVAR based radar assimilation system to improve the convective precipitation forecast [3].

Reference:

1. Thiruvengadam P, Indu J, Ghosh S. 2019 Assimilation of Doppler Weather Radar data with a regional WRF-3DVAR system: Impact of control variables on forecasts of a heavy rainfall case. Advances in Water Resources
2. Thiruvengadam P, Indu J, Ghosh S. 2020 Significance of 4DVAR Radar data assimilation in Weather Research and Forecast model based Nowcasting system. Journal of Geophysical Research Atmospheres
3. Thiruvengadam P, Indu J, Ghosh S. 2020 Improving Convective Precipitation Forecasts using Ensemble-Based Background Error Covariance in 3DVAR Radar Assimilation System. Earth and Space Science