Ensemble-Based Analysis of Errors in Atmospheric Profiles Retrieved from GNSS Occultation Data

We present results of an empirical error analysis based on simulated GNSS radio occultation data. Occultation observations were simulated for one day adopting the planned European weather satellite METOP as Low Earth Orbit satellite and its GNSS Receiver for Atmospheric Sounding as sensor. An ensemble of 300 occultation events with 100 events in each of three latitude bands (low, middle, high) was chosen and excess phase path profiles were computed involving quite realistic atmospheric modeling and observation system modeling. The rms error of the ionosphere corrected phase path profile sampled at 10 Hz was found to be 2–3 mm, at meso-and stratospheric heights, and the atmospheric Doppler error to be ∼3 mm/sec, mimicking realistically modern receiver performance. Atmospheric profiles were retrieved by applying an advanced geometric optics bending angle retrieval algorithm followed by Abelian refractivity retrieval and, subsequently, by dry air retrieval in the stratosphere and optimal estimation retrieval in the troposphere. The retrieved profiles were referenced to the “true” co-located ones of the ECMWF analysis field used as atmospheric model. Based on these data, we empirically estimated bias profiles and covariance matrices (standard deviations and correlation functions) for the retrieval products bending angle, refractivity, pressure, geopotential height, temperature and specific humidity. Specific results include: Refractivity exhibits a relative standard deviation of 0.1–0.75% and a relative bias of <0.1% at 5–40 km height. Temperature shows a standard deviation of 0.2–1 K at 3–31 km height and a bias of <0.5 K below 33 km and of <0.1 K below 20 km. The obtained empirical errors are conservative error estimates and provide a valuable basis for further retrieval algorithm improvements and for proper specification of observational errors in data assimilation systems.