Dataset for "Evaluating the Performance of Empirical Models of Total Electron Density and Whistler-mode Wave Amplitude in the Earth’s Inner Magnetosphere"

This is the dataset for the paper "Evaluating the Performance of Empirical Models of Total Electron Density and Whistler-mode Wave Amplitude in the Earth’s Inner Magnetosphere", submitted to Frontiers in Astronomy and Space Sciences by Qianli Ma.
<br>
Empirical models have been previously developed using the large dataset of satellite observations to obtain the global distributions of total electron density and whistler-mode wave power, which are important in radiation belt dynamics. In this paper, we apply the empirical models to construct the total electron density and the wave amplitudes of chorus and hiss, and compare them with the observations along Van Allen Probes orbits to evaluate the model performance. The empirical models are constructed using the Hp30 and SME (or SML) indices, and considering the plasmapause defined using the occurrence rate of high-density regions. The total electron density model provides a high correlation coefficient with the observation overall, while large deviations are found in the dynamic regions near the plasmapause or in the plumes. The chorus wave model generally agrees with the observation when the plasma trough region is correctly modelled and for the modest wave amplitudes of 10 – 100 pT. The model overestimates the wave amplitude when the chorus is not observed or weak in the plasma trough, and underestimates the wave amplitude when large amplitude chorus is observed. Similarly, the hiss wave model has good performance inside the plasmasphere when modest wave amplitudes are observed. When the modelled plasmapause does not agree with the observation, the chorus and hiss waves are miss-identified by the model compared to the observation, and large modelling errors occur. In addition, strong (&gt;200 pT) hiss waves are observed in the plumes, which are difficult to capture using the empirical model. We also evaluate four metrics for different empirical models which are parameterized by different indices. Among the tested models, the empirical model considering a plasmapause and controlled by Hp* (the maximum Hp30 during previous 24 h) and SME* (the maximum SME during previous 3 h) or Hp* and SML has the best performance with low errors and high correlation coefficients. Our study indicates that the empirical models are applicable for the prediction of density and whistler-mode waves with modest power, but large errors could occur especially when the plasmapause is highly dynamic or in the plumes.