Diurnal, seasonal, and solar cycle effects of the variability (VR) of the critical frequency of sporadic E layer (foEsq) are investigated at Ibadan (7.4°N, 3.9°E, 6°S dip) in the African sector during high solar activity (HSA) year of 1958 (Rz = 181), moderate solar activity (MSA) year of 1973 (Rz = 30), and low solar activity (LSA) year of 1965 (Rz = 17). The diurnal variation of foEsq VR is characterized by post-midnight (32%–78%) and pre-midnight (20%–84%) peaks during high solar activity (HSA), the only epoch of the three showing these peaks and a diurnal trend. While the daytime foEsq VRs of the three epochs show no seasonal trend, pre-midnight and post-midnight, the foEsq VRs during HSA and LSA show seasonal trends. Similarity is observed in the curve of reciprocal of percentage occurrence of Esq and that of foEsq VR, indicating inverse variation of percentage occurrence and foEsq VR. Longitudinal influence is observed in the diurnal variation of HSA and MSA July foEsq VR of Ibadan (7.4°N, 3.9°E, 6°S dip) in the African sector, which is in the neighborhood of the Greenwich Meridian (GM); Singapore (1.3°N, 108.3°E, 17.6°S dip) in the Asian sector, east of GM; and Huancayo (12°S, 284.7°E, 1.90 dip) in the American sector, west of GM.
 The ionosphere is a unique part of the near-Earth environment. Above and below the ionosphere are phenomena which cause the state of the ionosphere to vary. Basu et al.  mentioned that extreme ionospheric variability portends a dangerous threat to transionospheric radio propagation and navigation systems. This is particularly true of sporadic E ionization, which is known to scatter HF radio signals [Ratcliffe, 1970; Cohen, 1967]. Bradley et al.  pointed out that fractional day-to-day VR in the lower ionosphere, i.e., below 200 km, is less investigated than that in the upper ionosphere (i.e., above 200km). The need for the variability or expected deviation from climatological means of ionospheric characteristics has been mentioned by Bilitza . Pulinets et al.  pointed out that the variability of the ionosphere, especially on a day-to-day basis, has not been sufficiently studied.
 This study aims at characterizing the variability of the critical frequency of equatorial sporadic E layer (foEsq) at Ibadan in the African sector on diurnal, seasonal, and solar cycle scales. The day-to-day VR of sporadic E in equatorial Africa will augment the International Reference Ionosphere model, which is the most commonly used model of the ionosphere. Ezquer et al.  pointed out that a model of the ionospheric VR would be useful for users of such model as well as for operators or satellite designers. Comparison of foEsq at Ibadan (7.4°N, 3.9°E, 6°S dip), Singapore (1.3°N, 108.3°E, 17.6°S dip), and Huancayo (12°S, 284.7°E, 1.9° dip) is also undertaken for investigation of the longitudinal influence of foEsq.
 Equatorial sporadic E (Esq) ionization is one of the three major types of sporadic E ionization phenomena mentioned in the literature. The other two are the type in the aurora and polar cap regions due to the precipitation of energetic particles and the type found in the midlatitude, the occurrence of which is explained by the wind shear theory of Whitehead . Esq is not a density layer, while the other types are. The sporadic E or Es phenomenon was discovered when Appleton and Naismith [1939, 1940] in UK and Berkner and Wells  in Huancayo found ionization in excess of that of the normal E layer. It is obviously a different phenomenon from the normal E layer ionization. Figure 1 illustrates the unusual high ionization of sporadic E compared to the regular E region ionization, though both regular E and Es layers exist at the same altitude [Liperovskaya et al., 2003].
Esq, the sporadic E phenomenon that occurs in the equatorial region, is a plasma irregularity layer, and the mechanism for its detection is backscatter instead of reflection. The plasma irregularity is due to irregularities in the electrojet equatorial (EEJ) current [Resende and Denardini, 2012], which cause E × B plasma gradient instability at the base of the E region, E being the electric field arising from the EEJ, and B being the horizontal magnetic field in the neighborhood of the equator. The EEJ is the eastward current flowing by day in a narrow belt flanking the dip equator.
 The data used for this study were obtained by the union Radio Mark II recorder-type ionosonde at Ibadan (7.4°N, 3.9°E, 6°S dip) and Singapore (1.3°N, 108.3°E, 17.6°S dip). The Huancayo (12°S, 284.7°E, 1.9° dip) data were obtained from the Space Interactive Data Resource website (http://spidr.ngdc.noaa.gov/).
 The ionosonde was developed at the Radio Research Station in Slough. Its transmitter and receiver are two separate subunits kept in tune by a frequency-sensitive servo system as the transmitter frequency is swept over the range 0.7–25 MHz in a sweep time of 5 min duration. The sounder sends out pulses at a repetition rate of 50 s−1 with a peak power of up to 1 kW. The photographic records of variation of virtual height with frequency, called ionograms, give the critical frequency values. The details of the ionosonde are given by Somoye [2009a]. Figure 2 shows a Mark II ionogram showing E and Esq traces, in which the former is indicated by the arrow labeled A and the latter by the arrow labeled B.
 The critical frequency of equatorial sporadic E layer, foEsq, during high solar activity (HSA) year (1958, Rz = 181), moderate solar activity (MSA) year (1973, Rz = 30), and low solar activity (LSA) year (1965, Rz = 17) are used for the Ibadan station. By definition, foEsq is attributed to the maximum frequency observed for the ordinary trace of the Es layer. However, since the ordinary and extraordinary traces are not easily distinguishable, a top frequency for Es is used as a proxy of the foEsq [Bibl et al., 1955; Rishbeth and Garriot, 1969]. The data are grouped into four seasons of March equinox, comprising February, March, and April; June solstice, comprising May, June, and July; September equinox, comprising August, September, and October; and December solstice, comprising November, December, and January. Following the examples of Rishbeth and Mendillo , the data were also grouped into different local time (LT) bins (00–05 h, 06–11 h, 12–17 h, and 18–23 h). We could not investigate the seasonal influence for Singapore and Huancayo due to unavailability of all-year-round data of the years considered.
 The relative variability (VR) is obtained by using the monthly mean, μ, at each hour and the standard deviation, σ. Following the example of Forbes et al.  and Bilitza et al. , we define VR as
 The extent of spread or deviation of each data from the calculated mean is usually given by VR. The VR given by equation (1) has an advantage over the one obtained from the quotient of interquartile range and median used by Kouris and Fotiadis  and Fotiadis et al. , though the latter is easier to interpret in terms of probability. While the VR obtained from equation (1) considers the whole data point, the latter method uses only 50% of the data [Bilitza et al., 2004; Somoye and Akala, 2010].
3 Result and Discussion
 Figure 3 shows the diurnal plots of foEsq VR separately for high, moderate, and low solar activity periods, respectively. The variations of foEsq VR with hours of the day is illustrated for the four seasons. In general, while foEsq VR is characterized by post-midnight and pre-midnight peaks during HSA due to onset and turnoff of solar ionization [Bilitza et al., 2004; Chou and Lee, 2008], no diurnal trend is found for the foEsq VR of both MSA and LSA. The diurnal trend of post-midnight and pre-midnight peaks has been observed for foF2 (critical frequency of F2 layer) VR, NmF2 (maximum electron density of F2 layer) VR, MUF (maximum useable frequency) VR, and h′F (virtual height of F layer) VR, though at the three epochs, by Bilitza et al. . Somoye [2009b]. Akala et al. , and Somoye et al. . Also, while post-midnight and pre-midnight peaks of the VR of these other ionospheric characteristics occur around 05 and 19 h, respectively, those of the foEsq VR during HSA occur around 06 h and 17 h, respectively.
 Generally, nighttime foEsq VR is found higher during HSA than during both MSA and LSA, being in the range 20%–78%, 15%–60%, and 5%–48% during the three epochs, respectively. Daytime foEsq VR is, however, slightly greater during both MSA and LSA (10%–40%) than during HSA (10%–30%) due to lower mean values during both MSA and LSA than during HSA, as shown in Figure 4, and higher absolute standard deviation during both MSA and LSA than during HSA, as illustrated in Figure 5. The higher daytime foEsq mean resulting in the lower foEsq VR during HSA than during MSA and LSA is most likely due to the greater intensity of EEJ during HSA than during other epochs [Onwumechilli, 1967]. On the other hand, the greater nighttime foEsq VR during MSA and LSA than during HSA is possibly due to the fact that Esq nighttime ionization is negatively correlated to the Zurich sunspot number, Rz [Chadwick, 1962]. This negative correlation may be as a result of nighttime reversal of EEJ, which causes the inhibition of instabilities responsible for Esq during the night [Rastogi, 1974]. The diurnal trend in the foEsq VR during HSA, which is absent during both MSA and LSA, is most likely due to the same reason for the greater daytime VR during both MSA and LSA than during HSA. It is interesting to note that the nighttime foEsq VR is greater than the foF2 VR during the three epochs. The diurnal variations of both the foEsq VR and the foF2 VR during the three epochs are shown in Figure 6. That the foEsq VR is greater than the foF2 VR is most likely due to the irregular nature of sporadic E ionization. Somoye [2009b] pointed out that the high nighttime relative variability of an ionospheric characteristic is not only dependent on its low mean value at night alone but also on how irregularly the ionization occurs. Also worthy of note is the similarity in the diurnal variation of foEsq VR and that of the reciprocal of percentage occurrence of Esq ionization, as shown in Figure 7, indicating that foEsq VR varies inversely as percentage occurrence of Esq ionization. Esq ionization occurrence is low at night [Oyinloye, 1968; Awe, 1971] due to decay of ionization at that time of the day. An average of 25% nighttime Esq occurrence during the three epochs is observed. Awe  mentioned that nighttime Esq percentage occurrence shows no clear dependence on solar activity. Daytime Esq occurrences of ~75% during HSA and ~45% during both MSA and LSA are observed. This decay of ionization is one of the factors responsible for high nighttime relative variability, the other factor being low mean values at night [Somoye, 2009b].
 In Figure 8, which shows the foEsq VR of 00–05 LT, 06–11 LT, 12–17 LT, and 18–23LT with the months of 1958, 1973, and 1965, respectively, the post-midnight foEsq VR is maximum during March equinox and December solstice, while the pre-midnight foEsq VR is maximum during March equinox and June solstice of HSA. Daytime foEsq VR shows no seasonal trend. No LT bin in MSA foEsq VR shows any clear seasonal trend. During LSA, the seasonal trend in foEsq VR is obvious only in the second half of the day, with peak values in March equinox and December solstice. The present results, under seasonal consideration, are in fair agreement with those of Akala et al. , who reported the highest variability of foF2 VR during March equinox and June solstice, and those of Bilitza et al. , Akala et al. , and Somoye [2009b], who found higher nighttime VR of NmF2/foF2 during June solstice and September equinox. On a general note, the seasonal trend of VR of the ionosphere is very likely dependent on the ionospheric characteristic and on the latitude and longitude of the station of observation.
 Figure 9 shows the diurnal variations of foEsq VR of Ibadan in the African sector, of Singapore in the Asian sector, and of Huancayo in the American sector during 1968 (HSA) and 1971 (MSA). The longitudinal influence on foEsq VR is discernible in that while at Huancayo, the HSA foEsq VR is much higher than that of MSA during the first half of the day, the foEsq VR of Ibadan during HSA is slightly higher than the MSA foEsq VR throughout the day. At Singapore, the HSA foEsq VR and the MSA foEsq VR show no clear distinction. Also, foEsq VR is greatest at Huancayo, west of the Greenwich Meridian (GM), followed by that of Singapore in the east of the Greenwich Meridian, and least at Ibadan, in the neighborhood of the Greenwich Meridian. We can explain this with the more intense E region current (and, therefore, the horizontal magnetic intensity, H) in the American sector, i.e., west of GM than in the Asian sector (east of GM) and the African sector (in the neighborhood of GM) [Onwumechilli and Ogbuehi, 1967]. The present result is in good agreement with that of Akala et al. , who reported the greatest foF2 VR for Huancayo—an equatorial station west of the Greenwich Meridian, followed by that of Vanimo—a low latitude station in the east of the Greenwich Meridian. Ouagadougou, an equatorial station in the neighborhood of the Greenwich Meridian is reported to have the least foF2 by these authors. The longitudinal effect is discernible not only in the foEsq VR values of the three stations but also in the manner in which the foEsq VR of HSA differ from that of MSA. While the difference in the foEsq VR of HSA and MSA is not pronounced at Singapore, the foEsq VR of the former is slightly greater than that of the latter at Ibadan. At Huancayo, the HSA foEsq VR is much greater than that of MSA in the first half of the day. Also, significant dependence on longitude is revealed in the behavior of day-to-day MUF VR, as reported by Fotiadis et al. .
 The characterization of the variability of sporadic E ionization at the equatorial station of Ibadan on diurnal, seasonal, and solar cycle scales has been undertaken. Also investigated is the diurnal influence of longitude on equatorial sporadic E ionization variability using data from Ibadan, Singapore, and Huancayo. We have the following observations.
foEsq VR is characterized by post-midnight (occurring at 06 h) and pre-midnight (occurring at 17 h) peaks during HSA, the only epoch showing a diurnal trend of foEsq VR.
 While nighttime foEsq VR is greater during HSA (20%–78%) than during MSA and LSA (15%–60%), daytime MSA and LSA foEsq VR (10%–40%) are greater than that of HSA (10%–30%).
 The daytime foEsq VRs of the three epochs show no seasonal trend, while post-midnight and pre-midnight foEsq VR of HSA and that of LSA during the second half of the day show seasonal trends.
 Similarity is observed in the plot of reciprocal of percentage (%) occurrence of equatorial sporadic E ionization and that of foEsq VR, indicating that percentage (%) occurrence varies inversely as foEsq VR.
 The foEsq VR at Huancayo (west of the Greenwich Meridian) is the greatest, followed by that of Singapore (east of GM), while that of Ibadan in the neighborhood of the GM is the least.
 The authors sincerely thank Adeniyi Jacobs for his useful suggestions.