Radio Science

F2 region maximum electron density height predictions for South American latitudes

Authors

  • R. G. Ezquer,

    1. Facultad Regional Tucumán, Universidad Tecnológica Nacional, Tucumán, Argentina
    2. Laboratorio de Ionósfera, Instituto de Física, Universidad Nacional de Tucumán, Tucumán, Argentina
    3. Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Buenos Aires, Argentina
    Search for more papers by this author
  • L. Scidá,

    1. Laboratorio de Ionósfera, Instituto de Física, Universidad Nacional de Tucumán, Tucumán, Argentina
    Search for more papers by this author
  • G. A. Mansilla,

    1. Facultad Regional Tucumán, Universidad Tecnológica Nacional, Tucumán, Argentina
    2. Laboratorio de Ionósfera, Instituto de Física, Universidad Nacional de Tucumán, Tucumán, Argentina
    3. Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Buenos Aires, Argentina
    Search for more papers by this author
  • M. Mosert,

    1. Facultad Regional Tucumán, Universidad Tecnológica Nacional, Tucumán, Argentina
    2. Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Buenos Aires, Argentina
    3. Complejo Astronómico El Leoncito (CASLEO), San Juan, Argentina
    Search for more papers by this author
  • M. F. Herrera

    1. Laboratorio de Ionósfera, Instituto de Física, Universidad Nacional de Tucumán, Tucumán, Argentina
    Search for more papers by this author

Abstract

[1] Values of the F2 region maximum electron density height (hmF2) calculated using ground ionosonde data at South American latitudes are used to check the validity of the International Reference Ionosphere (IRI) to predict this variable. With this in mind we compare hmF2 predictions given by the model when measurements of critical frequency of F2 region and propagation parameter M(3000)F2 were used as input parameter in IRI (hmF2IRI-Exp), against those obtained using the standard International Radio Consultative Committee (CCIR) option (hmF2IRI-CCIR). In this paper we used hmF2IRI-Exp values because hmF2 measurements were not available for the considered cases. Moreover, a comparison of the measured M(3000)F2 values with the CCIR predictions have been done. The results show that, in general, the standard predictions follow the diurnal tendency observed in the hmF2IRI-Exp values. At low latitudes the hmF2IRI-Exp values show oscillations not reproduced by the standard option. Cases with disagreements for 24 hours have been observed at high latitudes. Other cases with good agreement have been also obtained. The results suggest that, in general, the standard option of the model gives good hmF2 predictions at South American latitudes. Few cases showed deviation between 15 and 25%. As we expected, the obtained results suggest that the deviation between predicted and measured M(3000)F2 values is the main contribution for the deviation between hmF2IRI-CCIR and hmF2IRI-Exp. The comparison with the results obtained in previous work shows that the IRI performance in predicting M(3000)F2 and hmF2 is better than in predicting foF2.

1. Introduction

[2] For successful radio communication, it is essential to predict the behavior of the ionospheric region that will affect a given radio communication circuit. Such a prediction will identify the time periods, the path regions and the sections of high frequency bands that will allow or disrupt the use of the selected high frequency communication circuit. This need for prediction leads to modeling of the ionosphere. Several physical, empirical and semiempirical models [e.g., Anderson, 1973; Barghausen et al., 1969; Bent et al., 1976; Llewellyn and Bent, 1973; Bilitza, 1990; Anderson et al., 1987] have been developed to predict ionospheric variables.

[3] In a previous work, Ezquer et al. [1996] used measurements of the critical frequencies of the ionospheric regions (foE, foF1 and foF2) obtained at South American stations for different solar conditions and seasons to check the validity of the International Reference Ionosphere (IRI) model [Bilitza, 1990] to predict these frequencies. They found good predictions for foE and foF1 when compared. The degree of accuracy among experimental and predicted foF2 values was lower than those observed for the other frequencies, which is well known and is due to higher variability in the F2 region, and cases with strong disagreements were observed by Ezquer et al. [1996].

[4] In order to complete the study of Ezquer et al. [1996], in the present paper values of the F2 region maximum electron density height (hmF2) calculated using data of foF2 and M(3000)F2 obtained at South American latitudes are used to check the validity of the International Reference Ionosphere to predict this variable. Taking into account that for this study hmF2 measurements were not available, we compare hmF2 predictions given by the model when measurements of critical frequency of F2 region and propagation parameter M(3000)F2 were used as input parameter in IRI (hmF2IRI-Exp), against those obtained using the standard CCIR option (hmF2IRI-CCIR). In the IRI model foE measurements cannot be used as an input coefficient. Only cases with 24 hours measurements were considered.

2. IRI Model

[5] The Committee on Space Research (COSPAR) and the International Union of Radio Science (URSI) established an international task group to develop and improve a standard model of the ionospheric plasma parameter. This model is the International Reference Ionosphere (IRI) [Rawer et al., 1981; Bilitza, 1986; Rawer and Bilitza, 1989, 1990; Bilitza, 1990]. COSPAR is interested in a general description of the ionosphere as part of the terrestrial environment for the evaluation of environmental effects on spacecraft and experiments in space. The main interest of URSI is the electron density part of IRI for defining the background ionosphere for radiowave propagation studies and applications.

[6] IRI is one of the most widely used empirical models, and has undergone several years of critical checking and improving by the international science community. The emphasis of IRI is to summarize a large collection of ground-based and spacecraft data to provide true height profiles of the ionosphere. IRI gives the altitude dependence of electron density, electron and ion temperatures and the composition of positive iones. For the worldwide description of the peak electron density, the International Radio Consultative Committee (CCIR) [1967a, 1967b] numerical maps are used as a choice. In this work we are interested in the F2 peak height.

[7] In the IRI model, hmF2 is obtained by its close correlation with the propagation parameter M(3000)F2 [Shimazaki, 1955; Bradley and Dudeney, 1973; Bilitza et al., 1979]. M(3000)F2 is defined as

equation image

where MUF is the maximum usable frequency that, refracted in the ionosphere, can be received at a distance of 3000 km. This factor has been routinely scaled from ionograms, and numerical maps [CCIR, 1967a, 1967b] are used has a choice in the model. The F2 peak height is calculated from M(3000)F2 with the empirical formula [Bilitza et al., 1979]

equation image

with the correction factor

equation image

and the solar activity functions

equation image
equation image
equation image
equation image

R12 is the 12-month-running mean of solar sunspot number, and Ψ is the magnetic dip latitude

equation image

which is related to the magnetic inclination (short: dip) ψ of the Earth's magnetic field at 300 km altitude.

[8] In this paper we calculate hmF2 using the CCIR options in IRI and also using ground ionosonde measurements as input parameters in the model, from now on: hmF2IRI-CCIR and hmF2IRI-Exp values, respectively. For this purpose, the internet online version of IRI (IRI 95), (http://nssdc.gsfc.nasa.gov/space/model/models/iri.html), has been used. We assume that hmF2IRI-Exp is a more realistic value of the maximum electron density height. The use of the URSI model is beyond the scope of the paper.

3. Data

[9] Hourly monthly median values of M(3000)F2 and foF2 measured at the stations listed in Table 1 were used to calculate hmF2. We consider equinoxes and solstices for years of low (1965, 1977, 1985) and high (1958, 1969, 1980) solar activities. Only typical results are shown in this paper.

Table 1. Considered Stations
 Geodetic Coordinates
Huancayo(−12.05, 284.67)
Tucumán(−26.90, 294.60)
Bs. Aires(−34.55, 301.30)
Port Stanley(−51.70, 302.20)
Ushuaia(−54.80, 291.70)
Islas Argentinas(−65.20, 295.70)

[10] It is well know that the ionosphere produces several effects on transionospheric radio waves. These effects are proportional to the number of free electrons encountered by the wave on its passage through the ionosphere (total electron content, TEC). The highest TEC values in the world occur in the near-equatorial region. This region extends approximately 20° either side of the magnetic equator, with the highest value not at the equator, but rather at the so-called “equatorial anomaly (EA) peaks region” at approximately ±15° from the magnetic equator. Tucumán is placed near the southern peak of the EA. Ezquer et al. [1995, 1998] showed that the IRI model underestimated TEC measured above Tucumán. Their results suggest that the ionization contribution from the equator, which causes the EA, affects the ionosphere over Tucumán producing an electron density profile broader than those assumed by the model.

4. Results and Discussion

[11] Figure 1 shows the results for equinox and low solar activity. It can be seen that the IRI-CCIR predictions follow the tendency of the curve obtained with measured M(3000)F2 factor (M(3000)F2Exp), at all stations. At low latitudes, hmF2IRI-CCIR is greater than hmF2IRI-Exp during daytime conditions and does not reproduce the oscillations shown by hmF2IRI-Exp. The best agreement is observed at Ushuaia daytime hours. For Islas Argentinas, hmF2IRI-CCIR is greater than hmF2IRI-Exp for all hours of the day. The disagreement observed at Tucumán could be produce by the influence of the EA on the Tucumán's ionosphere which is not well reproduced by the model; and that observed at Islas Argentinas could be produced because Islas Argentinas latitude is close to the validity model boundary [Bilitza, 1990].

Figure 1.

The hmF2IRI-Exp (solid line) and hmF2IRI-CCIR (circles) values for Huancayo (HN), Tucumán (TU), Buenos Aires (BA), Port Stanley (PS), Ushuaia (UH) and Islas Argentinas (IA). Equinox, low solar activity.

[12] The results for equinox, high solar activity, are shown in Figure 2. At Tucumán, hmF2IRI-CCIR exceeds hmF2IRI-Exp from 0 LT to 17 LT. This situation is observed for almost all the day at Ushuaia. The best agreement between both curves is obtained for Port Stanley during daytime conditions.

Figure 2.

The hmF2IRI-Exp (solid line) and hmF2IRI-CCIR (circles) values for Tucumán (TU), Port Stanley (PS), Ushuaia (UH) and Islas Argentinas (IA). Equinox, high solar activity.

[13] Figure 3 shows the results for solstices considering different solar conditions. The best agreement between both curves is observed for Port Stanley.

Figure 3.

The hmF2IRI-Exp (solid line) and hmF2IRI-CCIR (circles) values for Tucumán (TU), Buenos Aires (BA), Port Stanley (PS) and Islas Argentinas (IA). Solstices.

[14] Taking into account that the calculated hmF2 depends on M(3000)F2 factor, as is shown in equation (2), a comparison among measured and predicted M(3000)F2 has been done. Figures 4, 5 and 6 show the results for the cases considered in Figures 1, 2, and 3, respectively. An opposite behavior to that observed for hmF2 can be seen. In these figures we included the absolute value of the deviation between prediction and measurement in percentage of measurement, calculated as:

equation image

It can be seen that for few cases AD is greater than 10%.

Figure 4.

The M3000F2Exp (solid line) and M3000F2CCIR (circles) values for Huancayo (HN), Tucumán (TU), Buenos Aires (BA), Port Stanley (PS), Ushuaia (UH) and Islas Argentinas (IA). Equinox, low solar activity. Absolute deviations in percent of M3000F2Exp (squares).

Figure 5.

The M3000F2Exp (solid line) and M3000F2CCIR (circles) values for Tucumán (TU), Port Stanley (PS), Ushuaia (UH) and Islas Argentinas (IA). Equinox, high solar activity. Absolute deviations in percent of M3000F2Exp (squares).

Figure 6.

The M3000F2Exp (solid line) and M3000F2CCIR (circles) values for Tucumán (TU), Buenos Aires (BA), Port Stanley (PS) and Islas Argentinas (IA). Solstices. Absolute deviations in percent of M3000F2Exp (squares).

[15] For the used ionosondes, the error when foF2 is measured is about 0.1 MHz, and the average error for the M(3000)F2 factor measurements is about 2%, which is lower than the AD observed in the considered cases.

[16] In order to check the incidence of M(3000)F2 disagreements on the deviation between hmF2IRI-CCIR and hmF2IRI-Exp, we calculated the following deviations:

equation image
equation image

Figure 7 presents cases with hmF2 deviations lower than 15%. Both curves show an almost symmetrical behavior suggesting that, as expected, the main contribution to the hmF2 deviation is that observed for M(3000)F2 factor.

Figure 7.

Deviations between hmF2IRI-Exp and hmF2IRI-CCIR, in percent of hmF2IRI-Exp (solid line), for Tucumán (TU), Buenos Aires (BA), Port Stanley (PS) and Ushuaia (UH). Cases with hmF2 deviations lower than 15%. M(3000)F2 deviations in percent of M(3000)F2Exp (circles).

[17] Figure 8 shows the highest observed deviations, which correspond to high latitude. hmF2 and M(3000)F2 deviations show similar behavior to that observed at Figure 7. Moreover, it can be seen that the obtained values of hmF2 deviations are not greater than 30%.

Figure 8.

Deviations between hmF2IRI-Exp and hmF2IRI-CCIR, in percent of hmF2IRI-Exp (solid line), for Ushuaia (UH) and Islas Argentinas (IA). Cases with hmF2 deviations lower than 30%. M(3000)F2 deviations in percent of M(3000)F2Exp (circles).

[18] Ezquer et al. [1996] showed that for South American stations, the maximum deviation of the predicted foF2 values from the measurements could reach values as high as 50% or more. Those results suggested that it would be possible to plan a HF circuit with a predicted frequency, which is greater than the maximum frequency that the circuit can support. Assuming that hmF2IRI-Exp is a more realistic value of the F2 region maximum electron density height, we can say that the deviation values obtained in the present paper, in general, are lower than those obtained previously for foF2. These results suggest that the model performance in predicting M(3000)F2 factor and hmF2 is better than in predicting foF2 for South American stations.

5. Conclusions

[19] In order to complete a previous work on ionospheric predictions for South American latitudes, in the present paper a study to check the validity of IRI model to predict hmF2 has been done. The results show that, in general, the standard predictions follow the diurnal tendency observed in hmF2IRI-Exp values. At high latitudes cases with disagreement for 24 hours have been observed. However, other cases with good agreement have also been obtained.

[20] The deviation between hmF2IRI-CCIR and hmF2IRI-Exp, in general, are lower than 15%. Few cases, corresponding to high latitudes, showed disagreements between 15 and 25%. An important contribution to hmF2 deviation is the deviation between M(3000)F2CCIR and M(3000)F2 Exp. The comparison with the results obtained by Ezquer et al. [1996] shows that the IRI performance in predicting M(3000)F2 and hmF2 is better than in predicting foF2.

Ancillary