Dynamics of the Polar Cap Ionosphere.
Part 2. Case Studies of Polar Patches above Casey, Antarctica.

A.M. Breed*, T.M. Maddern*, P.L. Dyson# and R.J Morris*

* Atmospheric and Space Physics Group,
Australian Antarctic Division, Channel Highway,
KINGSTON, TASMANIA, 7050, AUSTRALIA
(e-mail: anthon_bre@antdiv.gov.au)

#Department of Physics, La Trobe University,
BUNDOORA, VIC. 3083, AUSTRALIA
(e-mail: p.dyson@latrobe.edu.au)

ABSTRACT

During 1997 and early 1998 a series of campaigns were undertaken at the Australian Antarctic polar cap station Casey (-80.8o geomagnetic latitude) with the aim of identifying and gaining further understanding of polar patches. The main instrument used for these campaigns was a UMLCAR (University of Massachusetts Lowell, Centre for Atmospheric Research) Digital Ionosonde (DPS-4), running a 3-minute cycle of Doppler ionograms and drift velocity measurements. Patches identified from group range and critical frequencies on the ionogram records were compared with drift measurements and other geophysical data sets. Good correlations were observed with Total Electron Content (TEC) derived from Global Positioning System (GPS) satellite observations with a receiver also located at Casey. The field-of-view (FOV) for GPS observations allowed patches to be observed over large distances, providing some insight into their origin. The influence of the Interplanetary Magnetic Field (IMF) on the formation and occurrence of patches was also considered. Comparison with plasma drift velocities, (derived from DPS drift measurements), confirmed that patches are associated with fluctuations in horizontal drift velocity. In the current study, peaks in the magnitude of the horizontal drift velocity correlate with the patch edges. This paper considers case studies of three days of patch observations in April 1998.

1. INTRODUCTION

1.1 Morphology

The term polar patch (hereafter referred to as patches) is used to refer to enhanced regions of ionospheric F-layer plasma density, convecting across the polar cap in a generally antisunward direction.  (Similar enhancements observed outside the polar cap are generally called auroral blobs.)  The patches have horizontal scale sizes of 100 – 1000 km and are observed travelling at velocities in the range 300 – 1000 m/s.  The patches themselves often contain smaller scale structure and irregularities (1 – 10 km range), particularly on the edges as determined from scintillation effects.

The first observations of these phenomena were made by Hill, (1963), but it appears no intensive studies of patches were undertaken until those reported by Buchau et al., (1983, 1985) and Weber et al., (1984,1986).  In-depth reviews of polar patch morphology, dynamics and formation processes are given by Crowley, (1996) and Rodger, (1998).
Patches are observed to have plasma densities comparable with the dayside sub-auroral ionosphere, suggesting that they originate in this region and are subsequently transported through the cusp region to the polar cap by the high-latitude convection.  Studies have shown that southward interplanetary magnetic field (IMF)  (Bz negative) is conducive to the formation of patches, as is a disturbed geomagnetic field (Kp greater than 4) (Buchau et al., 1983; Weber et al., 1984),.  Patches may also form simultaneously in geomagnetically conjugate regions (Rodger et al, 1994b).  The electron temperature of patches (determined from incoherent scatter radar measurements) is low (Anderson et al, 1988) indicating that they are not caused by particle precipitation.

Diurnal and seasonal occurrence rates of patches are discussed by Rodger and Graham, (1996).  Their data indicated maxima in occurrence around magnetic noon between equinox and winter seasons.

1.2 Formation Mechanisms

A number of possible formation mechanisms of polar patches have been put forward and discussed or modelled in the literature. Although there is no universal agreement on the validity of the individual mechanisms (some depend on how a polar patch is specifically defined), there does appear to be general agreement that the source of ionisation for patch formation is the enhanced ridge of plasma called the ‘tongue of ionisation’ created by the polar convection dragging sunlit plasma equatorward of the cusp into the polar cap (Steele and Cogger, 1996).  The method by which this plasma is ‘broken up’ into patches is a source of debate, although overall there seems to be approximately three main themes. They are :

  1. The rapid expansion or change in the polar convection patterns associated with either IMF Bz transitions (Valladares et al., 1998) or transient magnetopause reconnection (Flux Transfer Events (FTEs)) (Lockwood and Carlson, 1992), ‘modulating’ the plasma flow from the dayside into patches.
  2. The distortion and fragmentation of an existing steady plasma ridge (tongue of ionisation) by rapid changes in IMF By and possibly gravity waves (Sojka et al., 1993).
  3. The ‘chopping’ of an existing, steady plasma ridge into patches by fast plasma jets (Flow Channel Events (FCEs)) also associated with FTEs. The increased plasma velocities (several kilometres per second) decrease the expected F-region plasma lifetime, enhancing recombination and causing depletions in the plasma ridge (Rodger et al., 1994a; Valladares et al., 1994, 1996). This appears to be the most widely accepted formation mechanism.

1.3 Signatures and detection
Polar patches are detected by a wide range of ground-based instruments and have a correspondingly wide array of signatures. Most early observations relied on an enhanced 630 nm optical emission detected with photometers or imaged with intensified all-sky cameras. A number of recent studies have still used this technique of detection (Fukui et al., 1994; McEwen et al., 1995a, 1995b). Sojka et al., (1997) discuss problems comparing optical measurements with radio frequency observations.

More recently, signatures of patches have been detected with satellite-borne instruments either by in-situ sampling (Coley and Heelis, 1995; Kivanc and Heelis, 1997) or through remote sensing using visible, ultraviolet or x-ray imagers (eg. on board the Polar satellite). Also involving satellites are the techniques of Total Electron Content (TEC) measurements, using Global Positioning System (GPS) or Navy Navigation Satellite System (NNSS) satellite beacons, which allow patches to be detected at ground receivers as large increases in line-of-sight electron content or radio scintillation on satellite signals (Weber et al., 1986).

Ground based radio frequency signatures include increased absorption measurements on riometers ) (Rosenberg et al., 1993; Wang et al., 1994) and electron density structures observed with HF radars (incoherent scatter radar or coherent scatter radar (Ogawa et al., 1998a, 1998b).

Of more particular importance to the current studies are the signatures of patches observed on digital ionosondes. The most obvious effect of polar patches seen on ionosonde measurements are of course increases in critical frequency (foF2) associated with the increased plasma density. However, a more definitive signature appears in plots of h’t, (virtual height versus time at a constant frequency). James and MacDougall, (1997) have modelled the effects of electron density structures on ionosonde transmissions and predict the appearance of ‘U-shaped’ (or hyperbolic) traces in ionograms and h’t plots. These U-shaped traces are shown in the studies of MacDougall et al., (1996), where they are the primary method of patch identification. The formation of the U-shaped trace relies partly on the fact that echoes observed with the digisonde are not necessarily from overhead. As a high-density patch approaches the sounder, echoes are picked up at large, oblique ranges. The range of these echoes decreases as the patch approaches, and then increase again as it recedes from the station, hence producing the U-shape in the h’t plots. Figure 1 of MacDougall et al., 1996, shows examples of the patch associated U-shaped plots from measurements made with a Canadian Advanced Digital Ionosonde (CADI). With digital ionosondes (such as the DPS-4) parameters such as direction of arrival and Doppler shift of reflected signals are also available and can further contribute to the identification of patches in data records.

22K; 4 patches

Figure 1: A series of four patches identified on Casey digisonde records for April 9, 1998 (DOY 99). Virtual ranges are shown for both 3 and 4 MHz transmission frequencies.


In addition to these methods of detection, the DPS-4 also has the facility for drift mode operation in which high resolution measurements of echo location (angle of arrival) and Doppler shift can be made over a number of echo sources, heights and frequencies, by the method of Doppler sorted interferometry (DSI) (Reinisch et al., 1995; Scali et al., 1995), effectively allowing ‘Skymaps’ of echo sources to be produced. Although the signature of patches in this type of display is still under investigation, it is expected that the edges of large patches would appear as ‘fronts’ of echoes crossing the field-of-view, associated with the presence of small scale irregularities, while smaller patches may be seen in their entirety (Breed et al., 1999).

1.4 Current Studies


The current paper presents and discusses some case studies of polar patches, concentrating on patches identified on three days of a two-week campaign in April 1998. Some mention is also made of patches identified at other times. Patches were identified from the digisonde measurements visually by the U-shaped trace described above. Measurements of critical frequency, TEC, drift velocity and 630 nm optical emissions were then examined for corresponding patch signatures.

2. INSTRUMENTS AND ANALYSIS

The experiments described in this paper were located at Casey base, in the Australian Antarctic Territory. Casey lies at 66°17’ S and 110°32’ E geographic, (-80.8° geomagnetic latitude). Magnetic local time (MLT) at Casey is approximately UT+5.5 hours.

The primary instrument used in this study is the UMLCAR Digisonde DPS-4. This instrument operates as a pulsed, Doppler radar, and consists of crossed rhombic transmit antennae and four crossed loop receive antennae (and associated independent receivers) in a triangular array, allowing measurements of angle-of-arrival, polarisation, and Doppler shifts of ionospheric echoes. For the patch campaign in April 1998, the DPS-4 was running an operating schedule consisting of a beam-forming ionogram (BEM) followed by a drift mode program, repeated every three minutes.

Beam-forming ionograms use the directional capabilities of the DPS-4 receive array to sort returning signals (maximum amplitude signal in each range and frequency bin) into one of seven directional ‘beams’ (consisting of one ‘vertical’ (to 20 degrees from zenith) beam and six off-vertical beams (between 20 and 45 degrees from zenith)). The vertical beam is further divided into ordinary and extraordinary polarisations. Coarse (one bit) Doppler information is also assigned to the echoes.

In DPS-4 drift mode sounding echoes are sorted by Doppler shift (at a given frequency and virtual range), and accurate angles of arrival are obtained using phase shifts across the four receive antennae (hence the term Doppler sorted interferometry (DSI)). By varying the number of samples and the sample rate (by changing the number of pulses transmitted and the pulse repetition rate) the Doppler resolution and range of the Fourier transformed time series, and the number of possible echo sources (up to a maximum of 128 per range and frequency), can be controlled. Hence, from the drift mode measurements, a three dimensional ‘Skymap’ of echo sources and associated Doppler shifts above the sounder can be constructed. Using a least squares fit to at least three echoes apparent ‘drift’ velocities are calculated. The analysis is carried out using the Digisonde Drift Analysis (DDA) software package (Reinisch et al., 1995; Scali et al., 1995; Parkinson et al., 1999a).

To assist in identifying polar patches in digisonde data a suite of programs were to analyse and display particular data parameters. This involved extracting and plotting critical frequencies (foF2 and FxI) and virtual ranges (at fixed frequencies) from the BEM ionograms, using colour to show Doppler signs or azimuths of echoes. Drift data were used directly to obtain ionospheric velocities or displayed in the animated ‘ionomovie’ format (Breed et al. 1999).

Data from co-located GPS receivers were used to obtain TEC values and scintillation measurements. TEC measurements were corrected for satellite and receiver biases, and transformed from slant to vertical TEC values. This transformation assumes a median ionospheric height of 400 km to define a subionospheric point and uses the cosine of the zenith angle to project slant TEC to vertical TEC values (Breed et al., 1998). The subionospheric point is the spot on the Earth’s surface directly below the intersection of the receiver-to-satellite path and 400 km altitude. Scintillation measurements were corrected according to the method of Nichols et al., (1999). Due to the inclination of the GPS satellite orbits, and an imposed zenith angle limit of 50 degrees, the majority of measurements with these receivers applied to sub-ionospheric points several degrees geographic north of Casey station.

Interplanetary magnetic field (IMF) component values, in Geocentric Solar Magnetic (GSM) coordinates, were obtained from Wind spacecraft data (1.5 minute averages). These were corrected to allow for a propagation time delay from the spacecraft to the Earth’s bow-shock (assumed to be at 10 Earth radii) by using a simple approximation of distance divided by solar wind speed (also obtained from Wind spacecraft data). No allowance was made for any subsequent travel time caused by the slowing down of the solar wind close to the bow-shock or propagation time from the outer magnetosphere to the ionosphere that may contribute up to another 10 minutes to the delay.
Simple optical measurements at 630 nm from a wide-angle photometer and ionospheric absorption measurements from a standard, 30-MHz riometer at the station were also available. Both data sets will be briefly considered.
Other data available at Casey station, but not used in the present study, are induction and fluxgate magnetometers and an all-sky intensified video imager. These may be included in ongoing polar patch research and reported in future papers.

3. CASE STUDIES

The case studies discussed in this paper were from 3 days in April 1998. These intervals were not the only observations of polar patches in this period, but were picked as times displaying examples of large, well-defined patches, demonstrating their observation in DPS measurements and showing clear correlations with other data sets.

3.1 April 9, 1998 (Day 99)


Figure 1 shows the first case study of a series of patches observed on Day 99 (9-April-1998). (Titles on figures of the form “Day 8099” are derived from the digisonde data file system and consist of a single-character year and three-character day number. eg. Year 1998, Day 099). Figure 1 shows the h’t plots (group range as a function of time) for the frequencies 3 and 4 MHz, over a period of 5 hours. The patches are identified from U-shaped traces on the 4 MHz plot, as there is minimal effect on the 3 MHz group ranges in the current example. This arises because at this time (4 – 6 UT) on day 99 the 3 MHz signal was being reflected in the F1-layer (and on occasion the E-layer), and hence was not affected by the patches that appear to only disturb F2-layer heights (4 MHz signal). This is not always the case, and depends on time of day and the background (normal) electron densities. At nights or times of low background density, the patches are more likely to be seen at lower frequencies. The interpretation of patches seen on 4 MHz range plots but not in 3 MHz plots are that In figure 1 we have indicated 4 distinct patches. However, the critical frequency foF2 enhancements (see figure 2) suggest that there may be more, less distinct patches present in the data.

Figure 2: A comparison between patches identified on the 4 MHz h’t plot, foF2 critical frequencies and TEC from GPS satellites (different symbol for each satellite). April 9, 1998 (DOY 99).


Figure 2 shows the 4 MHz. trace from figure 1, combined with foF2 values (also derived from DPS ionograms, with three point smoothing applied) and TEC derived from GPS satellite reception (averaged and restricted to satellites with elevation angles greater than 50 degrees). A comparison between these three data sets must take into account that they have different fields of view. The h’t plot shows all 4 MHz echoes down to an angle of 45 degrees off vertical while foF2 measurements are restricted to angles less than 20 degrees. This should have little effect on comparison between these plots except that patches may be seen first on the h’t results. GPS TEC measurements relate to a small region 2-3 degrees geographic north of Casey and hence, depending on time of day and direction of patch drift, may exhibit patch signatures some time before or after DPS measurements.

The dashed vertical lines on figure 2 have been added to allow easy comparison between the three plots and indicate approximate times of closest approach to the station for the three patches. These times are estimated from the h’t plots as the minima of the U-shaped features.

It is immediately obvious that the patch occurrences do indeed correspond to peaks in critical frequency. It is noted that there are a number of closely spaced peaks in foF2 prior to 4 UT, most likely indicating a series of smaller patches.  Evidence of these is also present in the h’t plot, although much less distinct than the later patches.

The lower plot indicates increases in TEC assumed to correspond to all three patches indicated. There are two or more distinct traces shown in this panel corresponding to different GPS satellites and hence different observation locations (although still within a relatively small region north of Casey). Some time discrepancies exist in the observations of these increases and the closest approach of the patches, most significantly with the earlier patch (some 20 minutes), but also to a lesser extent with the second patch. A simple explanation for this would be the physical separation of the observations, with GPS TEC measurements corresponding to some 2-3 degrees North (geographic) of the h’t and foF2 measurements. Assuming a separation of around 3 degrees latitude, the 20-minute delay of the first patch would correspond to a patch velocity of around 300 m/s.  Horizontal drift velocities derived from DPS drift measurements decreased considerably (from around 600 m/s to less than 200 m/s) for approximately 20 minutes following a sudden change in IMF By/Bz at 4 UT, evident in figure 3.

Figure 3: Interplanetary magnetic field parameters By and Bz for April 9, 1998 (DOY 99).

A second possible explanation for the absence of direct correlation between observations involves the sudden changes in the IMF By and Bz occurring just prior to appearance of the first patch (see Figure 3). It is possible that these changes may have altered or disrupted the polar circulation pattern sufficiently for the patch to miss the TEC measurement points. DPS measurements around 4 UT indicate a drift direction shifted approximately 45 degrees away from the antisunward (geographic South) direction towards the post-midnight sector (ie. originating in the afternoon sector). An example of the limited spatial extent of some patches is demonstrated later in this paper. It is interesting to note as well that the time interval between the first two identified patches (approximately 40 minutes) appears to correspond to the time between the preceding south-north and north-south transitions in Bz suggesting a possible connection.

Figure 4:  A series of patches seen in 630-nm photometer data, compared with foF2 critical frequencies and TEC measurements. April 9, 1998 (DOY 99).

Figure 5: Interplanetary magnetic field parameters By and Bz for April 9, 1998 (DOY 99).

Figures 4 and 5 show a period later on the same day when optical measurements were available from a 630 nm wide angle photometer. This photometer has a 60 degree field of view and is collocated with the digisonde. Hence peaks in 630 nm airglow arising from patches should correspond temporally to peaks in foF2 critical frequency.  Figure 5 shows the state of the IMF during the period, indicating a generally southward (negative) Bz component conducive to polar patch formation (Buchau et al., 1983). Figure 4 includes the photometer measurements (uncalibrated), foF2 critical frequency and TEC. This figure shows almost a continuous series of enhancements (patches) apparent in the 630 nm band. These generally correspond well with both foF2 frequencies and TEC measurements, although there are large differences in the relative magnitudes of the increases as seen by the different techniques. This is in line with the results of Sojka et al. (1997).

3.2 April 10, 1998 (Day 100)

Figure 6: Virtual range versus time plots at 4 MHz frequency, showing Doppler shifts and directions of arrival for a series of patches identified on April 10, 1998 (DOY 100).


Figure 6 shows a h’t plot at 4 MHz on day 100, 1998. Again the characteristic U-shaped traces are seen indicating the presence of large patch features. In this plot, colour is used to show basic Doppler shift in the top panel and direction of arrival in the bottom panel. The change in Doppler can be clearly seen in the top panel as the patch passes the point of closest approach, from the positive Doppler (cyan) of the approaching patch to the negative Doppler (red) of the receding patch. A signature is not so easily seen in the bottom panel as most of the echoes have been classified as overhead (Vo) within 20 degrees of zenith.

Figure 7: A comparison between patches identified on the 4 MHz h’t plot, foF2 critical frequencies, TEC measurements (different symbols for each satellite), horizontal drift velocity and plasma drift direction (referenced to the antisunward direction). April 10, 1998 (DOY 100).


Figure 7 shows the same patch sequence in h’t, with corresponding critical frequencies and TEC measurements. Also included on this plot is the mean horizontal drift velocity determined from DPS drift measurements and the drift azimuth (the apparent direction of plasma motion). The centre line (0 degrees) in this lower panel represents the antisunward direction.. The vertical dashed lines again indicate the approximate times of closest approach of the patches. The identification of an unambiguous single patch becomes difficult in the period after 9:30 UT, with foF2 and TEC indicating the possible presence of 2 or more overlapping patches. What can be seen from this figure is again the close relationship between the U-shaped feature and peaks in foF2 and TEC. The fourth panel of this figure also indicates an apparent link between the patches and horizontal drift velocity, with minima observed at the points of closest approach to the station and surges in velocity associated with the patch edges. This is in agreement with some of the findings of MacDougall et al, (1996). From this figure there appears to be no conclusive signatures of patches in the drift azimuth directions. The first patch in the series is associated with a 90-degree (dawnward) fluctuation in drift azimuth, while the second patch appears to have generally antisunward flow. The third patch and associated drift directions could possibly be described as showing a vortical motion (ie. drift azimuth swinging from +90 degrees through antisunward to –90 degrees.

3.3 April 16, 1998 (Day 106)
Figure 8: A comparison between patches identified on the 4 MHz h’t plot, foF2 critical frequencies, TEC measurements (different symbols for each satellite) and horizontal drift velocity. April 16, 1998 (DOY 106).

The last case study is shown in figure 8, again in the form of a h’t plot. As with the previous examples, this is compared with foF2 frequencies, TEC and horizontal drift velocity (smoothed in this plot with a 3 point running mean filter). Up to 6 patches are indicated by the vertical dashed lines. This period differs significantly from the previous cases in that up until approximately 7:30 UT the IMF Bz component is consistently northward (positive) (see figure 9).
Figure 9: A comparison between patches identified on the 4 MHz h’t plot, plasma drift direction (referenced to the antisunward direction) and IMF Bz component. April 16, 1998 (DOY 106).
As a consequence of this, the drift azimuths tend to show sunward flow although in general the directions are quite variable, up to this time. At 7:30 UT as Bz turns southward (negative), the drift directions appear to stabilise and swing more antisunward, although at times showing up to 90 degree deviation as Bz oscillates between northward and southward. As shown on figure 8, there is again close correlation between the U-shaped patch signatures, peaks in foF2 and TEC (although there was a limited amount of this data available on this day) and minima in horizontal drift velocity. The TEC data available (shown in more detail in figure 10) displays the apparent narrow longitudinal extent of the final patch feature.

Figure 10: TEC measurements from GPS satellites for a single patch identified in Figures 10 and 11, showing positions and measurements from the two individual satellites visible at the time. April 16, 1998 (DOY 106).

This figure shows the TEC as measured along the ray paths to the two GPS satellites visible at the time. A large increase in TEC (from 10 – 20 TEC units) is seen on satellite PRN26 at a subionospheric longitude of 109.2°E, while no similar increase is seen on satellite PRN2 at a subionospheric longitude of 112.5°E.  At this time of day (10-11 UT, late afternoon local time), patches would be expected to approach from the west (geographic) (if travelling approximately antisunward) and hence would tend to be observed on a more westward satellite as found in this case. However, the complete lack of any increase on the second satellite, only approximately 150 kilometres away, suggests a sudden, large density gradient at the edge of this particular patch. This is also suggested by the large variation seen in foF2 in figure 8.

4. DISCUSSION AND CONCLUSIONS

Although still only preliminary at this stage, the results from these case studies successfully demonstrate the use of the DPS-4 digisonde in identifying and studying polar cap ionosphere features such as polar patches. The polar patches have been identified using a U-shaped signature found in Northern Hemisphere studies by James and MacDougall (1997). The patches have been compared and correlated with ionogram critical frequencies, Doppler shifts and directions, Total Electron Content from GPS satellite observations, plasma drift velocity and direction (also from DPS-4 measurements), 630 nm photometer measurements and Interplanetary Magnetic Field parameters.

The significant finding from these case studies is the apparent correlation of horizontal drift velocities with patches. Surges in the observed velocity are associated with the patch edges. No definite conclusions are made regarding this phenomenon at this time due to the possibility that these velocity surges are not part of the bulk plasma flow, but may in fact be an ‘artifact’ caused by turbulence or instabilities at the patch edges (Rodger, 1998). However, as one of the main creation mechanisms for polar patches are Flow Channel Events (fast plasma jets), it seems reasonable to expect remnants of such events to be observable in the patches (in particular, as most of the patches shown in this study are on the dayside, close to the cusp region where they are formed).

With further study and analysis (incorporating scintillation measurements, SuperDARN radar data and other ionosondes) we expect to be able to resolve some of the questions left unanswered in the present study.

Acknowledgments

R. P. Lepping and the Wind MAG (MFI) Data Processing team of NASA/Goddard Space Flight Center are thanked for allowing us to use Wind IMF data. The Australian Surveying and Land Information Group (AUSLIG) and La Trobe University are thanked for the use of data from their GPS receivers located at Casey.

REFERENCES

Anderson, D.N., Buchau, J. and Heelis, R.A., Origin of density enhancements in the winter polar cap ionosphere., Radio Science, 23, No. 4, pp. 513-519, 1988.

Breed, A.M., Goodwin, G.L., and Silby, J.H., Total Electron Content measurements in the southern hemisphere using GPS satellites, 1991 to 1995, Radio Science, 33, No. 6, p. 1705, 1998.

Breed, A.M., Maddern, T.M., Dyson, P.L. and Morris, R.J., Dynamics of the Polar Cap Ionosphere Part 1: Applications of Digisonde animations in ionospheric research., Australian National Antarctic Research Expeditions (ANARE) Research notes, current volume, 1999.

Buchau, J., Reinisch, B.W., Weber, E.J., and Moore, J.G., Structure and dynamics of the winter polar cap F-region., Radio Science, 18, No. 6, pp. 995-1010, 1983.

Buchau, J., Weber, E.J., Anderson, D.N., Carlson, H.C., Moore, J.G., Reinisch, B.W. and Livingston, R.C., Ionospheric structures in the polar cap : Their origin and relation to 250-MHz scintillation., Radio Science, 20, No. 3, pp. 325-338, 1985.

Coley, W.R. and Heelis, R.A., Adaptive identification and characterisation of polar ionization patches., Journal of Geophysical Research, 100, No. A12, pp. 23819-23827, 1995.

Crowley, G., Ionospheric patches and blobs., The Review of Radio Science 1993-1996., Ed.: W. Ross Stone, Oxford University Press, pp. 619-649, 1996.

Fukui, K., Buchau, J. And Valladares, C.E., Convection of polar cap patches observed at Qaanaaq, Greenland during the winter of 1989-1990., Radio Science, 29, No. 1, pp. 231-248, 1994.

Hill, J.R., Sudden enhancements of F-layer ionization in the polar regions., Journal of Atmospheric Science, 20, pp.49, 1963.

James, H.G. and MacDougall, J.W., Signatures of polar cap patches in ground ionosonde data., Radio Science, 32, No. 2, pp. 497-513, March 1997.

Kivanc, Ö. And Heelis, R.A., Structures in ionospheric number density and velocity associated with polar cap ionization patches., Journal of Geophysical Research, 102, No. A1, pp. 307-318, January 1997.

Lockwood, M. And Carlson, H.C., Production of polar cap electron density patches by transient magnetopause reconnection., Geophysical Research Letters, 19, No. 17, pp. 1731-1734, 1992.

MacDougall, J.W., Grant, I.F. and Hamza, A., Velocity fluctuations associated with polar patches., Radio Science, 31, No. 3, pp. 595-606, 1996.

McEwen. D.J. and Harris, D.P., Observations of F-layer patches and their convection over the north magnetic pole., Advances in Space Research, 16, No. 1, pp. 69-72, 1995a.

McEwen, D.J., Harris, D.P., MacDougall, J.W. and Grant, I.F., Drifting F-layer patches over the magnetic pole., Journal of Geomagnetism and Geoelectricity, 47, pp. 527-537, 1995b.

Nichols, J., Hansen,, A., Walter, T. and Enge, P., High latitude measurements of ionospheric scintillation using the NSTB., draft paper, Dept Aeronautics and Astronautics, Stanford University, 1999.

Ogawa, T., Nishitani, N., Pinnock, M., Sato, N., Yamagishi, H. and Yukimatu, A.S., Polar cap patches and auroral blobs observed with Antarctic HF radars: Preliminary results., Proceedings NIPR Symposium on Upper Atmospheric Physics, 11, pp. 11-18, 1998.
 

Ogawa, T., Nishitani, N., Pinnock, M., Sato, N., Yamagishi, H. and Yukimatu, A.S., Antarctic HF radar observations of irregularities associated with polar patches and auroral blobs : A case study., Journal of Geophysical Research, 103, No. A11, pp. 26547-26558, November 1998.

Parkinson, M.L., Breed, A.M., Dyson, P.L. and Morris, R.J., Application of the Dopplionogram to Doppler-sorted interferometry measurements of ionospheric velocity., Radio Science, 34, No. 4, pp. 899-912, 1999a.

Parkinson, M.L., Breed, A.M., Dyson, P.L. and Morris, R.J., Signatures of the ionospheric cusp in digital ionosonde measurements of plasma drift above Casey, Antarctica., Journal of Geophysical Research, 104, No. A10, p. 22,487, 1999b.

Reinisch, B.W., Bullett, T.W., Scali, J.L. and Haines, D.M., High Latitude Digisonde Measurements and their Relevance to IRI., Advances in Space Research, 16, No. 1, pp. 17-26, 1995.

Rodger, A.S., Pinnock, M., Dudeney, J.R., Baker, K.B. and Greenwald, R.A., A new mechanism for polar patch formation., Journal of Geophysical Research, 99, No. A4, pp. 6425-6436, 1994a.

Rodger, A.S., Pinnock, M., Dudeney, J.R., Waterman, J., de la Beaujardiere, O. and Baker, K.B., Simultaneous two hemisphere observations of the presence of polar patches in the nightside ionosphere., Annales Geophysicae, 12, pp. 642-648, 1994b.

Rodger, A.S. and Graham, A.C., Diurnal and seasonal occurrence of polar patches., Annales Geophysicae, 14, No. 5, pp. 533-537, 1996.

Rodger, A.S., Polar patches – Outstanding issues., Polar Cap Boundary Phenomena, Eds. J. Moen et al., Kluwer Academic Publishers, pp. 281-288, 1998.

Rosenberg, T.J., Wang, Z., Rodger, A.S., Dudeney, J.R. and Baker, K.B., Imaging riometer and HF radar measurements of drifting F-region electron density structures in the polar cap., Journal of Geophysical Research, 98, No. A5, pp. 7757-7764, 1993.

Scali, J., Reinisch, B., Dozois, C., Bibl, K., Kitrosser, D., Haines, M. and Bullett, T. ., Digisonde Drift Analysis., University of Massachusetts, Lowell, Center for Atmospheric Research, July 1995.

Sojka, J.J., Bowline, M.D., Schunk, R.W., Decker, D.T., Valladares, C.E., Sheehan, R., Anderson, D.N. and Heelis, R.A., Modelling polar cap F-Region patches using time varying convection., Geophysical Research Letters, 20, No. 17, pp. 1783-1786, 1993.

Sojka, J.J., Schunk, R.W., Bowline, M.D., and Crain, D.J., Ambiguity in identification of polar cap F-region patches: contrasting radio and optical observation techniques., Journal of Atmospheric and Solar-Terrestrial Physics, 59, No. 2 pp. 249-258, 1997.

Steele, D.P. and Cogger, L.L., Polar patches and the 'Tongue of ionization'., Radio Science, 31, No. 3, pp. 667-677, 1996.

Valladares, C.E., Basu, S., Buchau, J. And Friis-Christensen, E., Experimental evidence for the formation and entry of patches into the polar cap., Radio Science, 29, No. 1, pp. 167-194, 1994.

Valladares, C.E., Decker, D.T., Sheehan, R. and Anderson, D.N., Modelling the formation of polar cap patches using large plasma flows, Radio Science, 31, No. 3, pp. 573-593, 1996.

Valladares, C.E., Decker, D.T., Sheehan, R., Anderson, D.N., Bullett, T. and Reinisch, B.W., Formation of polar cap patches associated with north-to-south transitions of the interplanetary magnetic field., Journal of Geophysical Research, 103, No. A7, pp. 14657-14670, July 1998.

Wang, Z., Rosenberg, T.J., Stauning, P., Basu, S. and Crowley, G., Calculations of riometer absorption associated with F-Region plasma structures based on Sondre Stromfjord incoherent scatter radar observations., Radio Science, 29, No. 1, pp. 209-215, 1994.

Weber, E.J., Buchau, J., Moore, J.G., Sharber, J.R., Livingston, R.C., Winningham, J.D. and Reinisch, B.W., F layer ionization patches in the polar cap., Journal of Geophysical Research, 89, No. A3, pp. 1683-1694, 1984.

Weber, E.J., Klobuchar, J.A., Buchau, J., Carlson, H.C., Livingston, R.C., de la Beaujardiere, O., McCready, M., Moore, J.G. and Bishop, G.J., Polar cap F-Layer patches: Structure and dynamics., Journal of Geophysical Research, 91, No A11, pp. 12121-12129, 1986.

CONTENTS
Formating comments (E-mail: phil@ips.gov.au)
Date last altered: 2 Nov 99