Study of observational error and targeted observation in observing system simulation experiments (OSSEs) for NPOESS

Observing System Simulation Experiments (OSSEs) at NCEP have been focused on evaluating data impacts on a global scale and NPOESS instruments has been the main focus. It is recognized as ideal tool to achieve the goal THORPEX program, because through OSSE NWP community will be able to make impact on preparation of observation. During the calibration process of OSSEs, a larger impact from wind data as compared to temperature was confirmed. OSSEs to assess the forecast impact of Doppler Wind Lidar (DWL) were also conducted and addressed many valuable recommendations for designing DWL.

Many results seems tp be quite robust, however we also found the data impact depends on handling of surface data and observational error assignment. In this paper various finding about observational error assignments through OSSEs will be presented.

The work will be applied to assess the strategies of effective targeted observation of DWL for NPOESS. It is found that scanning is very important to gain significant impact from DWL and upper level data is more important. However, only for limited time (such as 10% of the time), DWL can be switched on. It is found that without targeting, the impact of DWL is very small.