The Value of Radio Occultation Data in Global NWP: An OSSE on the Impact of Profile Volume, Geographical Distribution, and Satellite Bias Correction

Le résumé fourni par la source

Abstract An observing system simulation experiment framework was developed to quantify the impact of increasing the number and improving the distribution of radio occultation profiles on global numerical weather prediction (NWP) skill. Experiments were conducted assimilating between 6,000 and 100,000 daily radio occultation soundings within a representative 2035 observing system. Results show substantial improvements in forecast skill even with a modest number of 6,000 profiles per day, particularly in the stratosphere, with progressively smaller gains as the number of radio occultation soundings increases. Assimilating 6,000 daily radio occultation soundings reduces stratospheric model bias by 1–3 K, acting as an “anchor” that stabilizes the data assimilation system, prevents model drift, and improves the bias correction of satellite radiances. Although improvements do not saturate, returns diminish near 50,000–100,000 radio occultation daily profiles. While this trend is consistent with earlier findings, our study provides a more definitive quantification of these impacts by using significantly higher vertical model resolution within a future-looking observing system framework. Additional experiments show that a more uniform spatial and temporal sounding distribution yields small forecast improvements relative to orbit-driven or polar-concentrated sampling. However, these benefits are much smaller than the impact of any of the configurations tested here relative to a configuration without radio occultation data. These results demonstrate the continued high value of radio occultation observations for global NWP and emphasize the benefits of both increasing the number of daily soundings and optimizing their global distribution.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.