Preliminary evaluation of all-sky radiance assimilation scheme with POD-3DEnVar method

Abstract

To meet the challenges in satellite data assimilation involving cloud and precipitation processes, an all-sky radiance assimilation scheme (referred to as the ‘scheme’) that includes hydrometeor analysis is constructed based on the Proper Orthogonal Decomposition (POD)-3DEnVar method. Three key elements of the scheme are: (1) four hydrometeor mixing ratios for cloud-water, cloud-ice, rain, and snow in the control variables of the POD-3DEnVar method; (2) an extension of the Community Radiative Transfer Model interface to include hydrometeor profiles; and (3) multiple physical process model forecasts that use multiple microphysical and cumulus parameterization schemes as ensemble samples to estimate the flow-dependent background error covariance of the control variables, including those for hydrometeors.To evaluate the performance of the scheme, single-observation experiments and observation system simulation experiments (OSSEs) were designed for a binary typhoon case. Results for the single observation experiments show that the scheme can accurately assimilate satellite radiance near typhoon centers and produce reasonable increments for the hydrometeor variables. The background error covariances of the hydrometeor variables have reasonable flow-dependent characteristics. Results for the OSSEs have root-mean-square errors for all control variables of the assimilation analysis field with respect to the ‘true’ field that are lower than those obtained without assimilation. This indicates that the scheme effectively reduces error for all control variables, including those for hydrometeors. Forecasting experiments reveal that these improvements in the assimilation of control variables result in improved forecasts. However, improvements in the hydrometeor variables are of shorter duration than are those of the other control variables. The impact of improvements in the assimilation on typhoon track and intensity are obvious in the forecasting field.