The geomagnetic storms are the disturbances of the Earth's magnetic field, and the impact of solar wind particles enhancement (Buonsanto, 1999;Gonzalez et al., 1994;Kumar & Kumar, 2019). Following geomagnetic storms, the ionosphere also has obvious disturbance. It shows the critical frequency of F2-layer (foF2) or total electron content (TEC) changes obviously. The large decrease of foF2 or TEC is the ionospheric negative storm, and the significant increase of foF2 or TEC is referred to as positive storms (Fagundes et al., 2016;Maruyama et al., 2004). In general, 𝐴𝐴 𝐴𝐴 region ionization is positively correlated to the density ratio of atomic oxygen [O] to the molecular nitrogen 𝐴𝐴 [N2] . Prölss (1987) showed that the negative storm was caused by the change of neutral gas composition. The enhanced ratio of 𝐴𝐴 [N2] /[O] would lead to the chemical loss rate increase, so the ionospheric electron density decreased.There are two main physical mechanisms of ionospheric positive storms. The first is the equatorward wind surge or disturbance wind (Fuller-Rowell et al., 1994). Joule heating raises the temperature of the upper thermosphere and ion drag drives high velocity neutral winds. The heat source drives a global disturbance wind. It propagates to low latitudes and even into the opposite hemisphere. The equatorward wind pushes the ionosphere up to high altitudes where the recombination is ineffective. As a result, the electron density increases under sunlit conditions (Maruyama et al., 2004;Zhao et al., 2008). Because the equatorward wind surge blows from high to low latitudes, the ionospheric disturbance has time delay. During quiet days, the neutral winds (or background winds) are poleward in daytime and turn equatorward after sunset. During storms, the neutral winds are the combination of background and disturbance winds. The daytime neutral wind, which is caused by Joule heating in the auroral region, even turns to equatorward (Fuller-Rowell et al., 1994). Different storms have different intensity, local time, and latitude coverage. de Jesus et al. ( 2016) studied a positive ionospheric storm at equatorial, low and middle latitudes in African sector. The factor was the equatorward disturbance winds and huge wind circulations.

We investigate the influence of assumed height for the thin shell ionosphere model on the Total Electron Content (TEC) derived from a small scale Global Positioning System (GPS) network. TEC and instrumental bias are determined by applying a grid-based algorithm to the data on several geomagnetically quiet days covering a 10 month period in 2006. Comparisons of TEC and instrumental bias are made among assumed heights from 250 km to 700 km with an interval of 10 km. While the TEC variations with time follow the same trend, TEC tends to increase with the height of the thin shell. The difference in TEC between heights 250 km and 700 km can be as large as ∼ 8 TECU in both daytime and nighttime. The times at which the TEC reaches its peak or valley do not vary much with the assumed heights. The instrumental biases, especially bias from the satellite, can vary irregularly with assumed height. Several satellites show a large deviation of ∼ 3 ns for heights larger than 550 km. The goodness of fit for different assumed heights is also examined. The data can be generally well-fitted for heights from 350 km to 700 km. A large deviation happens at heights lower than 350 km. Using the grid-based algorithm, there is no consensus on assumed height as related to data fitting. A thin shell height in the range 350 – 500 km can be a reasonable compromise between data fitting and peak height of the ionosphere.

No abstract