SEARCH

SEARCH BY CITATION

Keywords:

  • GPS;
  • ionosphere;
  • irregularities;
  • radar;
  • scintillation;
  • spread F

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental Setup
  5. 3. Observations
  6. 4. Results and Discussion
  7. 5. Summary and Final Remarks
  8. Acknowledgments
  9. References
  10. Supporting Information

[1] We present new results of a study of the interferometric coherent backscatter radar imaging technique applied to São Luís observations made on the night of 1 December 2005. The range-time-intensity (RTI) map of the observations shows echoes occurring near theF region peak and topside heights followed by echoing layers confined to bottomside F region heights. Analyses of the measurements made on this night allowed us to investigate the ability of the São Luís radar interferometer to provide information about the morphology of the scattering structures responsible for different types of equatorial spread Fechoing layers. Results show that topside echoes were produced by a vertically elongated, horizontally narrow scattering channel of irregularities associated with a large-scale plasma depletion (“bubble”) as evidenced by colocated GPS scintillation measurements. Bottomside echoes were caused by structured, eastward drifting scattering regions with limited vertical development. Bottom-type echoes, on the other hand, were detected at heights below the minimum altitude of the bottomside echoes and were caused by an undifferentiated scattering region. Our imaging results are discussed in light of current equatorial spreadFtheories and previous higher-resolution imaging observations made at the Jicamarca Radio Observatory.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental Setup
  5. 3. Observations
  6. 4. Results and Discussion
  7. 5. Summary and Final Remarks
  8. Acknowledgments
  9. References
  10. Supporting Information

[2] The manifestation of electron density irregularities in the equatorial ionosphere is commonly referred to as equatorial spread F (ESF) for historical reasons. First signatures of equatorial ionospheric irregularities were seen as frequency and range spread at F region heights in the ionograms measured by the early ionosondes [Booker and Wells, 1938]. ESF is now recognized as a threat to satellite-based communication and navigation systems [e.g.,Basu et al., 1988; Kintner et al., 2007]. Small-scale (hundreds of meters) ESF irregularities cause ionospheric scintillation on transionospheric radio signals [e.g.,Yeh and Liu, 1982]. Large-scale (several tens of km) ESF plasma depletions, on the other hand, cause abnormally large horizontal gradients in ionospheric delay that are difficult to model and predict [e.g.,Skone and Shrestha, 2002].

[3] Much of what is known today about the origin and morphology of ESF comes from radar observations. Ionospheric radar observations have been contributed significantly to both basic and applied studies of ESF. For instance, observations made by the ALTAIR radar in the Kwajalein Atoll have helped us better understand the conditions leading to ESF development and morphology [Tsunoda, 1981; Tsunoda and White, 1981; Hysell et al., 2006]. Additionally, recent studies have attempted simultaneous incoherent and coherent backscatter radar observations to better understand and improve the description of the ionospheric channel during ESF events [Caton et al., 2009; Costa et al., 2011].

[4] In particular, the radar observations at the Jicamarca Radio Observatory in Peru have greatly helped us to better understand ESF. Strong coherent echoes were detected during the initial incoherent backscatter observations made at Jicamarca. The early Jicamarca observations of ESF were described in a comprehensive study by Woodman and LaHoz [1976]. At that time, however, only sporadic (or campaign) observations of ESF could be made at Jicamarca. The development of the JULIA (Jicamarca Unattended Long-term Studies of the Ionosphere and Atmosphere) radar mode in the late 1990s allowed routine observations of field-aligned equatorial irregularities causing coherently scattered echoes. Based on the radar observations made at Jicamarca, it has been suggested that the nighttimeFregion echoing structures seen in range-time-intensity (RTI) maps seem to fall within three major categories of “layers” [e.g.,Woodman and LaHoz, 1976; Hysell and Burcham, 1998]. This classification and the main differences between the different types of layers should be kept in mind when comparing radar observations with measurements made by other instruments and with numerical simulations of ionospheric instabilities [Hysell, 2000; Woodman, 2009].

[5] The first type of echoing layer usually seen in RTI maps is the “bottom type.” Bottom-type layers appear as relatively weak echoing regions in the Jicamarca RTI maps, and they seem to occur virtually every night over Jicamarca during ESF season. Bottom-type layers are narrow in altitude (a few tens of km) and their occurrence is confined to bottomsideFregion heights. The most striking feature of bottom-type layers, which was determined from interferometric radar observations, is the fact that they tend to move in the westward zonal direction. In some cases, bottom-type layers do not seem to move at all, or they move very slowly in the eastward direction. The second, and perhaps the most impressive, type of echoing structure is the “topside” layer, also referred to as radar “plume” because of its appearance in RTI maps [e.g.,Woodman and LaHoz, 1976]. Topside plumes are believed to be a manifestation of the so-called ESF depletions or plasma bubbles commonly observed by satellites and airglow instruments [e.g.,McClure et al., 1977; Sobral et al., 1980; Makela, 2006]. Topside structures are well developed vertically and, as the name suggests, reach topside ionospheric heights. Radar plumes can reach over 1,000 km altitude during moderate or high solar flux conditions. Finally, the third type of Fregion echoing layer commonly observed in the Jicamarca RTI maps is called the “bottomside” layer. Bottomside layers are layers that are more vertically developed than bottom-type layers, but are less developed than topside plumes. Bottomside layers are usually confined to heights around and below theFregion peak. RTI maps show bottomside layers as echoing regions that are broader (in altitude) and more structured than the bottom-type layers. Despite being observed very frequently, the bottomside layer does not seem to receive the same amount of attention that is given to topside events [Hysell, 1998]. This is, at least in part, because bottomside layers do not seem to be associated with the same level of low-latitude ionospheric disturbances that accompany topside plumes.

[6] While much about ionospheric irregularities can be learned from conventional (single-antenna) radar observations, interferometric radar imaging can greatly help to advance our understanding of ESF. There is not only interest in a better understanding of the morphology of ESF structures, and how these structures are related to different types of ionospheric processes, but also interest in the variability of these structures (and processes) with geophysical factors; longitude, in particular. Virtually all the low-latitude radar imaging results come from Jicamarca observations, with exception of the results ofRodrigues et al. [2008], which were limited to observations of a horizontally modulated bottom-type layer. Here, we present the results of a follow-up study of the application of the interferometric imaging technique [Hysell, 1996] to observations of more diverse types of ESF echoes made by the 30 MHz coherent backscatter radar interferometer installed in the equatorial site of São Luís, Brazil.

[7] The paper is organized as follows: Section 2 provides an overview of the experimental setup for Fregion observations in São Luís. It also provides a summarized description of how interferometric in-beam radar images are constructed.Section 3 describes the radar measurements made on 1 December 2005 when a miniplume was observed, followed by echoing layers that were confined to the bottomside F region. We also discuss the scintillation activity measured by a colocated GPS scintillation monitor during the radar observations. In section 4, imaging analysis results are presented and discussed in terms of current ESF theories and previous observations. A summary of our findings and final remarks are given in section 5.

2. Experimental Setup

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental Setup
  5. 3. Observations
  6. 4. Results and Discussion
  7. 5. Summary and Final Remarks
  8. Acknowledgments
  9. References
  10. Supporting Information

2.1. Instrumentation

[8] A low-power 30 MHz coherent backscatter radar has been used for observations ofE and F region irregularities over the equatorial site of São Luís, Brazil (2.59°S, 44.21°W, −2.35° dip lat). The radar site is located in the opposite side of the South American sector (along the magnetic equator) with respect to the Jicamarca Radio Observatory (JRO), which has been the main source of ionospheric radar observations at low latitudes. Observations at Jicamarca and São Luís, for instance, would allow studies of ionospheric irregularities across the South American sector [e.g., Swartz and Woodman, 1998]. The location of the São Luís and Jicamarca sites are shown in Figure 1a. Initially, the radar was designed for single-baseline observations ofE and F region ionospheric irregularities [de Paula and Hysell, 2004]. This setup allowed conventional radar observations (SNR and Doppler velocities) of 5 m irregularities as a function of range and time [e.g., de Paula et al., 2004; Rodrigues et al., 2004]. Because two separate antenna sets (arrays of 4 × 4 Yagis each) pointing vertically were available, the zonal velocity of the echoes could also be estimated using interferometric techniques [Kudeki et al., 1981].

image

Figure 1. (a) Map showing the location of the geomagnetic equator (dashed line) according to IGRF for December 2005 and the location of the Jicamarca and São Luís observatories. (b) Disposition of the antenna sets used in the estimation of nonredundant lags of the spatial spectrum.

Download figure to PowerPoint

[9] In 2004, two new antenna sets were added to the radar system and a total of four independent antenna sets started to be available for reception. The antenna sets are nonuniformly spaced, and are aligned in the magnetic east-west direction.Figure 1bshows the distribution of the antenna sets. The new antenna sets were installed to allow the computation of different nonzero lags of the spatial correlation function of coherently backscattered echoes, and application of interferometric imaging techniques. First results of the analysis of in-beam images generated using São Luís observations were presented byRodrigues et al. [2008]. They focused, however, on observations of a bottom-type layer only.

[10] The observations of F region irregularities made by the São Luís radar and used in this imaging study were carried out with two 4 kW transmitters, each one of them connected to a different antenna set for transmission. Reception is made by the four antenna sets. The observations presented here were made using 28 bit coded pulses. The baud length and range resolution of the observations are 2.5 km. The interpulse period (IPP) is 9.33 ms (1,400 km), and 250 samples are collected in each IPP starting at 200 km altitude. A summary of the main parameters of the observations is shown in Table 1.

Table 1. Main Radar Parameters for Observations of F Region Echoes
ParameterValue
Peak power8 kW
Code length28 bauds
Baud length2.5 km
IPP1400 km
Number of samples250
Initial sampling height200 km
Number of coherent integrations1
Number of FFT points4

2.2. In-Beam Radar Imaging

[11] The goal of the radar imaging technique is to obtain the angular distribution of the scattering structures responsible for the observed echoes. This distribution can be described by a real function f, which in radio astronomy and previous radar imaging studies is referred to as the brightness distribution. Interferometric radar measurements, however, provide estimates of the so-called visibility functiong, which is the covariance or spatial cross-correlation function for a limited number of antenna separations (or lags). The brightness and visibility functions, nevertheless, are closely related. For the case of equatorial field-aligned irregularities, for instance, the relationship between brightness and visibility can be described by [Hysell, 1996]:

  • display math

where k is the wave number of the radar signal, d is the spacing between antennas, ψ is the zenith angle in the magnetic equatorial plane, and A(ψ) represents the two-way antenna radiation pattern. We do not specify this function, butHysell and Chau [2006] showed that proper specification of A(ψ) can help reduce artificial features outside the sector illuminated by the radar. The variable ω is the Doppler frequency, and makes explicit that the covariance can be computed as a function of Doppler spectral bin. Hysell [1996] developed an algorithm that uses the maximum entropy (MaxEnt) approach to estimate the brightness distribution from visibility measurements and applied it to ionospheric radar observations at Jicamarca. The MaxEnt maximizes the Shannon (or information) entropy constrained by the measurements and their uncertainties. The MaxEnt algorithm of Hysell [1996] was used to create interferometric radar images from the observations made by the São Luís radar that are presented in this study. More details about the algorithm are given by Hysell [1996] and Hysell and Chau [2006].

[12] Two-dimensional in-beam radar images are constructed by stacking one-dimensional brightness distributionsf(ωψ) estimated for each range gate. Four spectral bins are used and different colors are assigned to each bin. Green is used to represent the zero-frequency component, while red, blue and magenta are used to represent red-shifted, blue-shifted and Nyquist components, respectively. The four images are added to produce the composite radar images. Pure color images indicate narrow spectral features, while color combinations indicate broad spectra.Rodrigues et al. [2008]provide additional details about the construction of in-beam images using São Luís observations.

3. Observations

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental Setup
  5. 3. Observations
  6. 4. Results and Discussion
  7. 5. Summary and Final Remarks
  8. Acknowledgments
  9. References
  10. Supporting Information

[13] Figure 2 shows a summary of the radar observations used in this study. It also shows complementary scintillation measurements available for this investigation.

image

Figure 2. (a) Range-time-intensity (RTI) map of the observations made by the São Luís radar on 1 December 2005. Vertical dashed lines indicate the times for which in-beam images have been generated and are shown inFigures 3 and 4. (b) Amplitude scintillation index (S4) measured by a GPS scintillation monitor for satellites with elevation greater than 45°. S4 > 0.1 are indicated by black crosses. (c–f) Maps indicating the location of the São Luís radar (star) and the location of the ionospheric piercing points (at 350 km) of satellites with elevation greater than 45°. The solid line shows the Brazilian coastline, and the dashed line shows the location of the geomagnetic equator. Similar to Figure 2b, satellites with S4 > 0.1 are indicated with black crosses.

Download figure to PowerPoint

3.1. Radar Echoes

[14] Figure 2ashows the range-time-intensity (RTI) map for observations made by the São Luís radar on 1 December 2005. The measurements made on this night were chosen, among other factors, because of their similarity to observations made at Jicamarca on 20 October 1997, also used to investigate features of scattering structures responsible for bottomside echoes with higher resolution than is currently possible with the São Luís radar interferometer [Hysell, 1999]. In addition, the São Luís measurements on this night show relatively strong echoes occurring throughout the entire period of interest here, which helps the accuracy of the resulting in-beam images. One noticeable difference between the São Luís and Jicamarca measurements is that Jicamarca clearly shows the presence of a bottom-type layer preceding the appearance of topside echoes. Bottom-type layers are often seen at Jicamarca, but do not seem to be observed as often in São Luís. This is, at least in part, due to the lower sensitivity of the São Luís radar compared to Jicamarca.

[15] The RTI map shows that echoes started to be observed near 350 km altitude and around 22:10 UT. This height is below but close to the F region peak according to ionosonde measurements. The São Luís ionosonde measurements showed that range spread F started before the echoes, around 21:45 UT and that, the height of the F2 peak was at 388 km immediately before ESF onset. The temporal resolution of ionosonde data is 15 min. The echoing layer gains altitude reaching the topside ionosphere suggesting the occurrence of ESF depletions (bubbles). After that, however, around 22:40 UT the echoes were confined to lower altitudes in the bottomside F region.

[16] We point out that two distinct patterns in the echoes occurring at bottomside F region heights can be observed. First, below approximately 260 km altitude the echoing layer is smooth and undifferentiated. Above 260 km, however, the echoes form a more structured and somewhat stronger (in terms of SNR) layer. It is impressive how the observations made by the radar on this day resemble the measurements made by the Jicamarca radar on 20 October 1997 [Hysell, 1999]. Jicamarca observations also show a smooth echoing layer below the bottomside layer, despite the data gap at earlier times. The similarity between the observations at Jicamarca and São Luís is indicative of common physical processes responsible for ESF occurring at different longitude sectors and times.

3.2. GPS Scintillation

[17] We also investigated the state of the equatorial ionosphere near the São Luís site during the radar observations by means of scintillation of transionospheric radio signals. Scintillation at GHz frequencies can help us to confirm whether or not radar echoes were associated with large-scale plasma depletions (ESF bubbles). During low solar flux conditions and postmidnight hours, topside depletions might not be associated with detectable levels of L band scintillation. However, moderate scintillation levels on L band signals at equatorial latitudes, especially in the early evening, is a strong indicator of topside ESF plasma depletions. After midnight and during low solar activity periods, the correlation between ESF bubbles and L band scintillation is not as obvious as a result of lower ionospheric densities and decay of small-scale irregularities. Measurements of the GPS L1 (1.575 GHz) signal made by a GPS-based scintillation monitor [Beach and Kintner, 2001] were available for this study. The monitor is colocated with the radar. While the radar is sensitive to electron density irregularities in the vertical direction with scale sizes matching the Bragg condition (λradar/2 for backscatter perpendicular to B), scintillation of radio signals are the result of diffraction caused by irregularities with scale sizes of a few hundreds of meters (Fresnel scale) transverse to the propagation path of the radio signal.

[18] Figure 2b shows the variation of the S4 index computed for L1 signals of all the GPS satellites above 45° elevation angle. The S4 is a widely used index that indicates the severity of amplitude scintillation, and is given by [e.g., Yeh and Liu, 1982]:

  • display math

where I is the signal intensity (power), and the angle brackets denote an ensemble average. In practice, time averages are used. L1 measurements were made at a 50 Hz sampling rate and S4 values were computed every one minute (3,000 samples).

[19] Figure 2b shows S4 values greater than 0.1 as black (dark color) crosses, while lower values shown as green (light color) dots. Figure 2b shows that moderate scintillation occurred on this night with S4values exceeding 0.4 even for high-elevation satellites. This indicates that plasma bubbles occurred near São Luís. Note that at the magnetic equator, plasma bubbles usually cause only moderate scintillation levels at GHz frequencies [Aarons, 1982] because of the reduced background electron density in the F region caused by the Fountain Effect that forms the equatorial anomaly [Schunk and Nagy, 2000; Kelley, 2009].

3.3. Radar Echoes and Scintillation

[20] Figures 2a and 2b indicate that moderate scintillation was only observed when topside echoes appeared in the RTI map. There is, however, a small time offset between the plume appearance, and the time when moderate S4 values (0.2–0.4) were observed. Figures 2c–2f show the piercing points (at 350 km) of the signals for which the S4 values were computed. They show that the piercing points of GPS signals were a few tens of km to the magnetic east with respect to the radar site location. ESF plasma depletions are aligned with the geomagnetic field and tend to move in the eastward direction during geomagnetically quiet conditions. This difference in magnetic longitude would explain the small time offset between the miniplume echoes in the RTI map and the onset of the most intense scintillation.

[21] Another interesting point to observe in Figures 2a and 2b is that our observations suggest that bottomside radar echoes might also be associated with detectable scintillation levels at GHz frequencies depending on receiver's background S4 values. In the absence of ESF, the S4 background values did not exceed 0.05. When bottomside echoes were more clearly seen, however, around ∼2300 UT the S4 index reached ∼0.15 (see Figure 2b). We must point out, however, that the piercing points of the scintillating signals were not exactly over the radar. The piercing points were to the north of the radar site but only a few km to the east with respect to the magnetic meridian (see Figure 2f). The magnetic declination in São Luís is approximately 21°W. Additionally, the piercing points were calculated using the mean height of the ionosphere (350 km) commonly used in GPS studies instead of the height of the diffracting layer. If the mean height of the bottomside layer (∼300 km) is considered, the piercing points would be even closer to the radar site. Our association of the bottomside layer with detectable levels of scintillation at GHz frequencies uses the common assumption that ESF irregularities follow the magnetic flux tube. While irregularities causing S4 ∼ 0.15 in L band signals do not represent a major threat, they might cause moderate scintillation levels on VHF signals. This is because the severity of amplitude scintillation depends on the signal frequency with S4 ∝ fn, where f is the signal frequency, and n is an index that varies between 0.8 and 1.6 [Yeh and Liu, 1982]. Further investigations of the relationship between bottomside layers, and scintillation on L band signals and VHF links will be carried out in a future more comprehensive study. VHF observations such as those made by SCINDA (Scintillation Network Aid) would be valuable to such study.

4. Results and Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental Setup
  5. 3. Observations
  6. 4. Results and Discussion
  7. 5. Summary and Final Remarks
  8. Acknowledgments
  9. References
  10. Supporting Information

[22] Figures 3 and 4 show results of our interferometric imaging analyses. They show sequences of images of the scattering structures within the radar beam. The images were produced with a temporal resolution of approximately 15 s. Here, we selected and present a series of images that best describe the temporal and spatial evolution of the scatterers producing the observed echoes.

image

Figure 3. Interferometric in-beam images of the scattering structures producing the echoes near the mainF region peak and topside as measured by the São Luís radar on 1 December 2005.

Download figure to PowerPoint

image

Figure 4. Interferometric in-beam images of the scattering structures producing the echoes in the bottomsideF region as measured by the São Luís radar on 1 December 2005.

Download figure to PowerPoint

[23] Each image shows the resolved spatial distribution of the scatterers within the beam at the indicated time (below each image). The positions of the scatterers are shown as a function of altitude and zenith angle (in the east-west magnetic plane). The images cover 300 km in the vertical direction, starting at 200 km altitude and extending up to 500 km altitude. The field of view of the images is approximately ±10°, but scatterers are better resolved within ±8°. The field of view of the images is dictated by the radar's antenna beam pattern. The small antenna arrays used by the São Luís radar allow us to produce images with a relatively large field of view. Our longest baseline (15λ), however, is too short to produce images with the same angular resolution possible at Jicamarca. The angular resolution of the São Luís images is approximately 0.4°.

4.1. Topside F Region Echoes

[24] Figure 3 (top) shows the distribution of the scatterers for the first few minutes after echoes started to be observed on the night of the observations (see Figure 2). The images demonstrate that the echoes, initially, come from a localized region around 350 km altitude. This is a height that is not much below the F2peak height (∼390 km) as measured by a colocated ionosonde. Radar plumes are associated with large ESF ionospheric plasma depletions that are produced by the Generalized Rayleigh-Taylor (GRT) instability. The GPS scintillation measurements confirm that an ESF depletion developed on the night of the observations. As first pointed out by G. Haerendel (Theory of equatorial spreadF, preprint, Max-Plank Institute fur Physik und Astrophysik, Garching, Germany, 1973), the stability of the equatorial ionosphere depends on flux tube integrated parameters rather than on local values. One of the most important parameters contributing for the linear growth rate of the GRT instability is the gradient of the flux tube integrated electron density [Sultan, 1996], which can maximize at heights that are well above the maximum in the local vertical electron density gradient and closer to the F region peak. Therefore, the fact that first echoes were found to occur near the F region peak could be indicative of the dependence of unstable regions on flux tube integrated parameters.

[25] The first scatterers seem to develop over the radar site without showing, for at least a few minutes, a clear sign of zonal motion. The scattering structure, however, develops in both zonal and vertical directions as time progresses. Figure 3 (bottom) continues to show the vertical development of the scattering structure. The images now show that a vertically elongated scattering channel is formed, and that it reaches topside ionospheric heights. The channel is only 4–6 degrees (or 24–36 km) wide at 350 km altitude. Previous studies at Jicamarca showed that topside plumes can be caused by very narrow scattering structures. Hysell [1999], for instance, showed that more developed radar plumes were produced by scattering channels whose width could be as narrow as a few km. Once the scattering structure fully developed, it started to move in the eastward direction (to the right) following, presumably, the background zonal ionospheric motion. Climatological studies of the background zonal plasma motion show eastward drifts after sunset and throughout the night [Fejer et al., 2005].

[26] Looking at the RTI map in Figure 2a, one could erroneously infer that the scattering layer is much wider in the zonal direction than it actually is. For instance the echoes associated with the plume started around 22:10 UT and lasted until approximately 22:40 UT. Assuming a typical zonal velocity of 100 m/s, one would estimate that the scattering structure responsible for the echoes was almost 200 km wide. Because of the dynamic nature of the irregularities causing the echoes, only the interferometric images like the ones estimated from the São Luís measurements can unambiguously determine the spatial distribution of the scatterers. A few hot spots with scatterers producing narrow, red-shifted spectra can be easily identified in these images. This is indication of the fast upward motion of the small-scale irregularities within the scattering channel.

[27] Bifurcation of the scattering structure responsible for topside echoes, as reported in previous studies [e.g., Hysell, 1999], was not observed during this event. This is, perhaps, because we were not able to image the entire development of the radar plume at this time. The images seem to show only the initial stage of the plume development. We did observe, however, a quasiperiodic spacing between scattering regions. This periodicity can be more clearly observed in Figures 3 (bottom middle and bottom right). Initially, the spacing is only a few km but increases as the structure reaches higher altitudes. Looking again at Figure 2a, we can see that the vertical propagation of these structures produce the diagonal echo streaks observed in the RTI map between 22:10 and 22:40 UT and 325 and 450 km altitude. We hypothesize that these quasiperiodic scattering structures are signatures of electron density perturbations in the vertical direction that have been commonly observed by rocket measurements during ESF events. In situ measurements of the electron density profile made by rockets have exhibit coherent, shock-like structures [e.g.,Costa and Kelley, 1978; Kelley et al., 1987]. These steepened, vertically propagating structures have been reproduced by numerical simulations of collisional interchange instabilities [Zargham and Seyler, 1987; Hysell et al., 1994]. Meter-scale irregularities responsible for radar echoes would be located in the leading (steepest) edges of the interchange irregularities [e.g.,Hysell, 2000; Woodman, 2009].

4.2. Bottomside F Region Echoes

[28] As mentioned earlier, the RTI map of the observations made on the night of 1 December 2005 show two types of echoing layers at bottomside Fregion heights after 22:40 UT. Below approximately 260 km altitude, the echoing layer is smooth and undifferentiated. Above 260 km, on the other hand, the echoes are stronger and the layer is more structured. The echoes at higher altitudes seem to fall well within the class of bottomside layers that are commonly seen at Jicamarca. The echoes at lower altitudes, however, better match the description of bottom-type layers.

[29] Figure 4 shows the results of our imaging analyses for observations made after 22:40 UT when echoes are completely confined to the bottomside F region. The images show a very interesting feature of the scattering layers responsible for the echoes in the bottomside Fregion. They show that the echoes below approximately 260 km come from a beam-filling scattering region. The irregularities forming this layer do not seem to show any significant structuring or clear motion in the zonal or vertical direction. Echoes above 260 km, on the other hand, are produced by a localized scattering region that drifts across the radar beam in the eastward direction (from left to right). This drifting structure has limited vertical development, reaching only ∼320 km altitude. The scattering structure is also tilted to the west.

[30] The images of the scattering structure responsible for the bottom-type layer do not show any clear signature of primary waves. This can be a limitation of the low resolution of the São Luís images (∼0.4° in the zenith angle and 2.5 km in the vertical direction). The fact that bottom-type irregularities do not show clear zonal motion (as seen in our images) or that they tend to move in the westward direction suggests that they are confined in magnetic flux tubes that are controlled by theE region dynamo [e.g., Kudeki and Bhattacharyya, 1999; Hysell, 2000]. Assuming that neutral winds are eastward, Kudeki and Bhattacharyya [1999]suggested that bottom-type layers would be confined to a region of retrograde plasma motion. That region would provide favorable conditions to the development of wind-driven interchange instabilities.Hysell and Burcham [1998] pointed out that even that even though Eregion conductivities are significant on the flux tubes to which the bottom-type irregularities are limited to, they are not able to short out the polarization electric fields of growing plasma waves and to damp the instability. Based on the efficiency of the mapping of transverse (toB) electrostatic structures as a function of scale size [Farley, 1960; Hysell and Burcham, 1998] hypothesized that collisional interchange mode waves with wavelengths shorter than approximately 1 km would not be shorted out, and could exist in flux tubes controlled by the Eregion dynamo. This hypothesis was supported by numerical simulations. Like the São Luís images shown here, first radar imaging observations at Jicamarca also showed that bottom-type echoes were caused by undifferentiated scattering regions without signs of primary waves [Hysell, 2000]. The addition of a longer baseline for Jicamarca interferometric observations, however, allowed the imaging technique to better resolve the scattering structures. It was then shown that the scattering structures responsible for bottom-type layers were composed of kilometric primary waves [Hysell et al., 2004]. Furthermore, it was found that the primary waves would grow by advection as one would expect from the hypothesis put forward by Kudeki and Bhattacharyya [1999]. This also explains the lack of vertical development of bottom-type layers.

[31] The São Luís observations indicate the same phenomenology over the Brazilian sector. The images reinforce that bottom-type layers are caused by beam-filling scattering structures, and that primary waves must have wavelengths shorter than the resolution of our images, that is 2–3 km. Analyses of a large database of RTI maps measured by the JULIA radar at Jicamarca show that bottom-type echoes are generally seen first [Hysell and Burcham, 1998, 2002]. The bottom-type layers usually ascend in altitude, presumably, under the action of the prereversal enhancement (PRE) of the evening zonal electric field. Sometimes, after ascending to a certain altitude, they start to descend and bottomside or topside plumes are not observed. Other times, however, the bottom-type echoes give place to topside and bottomside layers. This transition would be caused by a change in the altitude of the flux tube, which then would cause a reduction in theEregion loading allowing the collisional interchange instability to develop. São Luís observations show the interesting case of simultaneous occurrence of bottom-type and bottomside layers. The bottomside layer is located at heights above the bottom-type layer. The boundary between bottom-type and bottomside echoes would then delineate a transition height for theE region loading.

[32] The São Luís images show that the scatterers responsible for the bottomside layer are structured with well-defined eastward motion, unlike the scattering region responsible for the bottom-type echoes. Again, because the low resolution of the São Luís images compared to Jicamarca observations, the tendency is to highlight features with scales greater than a few km in the images. Nevertheless, we are able to identify that the scattering structures responsible for bottomside and bottom-type layers are morphologically distinct. Supposedly, the bottomside scattering structure would be produced by intermediate-scale electron density structures caused by interchange instabilities driven by gravity or by a zonal electric field and would grow by convection.

[33] Finally, we would like to point out that the morphological features of this scattering structure responsible for the bottomside layer seen by the São Luís radar on this night resembles those of plasma perturbations produced by numerical simulations of the collisional shear instability (CSI) [Aveiro and Hysell, 2010]. Here we must assume that meter-scale irregularities are good tracers of larger-scale electron density perturbations, which has been shown to be true at times [e.g.,Tsunoda et al., 1982] but needs further investigation. Aveiro and Hysell [2010] suggested that bottomside scattering layers could be a result of the CSI in conditions less favorable to the GRT instability.

5. Summary and Final Remarks

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental Setup
  5. 3. Observations
  6. 4. Results and Discussion
  7. 5. Summary and Final Remarks
  8. Acknowledgments
  9. References
  10. Supporting Information

[34] We presented new results of a follow-up study of the interferometric coherent backscatter radar imaging technique applied to observations made by the São Luís radar. Initial results were presented byRodrigues et al. [2008]but were limited to observations of a bottom-type layer. Here, we presented and discussed in-beam radar images of scattering structures responsible for different types ofFregion echoes observed on the night of 1 December 2005. The range-time-intensity (RTI) map of the observations shows first echoes occurring near the mainF region and topside heights followed by an extended period of echoes confined to bottomside F region heights. These measurements provided an opportunity to investigate the morphology of the scattering structures responsible for different types of equatorial spread F echoing layers seen by the São Luís radar in the Brazilian sector. Our interferometric imaging results show that topside echoes were caused by a localized scattering structure. This structure was horizontally narrow (30 km), developed vertically from below the F region peak toward the topside, and moved in the eastward direction. We also observed quasiperiodicities in the vertical direction of this scattering structure. We associated them with steepened electron density structures previously observed by rockets experiments and reproduced in numerical simulation of collisional interchange instabilities. The topside echoes are associated with ESF depletions as evidenced by moderate GHz scintillation levels (S4 ∼ 0.4) measured by a colocated scintillation monitor. The RTI map of the observations also shows the simultaneous occurrence of two different types of echoing layers in the bottomside Fregion. One layer was weak, diffuse and confined to lower altitudes. The second layer is more structured, intense and located at higher range gates. The in-beam images reveal that these echoing layers are caused by two distinct types of scattering structures, which indicates different plasma instabilities operating simultaneously, but at different altitudinal ranges. Despite the low resolution of the São Luís images, they present an overall similarity with previous imaging observations made at Jicamarca. The resolution of the São Luís images could be improved with the addition of a longer baseline. Finally, we found that bottomsideF region echoes could be associated with very weak, nevertheless detectable, GPS (1.575 GHz) scintillation levels (S4 ∼ 0.15).

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental Setup
  5. 3. Observations
  6. 4. Results and Discussion
  7. 5. Summary and Final Remarks
  8. Acknowledgments
  9. References
  10. Supporting Information

[35] This study was supported by NSF award AGS-1024849. The authors would like to thank the technical staff of the São Luís Observatory. The authors would like to thank I. S. Batista and M. G. S. Aquino for providing us with the São LuíshmF2 ionosonde information. F.S.R. would like to thank D. Hysell for numerous valuable discussions about the São Luís radar and the interferometric radar imaging technique. São Luís radar was partially supported by FAPESP grants 99/00026-0 and 04/01065-0.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental Setup
  5. 3. Observations
  6. 4. Results and Discussion
  7. 5. Summary and Final Remarks
  8. Acknowledgments
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Experimental Setup
  5. 3. Observations
  6. 4. Results and Discussion
  7. 5. Summary and Final Remarks
  8. Acknowledgments
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
rds5974-sup-0001-t01.txtplain text document0KTab-delimited Table 1.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.