Simulation of observation and Calibration for Joint OSSEs at NOAA

An internationally collaborative effort for full Observing System Simulation Experiments (OSSEs), called Joint OSSE, was formed three years ago. The first Joint OSSE Nature Run has been produced by the European Center for Medium-Range Weather Forecasts (ECMWF). This Nature Run covers a 13 month period at T511 horizontal resolution with 91 vertical levels.

Full OSSEs allow the quantitative assessment of the analysis and forecast impact if a calibration is performed. Although a large initial investment is required for a full OSSE, using a full OSSE is the most reliable strategy today to assess the impact from prospective observing systems.

Ideally, all new instruments should be tested by OSSEs before they are selected for construction and deployment. OSSEs will be important in influencing the design of the instruments and the configuration of the global observing system. While the instruments are being built, OSSEs will also help prepare the operational Data Assimilation Systems (DAS) for the new instruments.

In a Full OSSE, all major observations used for the DAS have to be simulated as a control observation in addition to the observations to be tested by OSSEs. Using the control observations, the calibration of an OSSE verifies the simulated data impact by comparing it to a real data impact. Simulation of control observations and OSSE calibration are the most significant initial investments for a full OSSE before it can be used to evaluate data impact of future instruments, and this most fundamental work is not easy to be funded. NCEP and NESDIS have volunteered to simulate control observations. Calibrations of OSSE systems using the NCEP global data assimilation system are being conducted at ESRL and NCEP.

Flexible Strategies which allow coordination of limited resources over a wide area of expertise are proposed. Radiance data has been simulated using the Community Radiative Transfer Model (CRTM). The simulation of observational error will be discussed. Conventional data are simulated using the observed distribution. Tropical and midlatitude cyclone generation (TC-vital) has also been simulated using NCEP software. Progress has been posted to experts to share experience and to receive advices.

Simulated data and a data base for alternative simulations are being prepared and being made available through the NASA/GSFC/NCCS portal, where the Nature Run is also available. The simulation software is also provided through the Joint OSSE web site.

The calibration and simulation effort at NOAA is coordinated with work being done at NASA/GSFC/GMAO.