Calibration of Synthetic Observations For A Joint Observing System Simulation Experiment

Abstract

A cooperative effort between agencies including the National Centers for Environmental Prediction (NCEP), Earth System Research Laboratory (ESRL), European Centre for Medium-range Weather Forecasting (ECMWF), National Astronautics and Space Administration (NASA), and National Environmental Satellite, Data, and Information Service (NESDIS) is underway to construct a Joint Observing System Simulation Experiment (OSSE). The Joint OSSE uses a 13-month forecast from the ECMWF global operational model at T511 as a Nature Run. A comprehensive set of synthetic observations must be extracted from the Nature Run to simulate observations for every data type currently input into operational forecast models. These synthetic observations are then passed to the Global Forecast System (GFS) model via the Gridpoint Statistical Interpolation (GSI) package in order to perform forecasts for data impact studies. The data impact of synthetic observations must be calibrated against that of real observations to ensure that the results of these data impact studies closely reflect the actual data impact in the real world. The synthetic data impact will be calibrated by comparing to the real data impact results from ESRL-NCEP's real data denial experiments. The calibration using data denial experiments will be compared with the calibration using adjoint technique at NASA/GMAO. The calibration procedure for the synthetic observations is discussed, and results are shown for sets of synthetic data generated respectively by NASA/GMAO, NESDIS and NCEP.