[8-3]

Vertical localization strategies for radiance data assimilation within LETKF

J.-S. Kang (KIAPS), H.-W. Chun (KIAPS), B.-J. Jung (KIAPS, NCAR), J. Kim (KIAPS), and T. Miyoshi (RIKEN)

Radiance data from remote sensing instruments provide weight-averaged information in the vertical column of atmosphere. In order to assimilation such column data, LETKF algorithm needs to decide how much the data should give an increment for each vertical model level. There are several previous studies to suggest how to localize radiance data within EnKF data assimilation based on a vertical weighting (sensitivity) function of radiance data. We have examined the various methods to localize the radiance data of AMSU-A and IASI within KIAPS-LETKF data assimilation system implemented to NCAR CAM-SE model. KIAPS-LETKF system has been performed in a coupled system with KPOP (KIAPS Package for Observation Processing) that includes processes of bias correction, quality control, variable transform (RTTOV for radiance, ROPP for GPS-RO). Methodologies of vertical localization for radiance data within KIAPS-LETKF system have been investigated under the Observing System Simulation Experiments (OSSEs) assuming that radiance data have no systematic bias after quality control processes. As a result, IASI data shows much stronger sensitivity to the vertical localization than AMSU-A data, because it has much more channels with various shapes. We have also introduced an advanced method to use low-level channels of IASI data better.