TEC data used are from Koudougou GPS station. The receiver was provided by the Ecole Nationale de Télécommunication of Bretagne (ENST Bretagne) as part of the International Heliophysical Year (IHY) project initiated by Europe-Africa Study and Research Group (GIRGEA). This receiver has been installed at University Norbert ZONGO since November 2008. TEC data recorded cover solar cycle 24, except for the years 2008; 2018 and 2019. The geomagnetic index data aa and the dates of Sudden Storm Commencement (SSC) were used [23] to elaborate the pixel diagram. This diagram represents geomagnetic index aa evolution as a function of solar activity for each Bartels rotation [24]. The mean sunspot number (Rz) annual was used for the division of the solar cycle 24 into phases.

The geomagnetic index aa is deduced from K-index measured at two mid-latitude antipodal stations. This index measures the amplitude of global geomagnetic activity during 3-hour intervals normalized to geomagnetic latitude ±50˚. It was introduced by Mayaud [25] to monitor geomagnetic activity over the longest possible period. The daily average of the 8 tri-hourly values per day is noted Aa. An SSC corresponds to an abrupt change in the geomagnetic field followed by a geomagnetic storm that lasts less than an hour. The dates of SSCs and the values of the aa index since 1869 are available on the ISGI website (http://isgi.unistra.fr). Rz values are available on the OMNIWEB website: https://omniweb.gsfc.nasa.gov/form/dx1.html.

Legrand and Simon [26]; and Richardson et al. [27] [28] elaborated geomagnetic activities’ first classification from pixel diagram. They classified geomagnetic activities into four classes: 1) quiet activities associated with slow solar winds (V < 450 km/s); 2) recurrent activities caused by fast solar winds from coronal holes (V > 450 km/s); 3) shock activities related to shock waves due to Coronal Mass Ejections (CMEs); and 4) fluctuating activities caused by fluctuations in the Sun’s neutral blade. Ouattara and Amory-Mazaudier [29] continued, in the same dynamics, by improving this method.

This classification was further improved by Zerbo et al. [30], who pushed the limits of old classification (AC) by providing clarification of solar origin of about 20% of the geomagnetic storms in addition to the 60% explained by AC. In new classification (NC), days of quiet activity, associated with slow solar winds, correspond to days when Aa < 20 nT and disturbed activity to days when Aa ≥ 20 nT. Disturbed days include: 1) fluctuating activity days (AFs) or fluctuating events (FEs) caused by fluctuations in the Sun’s neutral blade, 2) shock events (SEs) including shock activity (SA) and cloud shock activity (CSA), and 3) recurrent event days (REs) including classical recurrent activities of AC (RAs) and Corotating Moderate Activities (CMA) due to stable corotating neutral winds.

Geomagnetic activity days selection is done using the pixel diagram (Figure 1) proposed by Simon and Legrand [31] and improved by Ouattara and Amory-Mazaudier [29] who organized it in columns and rows; then they defined a color code to identify different types of geomagnetic activities. A line in the diagram corresponds to a solar rotation (27 days). The SSC dates are indicated by circles surrounding the corresponding aa index value. The dates of the beginning days of Bartels cycle, the legend and the year are shown on the left, right and above the pixel diagram respectively. According to pixel diagram color code, different geomagnetic activities days are selected as follows:

1) Quiet activities (QA) corresponding to days with Aa < 20 nT which are represented by white and blue boxes;

2) Recurring events (RE) including:

a) classical recurrent activities (RA) of AC corresponding to days with Aa ≥ 40 nT and spanning one or more Bartels rotations without SSC; these days are represented by orange, red, and/or dark red boxes on at least two successive days without SSC and on at least two solar rotations;

b) and CMA days corresponding to days with 20 nT ≤ Aa < 40 nT and spanning one or more Bartels rotations without SSC; these days are represented by yellow or green boxes on at least two successive days without SSC and on at least two solar rotations

3) Shock events (SE) including:

a) Classical shock activities of AC (SA) corresponding to SSC days with Aa ≥ 40 nT; these days are represented by a set of 1, 2, or 3 days represented by orange, red, and/or olive-red boxes with SSC in phase beginning and no recurrence of SSC during 1, 2, 3, or 4 Bartels rotations;

b) and CSA days corresponding to SSC days where 20 nT ≤ Aa < 40 nT; these days are represented by a set of 1, 2, or 3 days represented by yellow and green boxes with SSC at phase beginning and no recurrence of SSC during 1, 2, 3, or 4 Bartels rotations

4) Fluctuating events (FE) which include all days not included in three previous classes.

Solar cycle is divided into phases according to criteria proposed by Ouattara and Amory-Mazaudier [29]. These criteria related to the annual average of sunspots number Rz are defined as follows: 1) minimum phase: Rz < 20; 2) ascending phase: 20 ≤ Rz ≤ 100 with Rz greater than that of previous year; 3) maximum phase: Rz > 100 noting that for weak solar cycles (with Rzmax< 100) the years of maximum phase correspond to those with an index Rz > 0.8 Rzmax, and 4) decreasing phase: 100 ≥ Rz ≥ 20 with Rz lower than that of the previous year.

However, since 2015, a new set of Rz values different from the one used by Ouattara and Amory Mazaudier [32] is available on the OMNIWEB website (https://omniweb.gsfc.nasa.gov/form/dx1.html). Considering the previous Rz values that are limited to 2014, cycle 24 is low because its peak that corresponds to the year 2014 is less than 100 (Rz = 78.9). Since the previous Rz values (R_{Z_previous}) do not exist beyond the year 2014, equations 1 and 2 were used to find the approximate values of R_{Z_previous} equivalent to those of the Rz new values (R_{Z_new}) for the missing years of cycle 24 (2015 to 2018) in order to be able to use the criteria of Ouattara and Amory-Mazaudier [32]. Table 1 summarizes solar cycle 24’ previous, new and approximate Rz values.

σ 1 = R Z a n c i e n R Z n o u v e a u (1)

σ 2 = ∑ 2008 2014 σ 1 7 (2)

Equation (2) gives σ 2 = 0.679709088

R_{Z_approximate} is the approximate value of R_{Z_previous}. It’s estimated by the Equation (3).

R Z _ a p p r o x i m a t e = R Z n e w × σ 2 . (3)

Table 2 shows results of solar cycle 24 cutting into phases.

Several studies have demonstrated TEC variability seasonal dependence. Some have revealed the existence of semi-annual variations characterized by electron density peaks at equinoxes [33] [34] [35] [36]; while others revealed a winter anomaly existence characterized by peaks in winter versus summer [33] [34] [35] [36]. Seasons are defined as follow: winter (December, January, February), spring (March, April, May), summer (June, July, August), and fall (September, October, November).

The error bars placed on TEC profiles of the QA, for the qualitative analysis of FE compared to QA, are obtained by δ = V where V is the variance given by Equation (4).

V = ∑ i = 1 N ( T E C i − T E C ¯ ) 2 N (4)

With TEC_i the hourly values of TEC; T E C ¯ the average hourly value of daily hourly values and N the total number of days depending on solar phase or season considered.

Based on the definition of storm strength as defined by Vijaya Lekshmi et al. [37], the relative deviation of TEC (δTEC), allowing the type and the intensity of the observed storm to be assessed, is defined by the Equation (5):

δ T E C = T E C F ¯ − T E C q ¯ T E C q ¯ × 100 (5)

where T E C F ¯ and T E C q ¯ denote the average hourly values of TEC in fluctuating and quiet periods respectively.

The storm is qualified as positive, respectively negative, when δTEC > 0, respectively δTEC < 0.

For δTEC values between ∓ 20% the storm is considered to be weak or moderate; otherwise, it is called an intense storm.

The solar cycle 24’s (January 1, 2008 to December 31, 2018) FE and QA days were counted in order to analyze their occurrences. The study included 2939 QA days and 697 FE days. The occurrence rates are obtained using the formula in Equation (6).

% O c c = N A N T × 100

where N_A is the number of QA or FE days per solar phase or season considered and N_T is the total number of days per solar phase or or by the considered season.

From analysis of impact of FEs on the diurnal variations of the TEC, it appears that:

1) FEs generate positive storms during all the day time and during all phases, except at maximum phase. Moderate negative storms are observed at solar maximum (from 0000 LT to 1800 LT) and weak negative storms at solar minimum around 0300 and 0700 LT. Concerning seasons, positive storms are observed during all hours in winter and autumn. For spring, negative storms are observed during all the day time, except from 2000 to 2100 LT. But for summer, positive storms are observed from all the day, except from 0000 LT to 0300 LT and from 2100 LT to 2300 LT.

2) Troughs are observed at 0500 LT and 2000 LT during all solar phases and seasons. Patel et al. [38] and Zoundi et al. [21] also found similar results by analyzing TEC values of Surat and Niamey stations respectively.

The troughs observed at dawn reflect a decrease in photoionization at night and an increase in recombination events for both the FE and QA [39] [40]. The abrupt increase in TEC observed at sunrise is related to extreme ultraviolet radiation (EUV) and E × B upward vertical drift [41] [42]. According to Krishnank et al. [43], this increase is attributed to southerly winds.

3) Peaks are observed around 1400 TL during all solar phases. Concerning seasons, peaks are observed around 1500 LT during all seasons, but for winter season they are observed around 1400 LT. Peaks observed for FE’s curves are more significant than those of QA’s curves, except during maximum phase and spring months. Peaks represent the greatest ionization during the day [21] [38].

4) The strongest positive storms are observed during the solar minimum phase and during the winter and fall months. Analyzing foF2 data variations at Dakar station, Sandwidi and Ouattara [44] observed similar results by showing that the weakest storms are observed during the increasing and decreasing phases and in summer months. FE predominate in the increasing and decreasing phases and in spring. This suggests that the strength of positive storms due to FE is not related to the occurrence of FE days.

5) Concerning, negative storms, the most intense are observed during maximum solar phase and spring season and the weakest during minimum solar phase (0300 LT and 0700 LT) and summer season (0300 LT and 2300 LT). Sandwidi and Ouattara [44] also found that the most negative storms are observed during maximum solar phase and spring months. It’s also observed that FE are predominant in maximum than in minimum phase; and at spring than in fall. This suggests that on solar phases and seasonal scale, the strength of FE’s negative storms is related to the occurrence of FE’s days.

6) Most of negative storms are observed at night (0000 LT - 0200 LT) when chemical recombination is predominant, except in spring when they are recorded during all the day time. According to Fuller-Rowell et al. [13] negative storms result from variations in thermosphere composition during magnetic storms. Thus, an acceleration of the recombination process (with an increase in the N₂/O ratio) can be observed at solar maximum (~0000 LT and ~0400 LT), in spring (~0000 LT to ~0600 LT), and in summer (0000 - 0400 LT and 2100 LT to 2300 LT) during FE. This shows that losses due to chemical recombination exceed the plasma accumulation due to mechanical effects of the TNWC [45].