Multiscale Analysis Schemes for the WRF Three-/Four-Dimensional Ensemble Variational Data Assimilation System (WRFDA 3/4DEnVar): Formulations and Case Results

To simultaneously assimilate all available observations at different spatial and temporal scales, multi-scale analysis schemes are implemented for the Weather Research and Forecasting (WRF) model 3/4-dimensional Ensemble-Variational Data Assimilation (WRFDA 3/4DEnVar) system. The key is applying scale-dependent localizations into ensemble-based background error covariance(EnBEC). This can be achieved through splitting the EnBEC into relatively small and large scale components and then applying the scale-dependent localization respectively. This also can be achieved by directly applying multi-scale localizations into the EnBEC. The formulations of the multi-scale analysis schemes is describled and case results from assimilating satellite cloud products along with conventional observations will be presented.