@manual{81681,
  author = {David Duncan and Niels Bormann and Mohamed Dahoui and Marijana Crepulja},
  title = {Assessment of the Arctic Weather Satellite in NWP},
  abstract = {The Arctic Weather Satellite (AWS) is an ESA mission launched on August 16th, 2024. AWS is notable for its small size, rapid development from concept to launch, and for providing the first observations of sub-millimetre wavelength channels for operational meteorology. AWS represents the first meteorological mission led by a space agency to manifest so-called “new space” concepts of increased miniaturisation, swift deployment, and independent platforms that enable relatively low-cost constellations of passive microwave (MW) radiometers. The payload of AWS is a newly developed MW instrument, which carries traditional MW sounding channels in the 50 GHz oxygen and 183 GHz humidity bands in addition to sub-mm humidity-sounding channels near 325 GHz. In addition to its stand-alone significance, AWS is also a pathfinder mission for the proposed EPS-Sterna constellation of radiometers that EUMETSAT is considering for launch in the late 2020’s to augment the coverage of EPS-SG MW instruments.<br />
This report assesses the performance of the AWS radiometer via comparison with equivalents from the ECMWF model, the Integrated Forecasting System (IFS). The vicarious approach to calibration and validation (cal/val) analysis takes the IFS as a calibration reference, permitting analysis of AWS radiometric performance in the context of similar instruments (e.g. AMSU-A and MHS) over a wide range of geophysical conditions. It is applied here in the all-sky assimilation framework, in which model clouds are&nbsp;simulated, but scenes with known model biases such as optically thick clouds, sea-ice, and difficult surfaces are screened out, depending on each channel’s sensitivity. The driving philosophy is to maximise the data sample whilst ensuring a balanced statistical sample.<br />
The work herein builds upon two previous EUMETSAT-funded projects, both in scientific terms and on a technical level. The first was a study on the potential impact of an EPS-Sterna constellation of radiometers, which led to several developments that underlie the assimilation of AWS and considered the potential assimilation of sub-mm channels for the first time (Lean et al., 2023; Lean and Bormann, 2024). Second was a study on Metop-SG instruments MWI and ICI in which the method for NWP-based cal/val in the all-sky framework was developed (Duncan et al., 2024b). This previous work set the stage for rapid evaluation of AWS and its unique observations.<br />
Monitoring capabilities for AWS were prepared in advance of the launch and the NWP-based analysis permitted swift feedback of instrument performance to ESA and EUMETSAT during the commissioning phase. Our analysis showed that data quality from AWS is generally good, with biases and noise performance comparable to AMSU-A and MHS but short of ATMS when assessed over broadly similar spatial scales. Of the performance requirements that can be adequately assessed with an NWP model, not all have been met in our estimation, most notably inter-channel biases between channels in feedhorn 1 and those of other horns. Some of these may be possible to address through updates in the level 1 processing (Kangas 2025, pers. communication), though for assimilation purposes most of these outstanding bias<br />
differences are adequately handled by the variational bias correction scheme.<br />
Assimilation follows the all-sky method as applied to heritage MW sounders. Of the 19 channels on AWS, here we assimilate 50 GHz channels 4-7 (equivalent to AMSU-A 5-8), 183 GHz channels 11-15 (equivalent to ATMS 18-22), and three 325 GHz channels (16-18). Four months of assimilation trials indicate a positive impact on short- to medium-range forecasts, particularly for humidity and winds. Z500 RMSE is improved by about 1% out to day 3 in the SH and by about 0.5% in the NH at day 2. Short-range forecasts of humidity and winds are improved, and there is a small improvement in tropospheric temperature as evidenced by radiosonde and ATMS background departures. Addition of the new 325 GHz (i.e. sub-mm) channels shows some benefit on top of the other assimilated channels on AWS. This band has similar water vapour sensitivity as 183 GHz but greater sensitivity to cirrus clouds, EUMETSAT Contract Report 3 AWS Assessment and inclusion of these channels slightly improves short-range forecasts of humidity, particularly at higher latitudes. Fuller exploitation of the cloud information in this band warrants investigation in the future.<br />
The microwave sounder on AWS is a valuable addition to the global observing system and represents the first small-satellite radiometer whose performance and stability meet the requirements of operational NWP. The significant benefit to forecasts from AWS assimilation comes partly from its unique orbital crossing time that complements larger backbone platforms, and this bodes well for the proposed EPSSterna constellation. Due to the positive impacts seen in forecast skill from assimilation of its radiances, AWS has been included in the ECMWF operational assimilation from July 2025.},
  year = {2025},
  journal = {EUMETSAT Contract Report},
  number = {RFQ/21/1383948},
  month = {09/2025},
  publisher = {ECMWF},
  address = {Reading},
  issn = {RFQ/21/1383948},
  url = {&nbsp;},
  doi = {10.21957/232efc59cf},
}