A method to predict thermospheric mass density response to geomagnetic disturbances using time‐integrated auroral electrojet index

Using the thermospheric mass density measurements from the European Space Agency's Gravity field and steady state Ocean Circulation Explorer (GOCE) satellite, we develop a new empirical geomagnetic disturbance time correction based on integrating the auroral electrojet (AE) index. For this, a US Naval Research Laboratory Mass Spectrometer and Incoherent Scatter Radar (NRLMSISE-00) model with no geomagnetic parametrization is subtracted from the GOCE densities, and the regressions between the time-integrated AE index and density residuals are computed as a function of latitude, solar time, and day of year. When we add this correction to the quiet time reference NRLMSISE-00 model, it increases the model's disturbance time correlation with the 270km normalized GOCE densities from 0.71 to 0.86. We assess the effect of integrating thermospheric density proxies with respect to time using both geomagnetic and solar indices and discover that the integration of AE, a(p), and the epsilon parameter significantly increase their correlation with the orbit-averaged GOCE densities. We compare the predictions of our empirical correction with the NRLMSISE-00 and Jacchia-Bowman (JB2008) models, and significant deviations from the measurements are discovered. The NRLMSISE-00 is confirmed to generally underestimate the density enhancement, and the latitudinal shape of the predicted response shows too low enhancements at middle latitudes. Even though these are not issues for the JB2008 model, it performs weaker than the NRLMSISE-00 at reproducing the orbit-averaged densities. This unexpected result is attributed to the weakness of the a(p) parametrization, which is used in the model during smaller disturbances.