SEARCH

SEARCH BY CITATION

Keywords:

  • equatorial spread F

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. Acknowledgments
  7. References

[1] Interpretation of the morphology of equatorial spread F (ESF) is usually given in terms of factors that modulate the growth rate of the Rayleigh-Taylor instability. That interpretation is, however, incomplete because only one means for growth-rate enhancement has been included; that is, the strength of the post-sunset rise of the F layer could become enhanced, when the solar terminator aligns with geomagnetic field lines. Maxima in ESF occurrence observed near equinoxes seem accountable by this means, but maxima observed near solstices require another source of enhancement. Seasonal migration and latitudinal alignment of the inter-tropical convergence zone (ITCZ) with the magnetic dip equator is suggested as the missing source, and shown able to fill that void.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. Acknowledgments
  7. References

[2] The occurrence frequency of equatorial spread F (ESF), which refers to plasma structure in the nighttime equatorial F layer, is known to have a complex morphology, one with both seasonal and longitudinal variations. Statistical description of that morphology, in terms of a probability that ESF will occur (PESF), can be cast as a product of two conditional probabilities [e.g., McClure et al., 1998], PESF = PS · PI, where PS and PI are probabilities that a seed perturbation of sufficient amplitude is present, and that the growth rate of the Rayleigh-Taylor (RT) instability (γRT) is large enough, respectively. A currently popular view is that the seasonal and longitudinal behavior of ESF occurrence can be described in terms of PI, assuming PS is always large enough [Maruyama and Matuura, 1984; Tsunoda, 1985].

[3] A basic difficulty with this interpretation is that it contains only one means for increasing PESF. The hypothesis for that means is that γRT becomes enhanced when the solar terminator aligns with geomagnetic field (equation image) lines [Tsunoda, 1985] and PESF becomes enhanced via PI. The reasoning is that (1) alignment produces the steepest longitudinal gradient in flux-tube-integrated Pedersen conductivity, (2) strength of the post-sunset rise (PSSR) of the F layer is proportional to this gradient, and, (3) the largest PESF should then accompany the strongest PSSR. Geometric alignment occurs at the equinoxes, in locations where magnetic declination angle is zero. The day of year of that alignment shifts away from equinoxes and toward the June (December) solstice, in locations where declination is eastward (westward). This hypothesis, which we will refer to as the solar terminator-equation image-alignment (STBA) hypothesis, seems to work in some longitudes, but it does not explain the high occurrence frequency of ESF over Peru during December solstice, but not June solstice [e.g., Rastogi, 1980; Basu and Basu, 1985]. Another source seems necessary to explain these asymmetrical distributions during the solstices. The other contributor to PI [Maruyama and Matuura, 1984] acts to reduce, not increase, γRT.

[4] A less popular view is one in which the modulation of PS is assumed responsible for the observed ESF morphology because γRT (hence, PI) is not large enough [Röttger, 1977, 1981; McClure et al., 1998]. In this scenario, gravity waves (GWs) are assumed to be generated by mesoscale convective systems (MCSs) in the troposphere, and to propagate up to the F region, where they induce seed perturbations in plasma density (N). These MCSs are usually found within the inter-tropical convergence zone (ITCZ), and ESF morphology is assumed to be controlled by the seasonal migration of the ITCZ in latitude. This view, however, has not attracted much of a following, perhaps because the evidence of ITCZ migration, as presented thus far to the ionospheric community, may have been less than convincing.

[5] In this paper, we address the question, what is the cause of ESF during solstices? Toward this end, we present the first tangible link between lower atmospheric disturbances and ESF. We show that occurrences of ESF during the solstices are likely produced by enhanced seeding (hence, enhanced PS). The findings, presented below, support the following hypothesis, that perturbations in the neutral gas, produced by GWs, elicit a response in plasma density (N) only when there is alignment in the phase fronts of GWs with equation image lines [e.g., Huang and Kelley, 1996]. We will refer to this as the gravity wave-equation image-alignment (GWBA) hypothesis. We show, for the first time, that GWBA occurs most often during solstice, when the ITCZ is collocated with the magnetic dip equator. When the STBA and GWBA hypotheses are combined, we arrive at a more complete description of ESF morphology, which seems to account for its major features.

2. Results

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. Acknowledgments
  7. References

2.1. ESF Morphology

[6] To demonstrate seed control of ESF during solstices, we use PESF from McClure et al. [1998], which was plotted as a function of longitude for the four seasons. Their data are from January 1978 to September 1980, a period of high solar activity. We have redrawn their plots and present them superposed in Figure 1a. (Order of presentation is from bottom to top, because available space had to be used to accommodate all of the plots.) Each curve represents behavior averaged over three months; letters indicate months of year. If STBA is important, curves for equinoxes (FMA, ASO) should be higher than curves for solstices (MJJ, NDJ), except where the magnetic declination is large. The most noticeable anomalies can be found in the South American and African sectors. The extremely large values of PESF for the NDJ curve in the South American sector illustrate the discrepancy mentioned by Basu and Basu [1985]. In the African sector, the large maxima during equinoxes seem reasonable, but asymmetric behavior during solstices cannot be explained.

image

Figure 1. (a) The curves for ESF probability, from McClure et al. [1998], for different month-sets of the year, are superposed to allow comparison of seasonal behavior at different longitudes. (b) A curve showing the sum of the curves in Figure 1a. (c) Map of the World showing locations of interest to this proposal, indicated by black circles, as well as other stations located near the magnetic dip equator, indicated by squares. A possible ground track of the C/NOFS satellite is also shown. (d) Plots showing the latitudinal migration of the ITCZ as a function of month of year, for four different locations [from Waliser and Gautier, 1993].

Download figure to PowerPoint

[7] Anomalies are also present in annual behavior, which we have depicted by simply adding the probabilities of each curve, as shown in Figure 1b. This sum curve has two maxima, a shoulder, and two minima. The larger maximum is situated over Africa, while the smaller maximum is situated over Micronesia, in the western Pacific region. The shoulder brackets South America and the eastern Pacific region. The probability of ESF occurrence during any season over Africa is comparable to that over South America during the December solstice. This implies that the large sum is a consequence of an almost perpetual presence of ESF over Africa. In contrast, the deepest minimum, which brackets Vietnam and the Philippines, appears to be produced by a virtual absence of ESF in that sector throughout the year. The other minimum, located to the east of Christmas Island, in the eastern Pacific region, is not as deep because there is some enhancement in ESF activity during the Fall equinox.

2.2. ITCZ Migration

[8] For information about the migration of the ITCZ with season, we use Waliser and Gautier [1993]. The results consist of occurrences per month of highly reflective clouds, as detected at visible and infrared wavelengths from satellites. The occurrences were binned as a function of geographic latitude and month of year, for different longitude sectors. Five (out of seven) of their plots, for longitude sectors of interest, are reproduced in Figure 1d; a sixth is presented as an inset, because of space constraint. The white curves indicate the mean locations of the ITCZ.

[9] We see that the migration of the mean ITCZ, averaged over 17 years, resembles a sinusoid. While the ITCZ migration in latitude is generally toward the summer hemisphere, the manner in which it does so, with month of year, clearly differs with longitude. Differences are seen both in modulation amplitude and in the northward displacement in the mean latitude of the sinusoid. Most important is the finding that the mean latitude of the ITCZ is situated near the magnetic dip equator during solstices. This relationship can be seen by comparing the white curves with the locations of the stations that are marked and labeled in Figure 1c; all are located or near the dip equator.

2.3. Maxima During Solstices

[10] We first show that behavior of PESF during the solstices agrees qualitatively with the GWBA hypothesis. That is, in the African sector, PESF is high (low) during MJJ (NDJ), and the mean ITCZ location in Figure 1d is closest to (farthest from) the dip equator during MJJ (NDJ). The same relationship is found in the Micronesian sector, where PESF is highest (lowest) during MJJ (NDJ), and the mean ITCZ location, for the West Pacific region, is closest to (farthest from) the dip equator during MJJ (NDJ). In the South American sector, the relationship is reversed, as expected, because the dip equator is in the southern hemisphere. That is, PESF is highest (lowest) during NDJ (MJJ), and the mean ITCZ is closest to (farthest from) the dip equator during NDJ (MJJ). The results appear convincing. In two of the three sectors, where dip equator is in the northern hemisphere, maxima in PESF occur during MJJ. In the third sector, where the dip equator is in the southern hemisphere, the maximum in PESF occurs during NDJ. Behavior in all three sectors can be explained in terms of the GWBA hypothesis. The reduction in PESF during the other solstice, in all three longitude sectors, is also consistent with this hypothesis.

2.4. Importance of Seeding

[11] Given that seeding has previously not been considered seriously as a source for ESF, it is interesting to note that seed amplitude in N during solstice may be as important as the strength of the PSSR (hence, γRT) during equinoxes. The finding that PESF in the African sector is as high during MJJ as PESF is during ASO would seem to support this premise. (In this sector, we would associate PESF during ASO and FMA with STBA.) Some additional support for this notion comes from a theoretical study, which indicated that the saturation amplitude associated with the RT instability would be much larger for cases in which the initial seed amplitude is large [Huang et al., 1993].

2.5. Persistent Minimum in East Pacific

[12] Behavior in this sector, unexpected at first glance, turns out to also be consistent with the GWBA hypothesis. That is, PESF should be low here during both MJJ and NDJ, because the mean ITCZ location remains in the northern hemisphere throughout the year, and the dip equator is in the southern hemisphere, at longitudes to the east of Christmas Island (CI in Figure 1c). The slight enhancement (depression) of ESF activity during ASO (FMA) can be attributed to a modulation in PI. by the STBA effect for a region of large-eastward magnetic declination [Tsunoda, 1985].

2.6. Exceptions

[13] While the GWBA hypothesis appears promising, two features in Figure 1 remain to be explained. The first is a dramatic transition in PESF behavior in the West African (Atlantic) sector, from that in the African sector to that in the South American sector. The maximum during NDJ is consistent with the fact that the mean ITCZ remains in the northern hemisphere, where the dip equator is located. While this enhancement can be explained in terms of GWBA, we cannot account for the minimum in PESF during MJJ, because GWBA should still be effective. One possibility, perhaps, is that PESF is reduced during MJJ because of a meridional neutral wind [Maruyama and Matuura, 1984].

[14] The most puzzling behavior is the apparent existence of a ‘permanent’ minimum in PESF in the Vietnam-Philippines (V-P) sector. The migration of the mean ITCZ and the location of the dip equator are essentially the same as those in Micronesia. The finding that PESF is low during equinoxes suggests that the PSSR is weak; the finding that PESF is low during solstices suggests that seeding must not be enhanced. Why the PSSR should be weak is not addressed here, but the conclusion that seed strength remains weak is surprising because MCS activity in the Indian and West Pacific sectors is very high and widespread in latitude, according to Figure 1. We can also discard the possibility that PESF results from AE-E are not be representative of ESF activity in that sector, because ionosonde observations from Vietnam also indicate that ESF activity is very low throughout the year [Hoang et al., 2010].

3. Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. Acknowledgments
  7. References

[15] We have presented the first tangible link between enhanced PESF during solstices and seeding of N perturbations by GWs. The evidence is in the form of enhancements in PESF whenever the mean ITCZ location becomes aligned with the dip equator. This behavior can be understood in terms of a transfer of neutral perturbations to the plasma, whenever the phase fronts of GWs are aligned with equation image. Alignment is required because excess charge of one sign gathers along a phase front of a GW, while excess charge of opposite sign gathers along a phase front that is half a wavelength away. If phase fronts are not aligned with equation image, the net excess charge would be reduced from electron motion between phase fronts, along equation image. Accordingly, transport of the plasma by the polarization electric field, which accompanies the excess charge, would also be reduced [e.g., Huang and Kelley, 1996]. It appears that the magnetic aspect angle (i.e., angle between phase fronts and equation image) would have to be less than equation image, where σP and σ0 are the Pedersen and direct conductivities, respectively. At the base of the F layer (say 300 km), equation image would be less than 10−4, which means the magnetic aspect angle would have to be less than 0.007°.

[16] Assuming GWs are launched isotropically (i.e., all azimuths) from a confined MCS region [e.g., Vadas, 2007], GWBA would most often be satisfied, when the mean ITCZ is near the magnetic dip equator. The implication is that only an extremely small subset of GWs that reach the thermosphere is responsible for the seeding of N perturbations. For example, meridional GWs that launch traveling ionospheric disturbances (TIDs) [e.g., Sterling et al., 1971] would not be involved in the seeding of ESF by this process.

[17] With the above discussion in hand, we can think of several explanations for the permanent minimum in the V-P sector. One, in particular, is suggested by the results reported by Maeda and Badillo [1966]. They found a tendency for ESF to be reduced during the approach of a tropical disturbance. (Their observations were made with an ionosonde and microbarographs in the Philippines.) This finding may be pointing toward a need to reconsider the polarization shorting issue in terms of a more realistic geometry. After all, if GWBA must be as precise as indicated above, we should account for the fact that GW phase fronts are spherical than plane and equation image lines are dipolar. This means that the radii of curvature of GW phase fronts would be smaller, the closer they are to the source region. Hence, GWBA with a geometry that changes with distance from the source could explain the results of Maeda and Badillo [1966], and zonal proximity to GW source regions may be a cause of the minimum in the V-P sector.

[18] Other processes could also be acting to produce the minimum in PESF in the V-P sector. For example, Chen [1993] showed that the quasi-biennial oscillation (QBO) of the zonal wind in the stratosphere seems to affect the occurrence frequency of ESF. In the Indo-East African sector, ESF occurrence decreased (increased) during the westerly (easterly) phase of the QBO. An opposite dependence was found for the Peruvian sector. Although this finding does not explain the perpetual minimum, it does point to the possibility that background winds may be a factor. For example, non-migrating tides could be affecting electrodynamical behavior with peculiar results appearing at specific longitudes [e.g., England et al., 2006].

[19] To summarize, all of the major features in ESF morphology seem attributable to two effects, STBA and GWBA, which appear to operate during equinox and solstice, respectively. Questions that remain, regarding the cause of the perpetual PESF minimum in the V-P sector, and a possible dependence of ESF activity on the distance from the source of GWs, point to a need to consider the conditions that affect GW propagation, as well as phenomena such as non-migrating tides, which could produce unusual longitudinal behavior. In closing, we note that large-scale wave structure (LSWS), which develops in the bottomside F layer prior to the occurrence of ESF [Tsunoda, 2005, 2008, 2009], is being measured routinely, for the first time, with radio beacon signals from an equatorial-orbiting satellite [e.g., Thampi et al., 2009]. We believe that LSWS is the plasma perturbation that results from GWBA and STBA. Monitoring of its characteristics should shed considerable light on outstanding questions.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. Acknowledgments
  7. References

[20] This research was supported by the National Science Foundation under grant ATM-0720396, and by the Air Force Office of Scientific Research under award FA9550-10-C-0004.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results
  5. 3. Discussion
  6. Acknowledgments
  7. References