Validation of operational GRAS radio occultation data

Abstract

GRAS RO (radio occultations) are validated against co‐located ECMWF and COSMIC data and by ECMWF impact trials. We focus on closed‐loop data at impact heights above 8 km. Results confirm the high GRAS quality and robustness, showing lower noise than COSMIC and more occultations per day/satellite. Mean differences to ECMWF and COSMIC from 18 km to 35 km show about 0.1% smaller GRAS BAs (bending angles). Around 40 km, ECMWF shows on average about 1% smaller BAs, which may be related to microwave radiances assimilation. Recent ECMWF updates, putting more weight on RO here, reduce this bias. COSMIC co‐locations reveal smaller GRAS BAs up to about 50 km, probably partly caused by COSMIC smoothing; this is currently revised. ECMWF forecast trials show similar positive GRAS, COSMIC impacts for Southern latitudes standard deviations, although GRAS provides about 60% fewer occultations. It also demonstrates that more RO instruments are beneficial, particularly for the tropics and Northern latitudes.