Yasutaka Ikuta, Japan Meteorological Agency
Data Assimilation with Adjoint Model including Three-Ice Bulk Cloud Microphysics
Abstract:
Japan Meteorological Agency has been developing a new meso-scale data assimilation (DA) system for further improvement of very short-range precipitation forecasts. The assimilation method in this system is a hybrid 4DVar using flow-dependent background error estimated from ensemble forecasts. Especially, this system aims at the assimilation of high temporal resolution observations whose importance is emphasized within DA research/development field in recent years. Some examples of such observations are radar reflectivity and satellite radiance, for whose assimilation the detailed information of hydrometeors is necessary. For this purpose, we developed tangent-linear/adjoint model including 6-class 3-ice 1-moment bulk cloud microphysics scheme. This tangent linear model of cloud microphysics can keep consistency of nonlinear trajectory and stability of linear calculation during 3-hour assimilation window by simplified microphysical processes. The backward trajectory for each hydrometeor from high temporal resolution reflectivity data including ice has been checked in single column model. The impact of the assimilation will be presented at the symposium.