Task 1.1 Evaluation of initial future EPS-Sterna constellations with 50 and 183 GHz

Task 1.1 Evaluation of initial future EPS-Sterna constellations with 50 and 183 GHz
Date Published
Eumetsat Contract Report
Katie Lean
Niels Bormann
Sean Healy
Abstract Observations from space-borne microwave (MW) sounding instruments currently provide a very important contribution to accurate Numerical Weather Prediction (NWP) forecasts. Complementing a continuing backbone of high-performance satellite platforms, small satellites are expected to become a significant component of the future observing system. The future EPS-Sterna constellation will consist of small satellites carryingMWsounding instruments. The EPS-Sterna instrument comprises well-established frequencies in the 50 and 183 GHz bands (plus complementary window channels) and a new set of channels around 325 GHz not seen on a space-borne platform before.
In this first phase of the project, we simulate the potential benefit of initial options for the EPS-Sterna constellations with focus on the 50 and 183 GHz frequencies. The different scenarios are assessed using the Ensemble of Data Assimilations (EDA) method where changes to the EDA spread (indicating changes in forecast uncertainties) inform the relative benefits of different constellations. A framework to flexibly simulate and assimilate MW sounders on small satellites within the all-sky framework has already been developed at ECMWF. We build on this existing work here and present a summary of the different steps and modifications required for the EPS-Sterna instrument. The simulated EPS-Sterna data and accompanying observation errors are added to a baseline of existing observations. Four EPS-Sterna scenarios are considered: a configuration proposed for nominal operational use consisting of six satellites in three orbital planes, two degraded options which add three satellites in three orbital planes and four satellites in two orbital planes respectively, and an enhanced option of eight satellites in four orbital planes. In addition to testing the different constellations, several EDA experiments are also run which change either the number of existing MW observations or deny entire Metop platforms relative to the same reference baseline.
The EDA analysis reveals that all four EPS-Sterna scenarios produce a significant positive impact for a range of atmospheric variables, including temperature, geopotential, wind and relative humidity. The impact increases with the number of satellites, and the six-satellite scenario gives just under twice the impact of the three-satellite scenario. Impacts are largest in the southern hemisphere and smallest in the tropics, in line with the trend for changes to real MW sounding observation numbers. Comparison to EDA impacts from changes to real data reveals that EPS-Sterna is able to achieve additional benefit that compares well to the magnitude of impact seen with similar numbers of existing MW sounders. Based upon EDA spread reductions across a range of variables, we would expect the six-satellite EPS-Sterna constellation to have an impact of between one and two full Metop platforms and thus provide significant additional benefits for global NWP, provided the EPS-Sterna data quality is acceptable. However, it is noted that this constellation of small satellites cannot replace the full, wide-reaching impacts of a high-performance, multi-sensor platform. Further sensitivity EDA experiments explore the impact of instrument noise performance by separately inflating or reducing the values. A degradation of 30% in instrument noise values for the six-satellite scenario equates to the loss of between one and two satellites in the constellation in the extra-tropics. A reduction in noise of around 60% in only the temperature sounding channels resulted in further spread reduction exceeding the benefit from adding two more small satellites in the extra-tropics, emphasising the
importance of pursuing good noise performance especially for temperature sounding channels.
We also use the EDA method to demonstrate potential large positive impact from combining the sixsatellite EPS-Sterna constellation with a future Doppler Wind Lidar (DWL) based on EPS-Aeolus (assessed in a separate study), the follow-on mission to Aeolus. The resulting benefit seen globally and across the pressure levels from assimilating the two missions together arises from good complementarity of the independent mission contributions to the EDA spread reduction and the largely additive nature of the separate impacts. For example, large positive impacts on tropical wind in the combined experiment is attributed mainly to the DWL while impacts on geopotential height in the extra-tropics have a higher contribution from EPS-Sterna.
URL https://www.ecmwf.int/en/elibrary/81511-task-11-evaluation-initial-future-eps-sterna-constellations-50-and-183-ghz
DOI 10.21957/0a695fcc39