The European Space Agency (ESA) Arctic Weather Satellite (AWS) currently planned for launch in 2024 could be a pathfinder mission in expanding the support of the EUMETSAT Polar System Second Generation (EPS-SG) Microwave Sounder (MWS) mission to global and regional numerical weather prediction (NWP) applications.
A successful AWS in-orbit demonstration in the 2024-2025 period would represent an opportunity for EUMETSAT to expand the products’ envelope of the EPS-SG mission for the users, by considering the implementation of a constellation of satellites with microwave sounding capability based on the AWS design.
ESA and EUMETSAT are cooperating on the preparatory activities that could possibly lead to a constellation of flying AWS recurrent models, providing sounding information on global humidity and temperature profiles to the users in near real time. The Phase A activities for the constellation definition are currently ongoing at EUMETSAT, and include various complementary scientific impact assessment studies, performed by CNRS/Météo France, the European Centre for Medium-range Weather Forecasts (ECMWF) and by a Consortium led by Met Norway (MetNO), including the Swedish Meteorological and Hydrological Institute (SMHI) and the Finnish Meteorological Institute (FMI). All studies are considering the same constellation scenarios, in terms of orbits, number of satellites, and instrument sampling.
The main goal of the various studies is to assess the impact of various possible constellation designs on NWP using different methodologies and focussing on different, but complementary aspects. CNRS/Météo France is performing a series of Observing System Simulation Experiments (OSSEs) considering a realistic representation of the global observing system, with focus on global NWP impact. Forecast lead times are considered over the range of few days.
ECMWF is performing an Ensemble of Data Assimilations (EDAs) series of experiments, assessing the benefit of the various constellation scenarios by measuring the reduction in variation across different forecast ensemble members, with focus on global NWP. The EDA methodology requires only data from the various constellations that need to be simulated, including a realistic observation error. Focus is on the short term forecast error reduction.
An additional study, performed at MetNO, SMHI and FMI, includes a series of regional OSSEs aiming at estimating the expected impact of the selected constellations scenarios with focus on regional NWP at high latitudes. Other important aspects to be investigated include the support of Nowcasting (NWC) and a societal impact assessment for high latitude regions and the Arctic.
This presentation describes the objectives and status for each of these studies and gives an outlook of the various future activities to support the definition of the AWS constellation.

The Arctic Weather Satellite (AWS) aims at paving the way towards a constellation of satellites, carrying a microwave instrument each, to give input to weather forecasting with short revisit times. However, already the prototype satellite will provide a novel element, as it is equipped with four channels around the 325 GHz water vapour transition. Existing operational microwave radiometers are limited to frequencies below 200 GHz. The upcoming Ice Cloud Imager (ICI) mission will also cover this range. Presently AWS has an earlier planned launch date than ICI and the later instrument will only have three 325 GHz channels. In addition, ICI is a conically scanning instrument with a footprint of 15 km while AWS is a cross-track scanner with a resolution around nadir of about 10 km.

For clear-sky conditions, AWS' 325 GHz channels will provide a somewhat improved precision for humidity, by complementing the channels around the 183 GHz transition. There is little information on surface emissivities around 325 GHz, and it is hard to judge if the addition of this frequency range can help constrain atmospheric humidities when the 183 GHz channels are disturbed by the surface. This happens in dry conditions and at high surface elevations. In the presence of clouds, there is a much more clear synergy between the two sets of channels. In assimilation of clear-sky character, 325 GHz should be the ideal complement to filter out cloud affected 183 GHz data. In simulations, this approach has also been shown to clearly outperform existing filtering methodologies. In fact, a significant fraction of the cloud-affected 183 GHz radiances can be corrected to create synthetic cloud-free counterparts with help of the 325 GHz channels. This indicates that the combination of 183 and 325 GHz allows to constrain humidity below and inside a broader range of clouds than what is now the case with just 183 GHz at hand. Assimilation of all-sky type should be the optimal manner to make use of this synergy between 183 and 325 GHz.

The information on ice hydrometeors provided by the 325 GHz channels has also inherit value, in line with the objective of ICI. Fewer cloud variables can be constrained by AWS, as ICI has additional channels further into the sub-mm range, up to 664 GHz. Still, AWS's partly smaller footprint and varying incidence angle can be beneficial for special studies, e.g. to verify assumptions on ice particle shape and orientation, and it can thus indirectly support ICI.

Context: The arctic region is poorly served by geostationary observations. The arctic weather satellite (AWS) will provide frequent coverage of the polar regions to support nowcasting and numerical weather prediction. The AWS Prototype Flight Model (PFM) satellite is the forerunner of a potential constellation of sixteen satellites that would supply a constant stream of temperature and humidity data.
Instrumentation: Onboard the AWS four receivers (19 channels) will perform passive remote sensing of the atmosphere with frequency coverage between 50 and 325 GHz. Radiometer Physics GmbH is in charge of developing and building the 325 GHz Front-End featuring 4 channels especially designed for humidity sounding and cloud detection. It relies on extensive heritage from the MetOp-SG Ice Cloud Imager receiver developments, but considers the New Space approach to prove innovative integration concepts in a cost-effective and timely manner. The 325 GHz receiver integrates in three aluminum based modules, all DC and RF functionalities that were previously separated (ICI), allowing for strong mass and size reductions.
New Space Approach: The involvement of private companies, like RPG, and investors in the commercial space sector has led to the “New Space” development. The most significant characteristics of the “New Space” approach include the use of new production methods, the incorporation of new cutting-edge technology, modularization and standardization, and the use of commercial off-the-shelf parts, as well as the willingness to take on higher risks while still remaining flexible enough to react quickly on customer needs. This approach enables us to incorporate novel business ideas in designing and manufacturing while still save development time and reduce costs.
Here, we will present the 325 GHz AWS Front-End with special focus on the chances and implications that New Space has on its design, development and manufacture.

The ESA Raincast study is a multi-platform and multi-sensor study to address the requirement from the research and operational communities for global precipitation measurements. It aims at identifying and consolidating the science requirements for a satellite mission that could complement the existing space-based precipitation observing system and that could optimally liaise with efforts currently made by other agencies in this area. One objective in the study is to provide criteria and guidelines in the design of future missions dedicated to global snowfall quantification.
Improvement in both the monitoring of high latitude precipitation and in our understanding on microphysical and dynamical processes that influence high latitude precipitation patterns, intensity and type must be driven by concerted observations of active radars and passive microwave radiometers. This has been recently demonstrated through the development of machine learning-based algorithms for snowfall detection and retrieval, exploiting global observational datasets built from passive and active microwave spaceborne sensors. In particular, the CloudSat/Calipso-based machine learning snowfall retrieval methodology developed for the GPM Microwave Imager (GMI) (SLALOM), which has been developed within the EUMETSAT Hydrology SAF in preparation for the EPS-SG Microwave Imager (MWI) mission, has proven to be very suitable for snowfall detection and retrieval. SLALOM is able to reproduce CloudSat snowfall climatology, but with better coverage (up to 65°N/S for GMI), outperforming other state-of-the-art GPM products.
The increasing number of operational cross-track scanning radiometers in the future (e.g., EPS-SG Microwave Sounder (MWS) mission) requires dedicated efforts to study the potentials of these radiometers to improve snowfall global monitoring. Moreover, the Arctic Weather Satellites (AWS) mission, carrying a cross-track scanning microwave radiometer covering the frequency range 50–325 GHz, will provide unprecedented spatial and temporal coverage at high latitudes. In this context, SLALOM has been recently adapted and applied to the currently available most advanced cross-track scanning radiometer, the Advanced Technology Microwave Sounder (ATMS), on board Suomi NPP, NOAA-20 and the future JPSS platforms. A dedicated study has been carried out to assess ATMS snowfall observation capabilities at high latitudes. The study is based on the use of a ATMS/CloudSat-Calipso coincident observation dataset. The main findings from the study will be presented by: 1) reporting on the different scientific aspects and on the complexity related to snowfall detection and quantification in extreme dry/cold conditions (e.g., sea ice/snow cover variability), 2) analysing and providing evidence of such complexity, and 3) proposing observation and retrieval strategies to be adopted in the future to improve detection and quantification of snowfall in the Arctic, also in view of the AWS mission. These findings pave the way towards the definition of synergistic approaches exploiting the future European AWS, EarthCare and CIMR missions.

The Meteorological institutes of Denmark, Finland, Norway, and Sweden have a long history of working together to advance limited-area NWP in conditions typical to these Nordic countries. The maintenance of the state-of-the-art operational NWP systems is currently taking place in the HARMONIE-AROME framework with close links to Meteo-France and several other European National Weather Services. The development of variational data assimilation in these mesoscale NWP systems facilitates efficient exploitation of satellite measurements. In anticipation of the new polar orbiter launches within the next 2-4 years, preparations are underway for the assimilation of radiances from the Arctic Weather Satellite (AWS).

AWS is designed as a prototype satellite to demonstrate the feasibility of low-cost microwave sounding from a low-Earth orbit. A successful demonstrator mission will pave the way for a constellation of satellites, thereby providing frequent data reception over high latitude regions. The AWS demonstrator mission is scheduled for launch in 2024. While the AWS satellite has a design-life time of 5 years the ESA mission is committed only for one year of operation.

Under ESA contract a Nordic Consortium has committed to make an early performance evaluation of the AWS data, with emphasis on regional NWP and high latitudes. The project is to be kicked off in December 2021 and will finish one year after launch. To be able to do a meaningful evaluation in this very short time frame, significant research and development efforts are required to be able to bring the Nordic NWP systems ready to digest and optimally use the AWS data by the time of launch. In addition this study will also try to assess what impact a possible future AWS constellation, providing an observation frequency over the Nordic region of upto and possibly beyond 1 hour, might have on short term weather forecasting and Nowcasting.

In all Nordic Meteorological Services, microwave sounding data assimilation is performed already in clear sky conditions. The Nordic ESA study will start to monitor and experiment with active AWS radiance assimilation using both temperature- and humidity-sensitive microwave sounding channels. At the time of launch, the operational use of these frequencies is likely restricted to cloud-free conditions. But substantial efforts are devoted to eventually extending the data use to all-sky conditions and improving the usage over sea ice, land, and snow surfaces. Additionally, research will be undertaken to investigate the potential use of channels in the sub-millimetre wavelengths. For example an enhanced cloud filtering for clear-sky assimilation using the 325 GHz channels will be evaluated.

In order to have access to real-time data with low latency from shortly after launch, a ground segment is set up using the satellite Direct Readout acquisition facilities available within the domain of the Nordic Institutes.

In parallel to the development of the AWS demonstrator mission EUMETSAT is currently conducting Phase 0/A studies for a future constellation of small microwave sounding satellites. This potential AWS constellation is foreseen as an expansion to the EPS-SG programme and is expected to be put forward for decision in 2025. If approved by the EUMETSAT member states the first satellites of this constellation will start flying around 2029. To support the decision process at EUMETSAT a number of dedicated global and regional assessment studies will be performed. In a Nordic study to start early 2022 and finish mid 2023 an Observing System Simulation Experiment (OSSE) over the Nordic/Arctic region will be conducted. In addition, a first look at how AWS data can be used to support Nowcasting (independent of what can be provided by short term regional NWP) will be addressed, and a study will include first steps towards making socio-economic benefit assessments of a future AWS constellation.

Here we will present the development and research plans of the Nordic Consortium to support ESA and EUMETSAT in the evaluation of the performance of the AWS demonstrator mission data and to assess the impact of a possible future AWS constellation, with focus on Nordic regional forecasting.

Observations from microwave (MW) instruments currently provide the greatest impact of the satellite data used in the ECMWF assimilation system. However, with some satellites having already been in flight for many years combined with a lower frequency of new missions, the constellation as it stands now is expected to decline in the coming years. Recent advances in technology have allowed the possibility of launching MW sounding instruments on small satellites with a performance that is expected to be adequate for Numerical Weather Prediction (NWP). These new small satellites are expected to complement a continued backbone of larger, high performance platforms. In this study, which is carried out in collaboration with ESA, we aim to investigate different potential future constellations of small satellites carrying MW sounding instruments with a focus on the optimal design for global NWP. A range of constellations have been chosen where the different designs have been primarily motivated by two aspects. Firstly, by varying the number of satellite platforms up to a large constellation of 20 small satellites, we aim to examine how much further benefit could be achieved with improving temporal sampling. Additionally, the trade-off between humidity sounding channels only or having additional temperature sounding channels will be considered.

The impact of these possible constellations will be evaluated by the Ensemble of Data Assimilations (EDA) method. The EDA consists of running a finite number of independent cycling assimilation systems, in which observations and the forecast model are perturbed to generate different inputs for each member. The small satellite data and accompanying observation errors will be simulated and the benefit from adding the data to the observing system is measured by reducing the spread of the ensemble members which reflects improvement to the uncertainties in analyses and forecasts.

Here we present the EDA methodology and provide an overview of the simulation framework which includes use of operational resolution model fields interpolated to the small satellite observation locations and a scattering radiative transfer model to allow use of the data in an all sky framework. Several EDA experiments will be run in which the different constellations of simulated small satellite data are added to a consistent baseline observing system. This baseline comprises the full observing system but with the number of existing MW sounders reduced to only four satellite platforms - two each in the Metop and JPSS orbits - to reflect a potential reduced constellation of larger platforms in the future. Preliminary evaluation of the first EDA experiments exploring the impact of the small satellite constellations will be presented.