Journal Article

Observation uncertainty and impact of Mode‐S aircraft observations in the Met Office limited area numerical weather prediction system

Simonin, David, Waller, Joanne A., Song, Taejun

2025

Overview

Aircraft observations derived from Mode‐Select Enhanced Surveillance (Mode‐S EHS) reports are a valuable, high temporo‐spatial resolution, source of upper‐air information that can be assimilated into numerical weather prediction models. At present temperature and wind Mode‐S EHS observations are assimilated into the Met Office's convection‐permitting model, the UKV. These observations are obtained from two different sources, an inhouse set of receivers and via the European Meteorological Aircraft Derived Data Centre (EMADDC). Currently, Mode‐S EHS data are assimilated using the same observation error standard deviation profiles as AMDAR data; however, differing observation processing is anticipated to result in differing error profiles for the Met Office and EMADDC data and for the AMDAR data. Therefore, we estimate new observation error statistics, including error correlations for the two types of Mode‐S EHS data. We also consider the impact of the different aircraft data on the UKV analysis. We find that the observation error standard deviation profiles for wind and temperature are dependent on observation type and season and differ from the current profiles used in the assimilation. Additionally, the Mode‐S EHS observation errors have a considerable spatial correlation that increases with height and is much longer than the spatial thinning distance. The estimated observation influence shows that Mode‐S EHS data are not optimally assimilated, and that the use of updated, observation‐type specific, error profiles is expected to improve the assimilation. The assimilation may be further optimized by modifying the observation thinning distance or including the correlated observation errors in the assimilation. Aircraft observations are a valuable source of information that can be assimilated into numerical weather prediction models. Using data from the Met Office regional system we assess the observation uncertainty and assimilation impact of aircraft observations. Our new results suggest that the current observation uncertainties used in the assimilation are not correct and, as shown by the calculated assimilation impact metrics, the data are not optimally assimilated. Assimilation of aircraft data may be improved by assigning updated observation error profiles and by modifying the observation thinning distance or accounting for correlated observation error statistics.

• Publisher: John Wiley & Sons, Ltd,Wiley
• Subject: aircraft‐based observations / data assimilation / degrees of freedom from signal / estimation of observation uncertainty / mode‐S / observation impact