17:15 〜 19:15
[MGI26-P03] Reduced non-Gaussianity in multi-scale background error by assimilating every-30-second radar observation: a case of idealized deep convection
キーワード：データ同化、気象レーダ
The non-Gaussianity of error probability distributions is a major challenge in EnKF-based methods when we assimilate radar reflectivity data for rapidly growing convective systems. The assimilation of phased array weather radar data with a very short interval of 30 seconds is an interesting approach to overcome this problem. The previous studies showed promising results in real-world cases, which had limitations in verifying the analysis accuracy. It was also difficult to distinguish the effect of non-Gaussianity from other factors which may also degrade the analysis and forecast accuracies, such as the errors in model physics, imperfect and nonlinear observation operators, limited observation coverage, and multi-scale background error.
In this study, we perform a series of idealized OSSEs for a convective cell triggered by a warm bubble and investigate the impact of assimilating radar observation with high frequency, focusing on the non-Gaussianity and the analysis accuracy. We used 100-member LETKF and synthetic radar reflectivity observation generated every 30 seconds. We generated the initial ensemble perturbations by shifting the location of the warm bubble and adding random band-pass filtered temperature perturbation. We compared the analysis fields after 50 minutes of data assimilation cycle of three different cases: 3D-LETKF with a 5-minute interval (using only 1/10 of full data), 4D-LETKF with a 5-minute interval, and 3D-LETKF with a 30-second interval.
We found that assimilating radar reflectivity every 30 seconds leads to a significant reduction of the non-Gaussianity of the background ensemble and the improvement of the analysis field, particularly for vertical velocity around the convective core (Panels 1 (g-i) in the figure). We also found the improvement in the analysis mean value of vertical velocity (Panels 1 (a-c)). However, the precipitation forecasts did not show significant differences in this idealized setting which ignores both model error and background error in spatial scales larger than mesoscale.
We performed another set of experiments adding perturbations on background thermal and wind vertical profiles of the initial condition, to imitate a more realistic situation with multi-scale uncertainty in the first guess. With the perturbed background profiles, we found a more significant impact of 30-second update on non-Gaussianity (Panels 2 (g-i)) and analysis vertical velocity fields (Panels 2 (a-c)), though the impact on precipitation forecasts was still not very significant. This may help interpret previous studies using real-world data.
In this study, we perform a series of idealized OSSEs for a convective cell triggered by a warm bubble and investigate the impact of assimilating radar observation with high frequency, focusing on the non-Gaussianity and the analysis accuracy. We used 100-member LETKF and synthetic radar reflectivity observation generated every 30 seconds. We generated the initial ensemble perturbations by shifting the location of the warm bubble and adding random band-pass filtered temperature perturbation. We compared the analysis fields after 50 minutes of data assimilation cycle of three different cases: 3D-LETKF with a 5-minute interval (using only 1/10 of full data), 4D-LETKF with a 5-minute interval, and 3D-LETKF with a 30-second interval.
We found that assimilating radar reflectivity every 30 seconds leads to a significant reduction of the non-Gaussianity of the background ensemble and the improvement of the analysis field, particularly for vertical velocity around the convective core (Panels 1 (g-i) in the figure). We also found the improvement in the analysis mean value of vertical velocity (Panels 1 (a-c)). However, the precipitation forecasts did not show significant differences in this idealized setting which ignores both model error and background error in spatial scales larger than mesoscale.
We performed another set of experiments adding perturbations on background thermal and wind vertical profiles of the initial condition, to imitate a more realistic situation with multi-scale uncertainty in the first guess. With the perturbed background profiles, we found a more significant impact of 30-second update on non-Gaussianity (Panels 2 (g-i)) and analysis vertical velocity fields (Panels 2 (a-c)), though the impact on precipitation forecasts was still not very significant. This may help interpret previous studies using real-world data.