J17.2 Sensitivity Experiments of the Control Parameters of the NCAR WRF 4DREKF Data Assimilation and Forecasting System

Thursday, 27 January 2011: 1:45 PM
2A (Washington State Convention Center)
Yonghui Wu, NCAR, Boulder, CO; and Y. Liu, L. Pan, T. Warner, and J. Pace

A 4DREKF (Four-Dimensional Relaxation Ensemble Kalman Filter, Liu et. al. 2010; this conference) mesoscale data assimilation scheme has been developed at NCAR with the Weather Research and Forecasting Version 3.2 model. 4DREKF combines the Newtonian Relaxation and the LETKF (Local Ensemble Transform Kalman Filter) type Kalman gain to formulate an ensemble-based advanced Four-Dimensional Data Assimilation (FDDA) algorithm. In this study, modeling experiments are conducted to evaluate the impact of key components and parameters of the 4DREKF system, including tempo-spatio interpolation of the Kalman gain for relaxation, the relaxation time windows and rate (nudging coefficients), two algorithms to weigh the Kalman gain for multiple observations, LETKF error-covariance localization and inflation, and ensemble sizes. The modeling result will be presented.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner