The First Comparison Between Swarm‐C Accelerometer‐Derived Thermospheric Densities and Physical and Empirical Model Estimates

Abstract

The first systematic comparison between Swarm-C accelerometer-derived thermospheric density and both empirical and physics-based model results using multiple model performance metrics is presented. This comparison is performed at the satellite's high temporal 10-s resolution, which provides a meaningful evaluation of the models' fidelity for orbit prediction and other space weather forecasting applications. The comparison against the physical model is influenced by the specification of the lower atmospheric forcing, the high-latitude ionospheric plasma convection, and solar activity. Some insights into the model response to thermosphere-driving mechanisms are obtained through a machine learning exercise. The results of this analysis show that the short-timescale variations observed by Swarm-C during periods of high solar and geomagnetic activity were better captured by the physics-based model than the empirical models. It is concluded that Swarm-C data agree well with the climatologies inherent within the models and are, therefore, a useful data set for further model validation and scientific research. Plain Language SummaryThe first systematic comparison of the thermospheric densities between Swarm-C accelerometer-derived densities is presented. The data are compared to two latest versions of a physics-based general circulation model and arguably the most widely used empirical models using multiple model performance metrics. The comparison at the satellite's temporal resolution provides a useful evaluation of the models' fidelity for orbit prediction and pertinent space weather forecasting applications. The results of this study show that the short-timescale variations observed by Swarm-C during periods of high solar and geomagnetic activities were better captured by the physics-based model than the empirical model. The results show the complex interconnectedness of solar activity and geomagnetic activity on model performance in terms of estimating density. The study also uses a machine learning exercise to demonstrate characteristics inherent to each model run. Swarm-C data are consistent with the models. It is concluded that Swarm-C data are suitable for further model validation and scientific research.

Density‐Temperature Synchrony in the Hydrostatic Thermosphere

“…Following the conclusions in Kodikara et al (2018), one could expect the seasonal monthly mean results presented above to not change much based on the Weimer (2005) or Heelis et al (1982) ionosphere convection models. Nevertheless, an extended comparison with different high-latitude forcing methods during storm/quiet times may help to quantify the sensitivity of -T synchrony in the high latitudes even further.…”

“…Kodikara et al () discussed that, compared to accelerometer‐derived densities, the performance of both Weimer () and Heelis et al () ionosphere convection models are comparable under 2014/2015 geophysical conditions. Both Kodikara et al () and Wu et al () provided examples of the Weimer () model outperforming Heelis et al () model during storm times. While the type of ionosphere convection model is not expected to influence the seasonal, monthly mean results much, it may have an impact on the storm time results.…”

“…TIE-GCM's ability to reproduce the thermospheric density and temperature dynamics has been validated in some studies (see Emmert, 2015;Maute, 2017;Qian et al, 2014;Shim et al, 2012, and references therein). In particular, Kodikara et al (2018) validated the TIE-GCM version 2.0 using accelerometer-derived density data for 2014/2015, which is part of the period considered in this analysis. Below is a comparison of ground measurements of thermosphere temperature from multiple stations with TIE-GCM.…”

Quantifying the Storm Time Thermospheric Neutral Density Variations Using Model and Observations

“…The Challenging Micro-Satellite Payload (CHAMP) and Gravity Recovery and Climate Experiment satellites are the most used satellites for the investigations of the neutral density and the associated atmospheric drag acting on satellites (Anderson et al, 2009;Bruinsma, 2015;Bruinsma & Forbes, 2010;Bruinsma et al, 2018;Huang et al, 2014;Liu et al, 2011;Picone et al, 2002;Xu et al, 2011). Recently, data from Swarm constellation has also been employed to derive the thermospheric neutral densities (Kodikara et al, 2018;Siemes et al, 2016;Zesta & Huang, 2016). In this kind of approach, the densities are calculated from the accelerometers on the spacecraft (Sutton et al, 2005).…”

“…Several metrics are employed to assess the model performances. For the neutral density studies, the most used metrics are the mean absolute error (MAE), bias (B), correlation (R), root mean square error (RMSE), standard deviation (Std), prediction efficiency (PE), ratio of maximum and ratio of average (Bruinsma, 2015;Elvidge et al, 2014Elvidge et al, , 2016Emmert et al, 2017;Kodikara et al, 2018;Pardini et al, 2012;Shim et al, 2012), and the version of the metrics in log space (Bruinsma et al, 2018;Picone et al, 2002;Sutton, 2018). Each of these metrics has advantages and disadvantages (Hyndman et al, 2006;Shcherbakov et al, 2013).…”