Abstract

In the tropical lower troposphere, the radio occultation (RO) bending angle can have complex and sharp structures. In current numerical weather prediction data assimilation, the estimated observation uncertainty (or error) of the RO bending angle is typically derived from a fixed-fraction uncerta... Show moreIn the tropical lower troposphere, the radio occultation (RO) bending angle can have complex and sharp structures. In current numerical weather prediction data assimilation, the estimated observation uncertainty (or error) of the RO bending angle is typically derived from a fixed-fraction uncertainty model of the RO bending angle. Therefore, the complex and sharp structures can exist in the observation uncertainty as well. This study investigates a potential analysis bias induced in the assimilation of COSMIC (Constellation Observing System for Meteorology, Ionosphere, and Climate) RO bending angles as a result of these complex and sharp structures in the observation uncertainty in the tropical troposphere. It is demonstrated that the complex structures can create a clear negative bias, up to -0.7%, in the bending angle space in the regional analysis with the Weather Research and Forecasting (WRF) model. The WRF analyses and backgrounds also show a larger bias from the COSMIC RO bending angles. Show less