�THE ATMOSPHERE VARIABILITY CLUSTER G2A��

Sean Bruinsma

CNES, Space Geodesy Office

Toulouse, France

​

ISWAT Working Meeting

10 February 2025

ISWAT Atmosphere Variability - G2A: Program

These teams are presently active in G2A:

1. Thermosphere model assessment and improvement (lead: Sean Bruinsma*)
• session G2A-01 in Jamaica on Wednesday AM1&AM2
2. Space weather and lower atmosphere (lead: Jia Yue*)
• no session, but we are doing a reentry study using WACCM-X
3. Satellite aerodynamic modeling (lead: Piyush Mehta*)
• session G2A-03 in Jamaica today PM1&PM2

​

* we are here the entire week

Plus a joint G2A - S2-02 session in Bahamas on Thursday PM1:

Solar proxies for thermosphere modeling

ISWAT Atmosphere Variability - G2A

Contributions to the special issue “Scientific Research and Applications” of the journal Advances in Space Research:

1. Thermosphere model assessment and improvement
• Description and comparison of 21st century thermosphere data (Bruinsma et al., 2022)
2. Space weather and lower atmosphere
• Contribution of the lower atmosphere to the day-to-day variation of thermospheric density (Yue et al., 2022)
3. Satellite aerodynamic modeling
• Satellite drag coefficient modeling for thermosphere science and mission operations (Mehta et al., 2022)

​

​

​

​

As well as to the G2A cluster paper

“Thermosphere and satellite drag”

​

The recommendations given in this paper will be listed next

Uncertainties in drag calculation

The components and data necessary for drag calculation

Sources of uncertainty and concerns:

• Upper atmosphere models
• Upper atmosphere data
• Solar activity: measurements and forecasts
• Geomagnetic activity: measurements and forecasts
• Satellite shape and aerodynamic coefficient modeling
• Orbit extrapolation and uncertainty propagation
• Operational concerns

Recommendations: atmosphere models

1. Empirical model development should be continued,
2. Models based on machine learning, or some hybrid form should be developed and tested.
3. Data assimilation models/approaches have demonstrated superior skill in modelling thermospheric densities in comparison to empirical and physics-based models and are the most promising way forward. (we need an ISWAT DA team – should be achieved this week)
4. The use of indirect observations (e.g. electron densities) to update estimates of thermospheric density must be further investigated.
5. Multi-model ensembles that combine the output of a range of ideally independent models, which can provide an intermediate operational “stopgap” whilst improvements in first-principle modelling, data assimilation techniques and high-rate data become available.
6. Implement a standard and objective model assessment procedure for comparing the performance of models and identify limitations. (CCMC/CAMEL)

Recommendations: upper atmosphere data

1. Consistency between existing and future density datasets must be guaranteed (topics session today)
2. An observation system is needed to achieve significant progress in thermosphere modeling. Simultaneous high quality mass density (e.g. GOCE, CHAMP) and composition (O2, N2, O, He) data by means of mass spectrometers (e.g. DE2) are necessary to revise the composition information in the models.
3. Uninterrupted high-resolution monitoring of geomagnetic storms is needed to augment the current small database of storm events. Only GRACE-FO (data of opportunity!) provides high-resolution total mass density data. (density will be an ESA product of NGGM/MAGIC)
4. temperature, composition, wind, and NO cooling measurements in the lower thermosphere from 100-200 km will complement the total density measurements and close the loop of physics. But observing in situ in that altitude region is very complicated due to the very high atmospheric drag experienced by a satellite, which imposes eccentric orbits and hence limited spatial coverage.
5. Inferring densities for altitudes above 600 km is complicated due to the very low drag and consequently data is sparse there too. Satellites with simple shapes (e.g. sphere, cube), thus minimizing errors due to radiation forces, and equipped with GNSS receivers could at least provide mean densities in support of Earth Observation satellites in the 700-800 km altitude range.

Recommendations: solar activity

1. Maintain the measurement of solar proxies widely used in operations (F10.7, F30, S10, MgII/M10, Y10). The measurement of solar surface magnetism is equally important because it is a key input for physical models. (F10.7 and F30 observations in the future also from Poland and Germany)
2. Provide a common framework allowing to compare the various uncertainties associated with these different solar inputs, what assumptions go into them, and for deriving community-consensus composites.
3. Expand the transition from solar proxies to EUV irradiance observations for operational use in thermosphere models. Recommendations have already been made as how to reconstruct the EUV spectrum from the measurement of a few lines or a few spectral bands. Sufficient overlap in time must be taken into account for consecutive missions to allow for cross-calibration, and in-flight calibration must be performed to enable correcting the measurements for instrument degradation.
4. Continue the uninterrupted measurement of the solar surface magnetic field, which is the primary input for flux surface and for dynamo models, which are respectively needed for short-term and long-term forecasts.
5. Provide a common framework for comparing the many different forecast models that exist : test them on a common time interval or better continuously, and define common metrics of performance. (S2-02 Scoreboards this meeting)

Recommendations: geomagnetic activity

1. Maintain solar wind measurements from L1, which improve predictions up to 12 hours ahead. The results show that most likely no other regression model could improve the prediction significantly.
2. Improvements for longer forecast horizons can only be obtained using different informative source data, such as those coming from solar wind predictions from global heliophysics models. In particular, flux rope modeling in order to correctly propagate the magnetic structure of the CME from the Sun to Earth will provide Bz at Earth.
3. Missions at other Lagrange points (L5, L4) may help.

Recommendations: satellite modeling

1. Operators need (to be motivated) to develop and share accurate and high-fidelity models (geometry and optical properties) for all new satellites and missions.
2. The community needs to come to an agreement on the use of gas-surface interaction models and arrive at consistent density data sets for improving and validating thermosphere models.
3. The community needs to identify a baseline method for modeling of physical drag coefficient.
4. The adopted method must be able to provide realistic uncertainty estimates.

​

NB: In case of collision avoidance, events are driven by the prediction error of the secondary objects in conjunctions, which are often non-active objects. So fitted ballistic coefficients will remain the norm in CA calculations, and small improvements to primary object prediction error are unlikely to manifest notable operational improvements.

Differences in drag calculations are mainly due to aerodynamic and satellite shape models used

(session this afternoon)

Problem: No standard, no consensus…

�THANK YOU

AND HOPE TO SEE YOU IN (ONE OF THE) G2A SESSIONS��

Sean Bruinsma

CNES, Toulouse, France

​

ISWAT Working Meeting

10 February 2025