ROM SAF - NRT Notification Detail

NRT Notification Detail

ROM SAF Report 49: An initial assessment of the quality of radio occultation data from EPS-SG-A1

DATE: 2026-04-28 14:30
TYPE: NRT-GTS

Description

Dear ROM SAF GTS User

Issued: 2026-04-28 14:27 UTC

We are pleased to release ROM SAF report 49 which contains an initial assessment of the quality of radio occultation data from EPS-SG-A1:
https://rom-saf.eumetsat.int/general-documents/rsr/rsr_49.pdf

Abstract

The EUMETSAT Polar System Second Generation Satellite A1 (EPS-SG-A1, hereafter referred to as Metop-D), was successfully launched on 13th August 2025. All the instruments on Metop-D, including radio occultation data, provide critical data for numerical weather prediction (NWP), weather and climate monitoring. Dissemination of radio occultation L1 bending angle data via EUMETCast first started on 6th November 2025, and the Met Office first received data on 17th November 2025, and these data have been monitored since.

Metop-D provides around 1,200 occultations per day, around double the volume from Metop-B and Metop-C separately. L1 bending angle data from Metop-D were compared with those from Metop-B, Metop-C, PlanetiQ and Spire. Notably, Metop-D shows improved O-B standard deviations particularly at higher and lower levels relative to all comparison satellites. Bending angle O-B biases are broadly similar to those of Metop-B, Metop-C, PlanetiQ and Spire in the core region, with slightly larger biases at higher levels (more positive) compared to all the other satellites and in the lower troposphere (more negative) compared to PlanetiQ and Spire. Met Office statistics also show a more positive mean bias in the upper stratosphere, while ECMWF statistics show a negative bias in the same region; this behaviour is consistent across all other satellites.

An NWP experiment assimilating Metop-D data was conducted over a three-month period and verified using the root mean square error, or RMSE. The addition of observations from Metop-D improves the forecast skill, with a mean impact of 0.19% and 0.20%, against observations and ECMWF analyses. The addition of these ~1,200 profiles per day (~1,100 after quality control) shows improvements in all regions across different variables with minimal degradation. The inclusion of Metop-D clearly decreases the standard-deviation ratios for the radio occultation observation type. There is little reduction in the standard-deviation ratios for other observation types.

The timeliness of Metop-D L1 data has not improved relative to Metop-B and Metop-C, averaging 52 and 25 minutes longer to receive data, respectively. This is understood to be due to the data transfer method that the Met Office uses while Metop-D is pre-operational. In comparison to FY-3E, Metop-D is much quicker, having received data 2 hours and 20 minutes sooner than FY-3E for the same time period. This allows the majority of observations to be included within the assimilation window. Timeliness is expected to improve once EUMETSAT makes Metop-D operational. Operational assimilation of bending angles from Metop-D by the Met Office is expected around June 2026.

Kind regards
The ROM SAF Team

Back to NRT Notifications List

Newer

Older

The ROM SAF products are licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0) by EUMETSAT. Whether you're displaying the product as-is or if you have used it as input data, please use the attribution suggested on ROM SAF usage policy and acknowledgement.