Resolute Bay VHF radar: A multipurpose tool for studies of ...physics.usask.ca/~jean/hocking 2001 2000RS001005.pdf · Resolute Bay VHF radar: A multipurpose tool for studies of tropospheric

Resolute Bay VHF radar: A multipurpose tool for studiesof tropospheric motions, middle atmosphere dynamics,meteor physics, and ionospheric physics

W. K. Hocking,1 M. Kelley,2 R. Rogers,3 W. O. J. Brown,3,4

D. Moorcroft,1 and J.-P. St. Maurice1

Abstract. A VHF radar has been established at a site near Resolute Bay in Nunavut,Canada (75�N, 95�W), which has the capability to make a variety of measurements relatingto the atmospheric and ionospheric environment in the polar regions. The site is veryclose to the north geomagnetic pole, and therefore the radar is well situated to makesome unique measurements. The system is a multipurpose instrument with good remotecontrol capabilities. It can be used as a wind profiler radar to study the lower troposphere,as a mesospheric radar to study polar mesosphere summer echoes (PMSE) in summer, asa meteor radar to determine winds in the altitude region of 80–100 km, and as anionospheric radar to study 3 m scale irregularities in the E and F regions. The radar hassome unique design features, partly dictated by the rough terrain in which it is sited. Inthis paper, the radar system is described, including description of some unusualapproaches to deal with special conditions at the site, and then some key early results arepresented. Important findings include error determinations for tropospheric windmeasurements, detection of PMSE, correlations between PMSE and atmospherictemperatures at 86 km altitude, measurements of mean winds and tidal characteristicsover a full year, and detection of various normal modes of oscillation in the 80–100 kmregion, especially in nonsummer months. Some of these features will be discussed here,but more detailed discussions will be left to related papers in this issue.

1. IntroductionIn recent years it has become clear that the polar

regions of the Earth’s atmosphere hold keys to un-derstanding of a variety of important processes.These include aurorae, polar mesosphere summerechoes (PMSE), the ozone hole, and global warming.Because of these issues a renewed effort has beendeveloped to establish polar atmospheric observato-ries, and the so-called “Early Polar Cap Observatory”(EPCO) is one of these facilities. This site has beenestablished at Resolute Bay, in the Canadian territoryof Nunavut (75�N, 95�W), which is conveniently close

to the north magnetic pole. One of the instrumentsdeveloped as part of this exercise is a new VHFatmospheric radar, which is the subject of this article.

This radar is capable of tropospheric meteorologi-cal wind measurements, middle atmosphere studies,meteor studies, and ionospheric work and is operatedin a variety of modes. These different modes will bediscussed, and preliminary results from the radar willbe reported. These results include validation of thebeam-pointing function of the radar, demonstrationof its capability for meteorological studies in thetroposphere and lower stratosphere, as well as theresults of our first searches for PMSE. The radar wasbuilt with a modest budget of less than Ca$300,000(US$200,000), which included logistical support, alltechnical assistance, transportation, construction, andcommissioning. After construction the system has rununattended, with only limited technical support. De-spite a rather modest transmitter power output thesystem has produced some important results, whichwill be summarized in this article. The radar beganfull-time unattended operation on April 20, 1998,although it had been used in campaign mode on a fewoccasions prior to this.

1Department of Physics and Astronomy, University of WesternOntario, London, Ontario, Canada.

2School of Electrical Engineering, Cornell University, Ithaca,New York.

3Department of Atmospheric and Oceanic Sciences, McGillUniversity, Montreal, Quebec, Canada.

4Now at Atmospheric Technology Division, National Center forAtmospheric Research, Boulder, Colorado.

Copyright 2001 by the American Geophysical Union.

Paper number 2000RS001005.0048-6604/01/2000RS001005$11.00

Radio Science, Volume 36, Number 6, Pages 1839–1857, November–December 2001

1839

2. Radar DesignThe Resolute Bay radar was designed as an instru-

ment which could be used in multiple ways. The mainintended applications were wind profiling in the tro-posphere, mesosphere-troposphere (MT) dynamicalstudies, meteor studies, and E region investigations ofcoherent echoes. The list is not exhaustive, however,since the system had been designed to be sufficientlyflexible that other experiments can easily be designedand implemented. The radar therefore comprisesseveral antenna fields which can all be driven by thesame transmitter. The system is designed so that thesedifferent options may be selected under computercontrol, enabling remote operation. The radar issituated on bare ground which is free of any vegeta-tion, and the ground is composed mainly of smalljagged rocks with typical dimensions of a few centi-

meters across. Figure 1 shows two photographs of theantenna fields used with the radar. The first (Figure1a) shows a view through the antenna field which isused for mesospheric and tropospheric studies (theso-called MT component), and the second photo-graph (Figure 1b) shows four Yagi antennas mountedon towers and pointing toward the southwest. Thislatter set is used for ionospheric studies.

2.1. Antenna ArraysFigure 2a shows a plan view of the site. There are

four main antenna systems which are utilized in ourwork. The primary instrument is a large crossedstructure, indicated in Figure 2a as the “main array.”The second antenna field is a row of four-elementYagi antennas mounted on vertical towers and situ-ated to the northeast of the main array. The third is

Figure 1. Two photographs of the radar arrays. A view through the center of the main array is shownin Figure 1a, while Figure 1b shows the towers used for ionospheric E region studies. The towerssupport four 4-element Yagi antennas pointed toward the southwest, toward Cambridge Bay.

HOCKING ET AL.: RESOLUTE BAY RADAR1840

Figure 2. (a) Plan view of the radar site. (b) An estimate of the polar diagram of the meteor radar asdetermined by counting meteor echo returns over an extended period of time. The white dashed linesshow the transmitter polar diagram at the �12 dB level (inner four loops) and �24 dB levels (outerloop) as calculated from a numerical computer model. Note that the real count rate distribution will bebiased by the natural tendency of meteors to be detected less efficiently overhead.

HOCKING ET AL.: RESOLUTE BAY RADAR 1841

the “transmitter array” for the meteor radar, and thefourth antenna array is the meteor receiver array,which is, in fact, embedded in the main array. Thevarious radars cannot generally be used simulta-neously, but it is possible to choose the differentantennas under computer control and alternate easilybetween the options. Such selection can be maderemotely, via modem communications. The arrays areused in the following ways. The main array is used intransmit-receive mode, just like a traditional windprofiler radar. Power is transmitted into all of theantenna elements in the main array, including thereceiver antennas for the meteor array. Each group offour antennas (called a quartet) is fed by separatecable. Returned power is received by the same anten-nas, combined, and diverted to a receiver via atransmit-receive (TR) switch. Cables are incorpo-rated into the path lines of each antenna, permittingphase delays which point the beam into differentdirections. At present the radar can be pointed ver-tically and 10.9� off vertical along the four arms of themain array. These arms point in directions offsetazimuthally by 19� from true north, east, south, andwest, as indicated in Figure 2a. The one-way 3 dBbeam half width is 2.0�, and in two-way mode it is 1.4�.This narrow beam is very suitable for studies ofturbulence and wind measurements in the tropo-sphere and mesosphere. The tilt angle of 10.9� waschosen because it places the first minimum of thepolar diagram in a vertical direction when tiltedbeams are used. In meteor mode, power is transmit-ted from the “meteor transmitter (Tx) array” andreceived separately on the four antennas of the“meteor receiver (Rx) antennas.” Prior to June 2000,signal was multiplexed sequentially into the receiveron a pulse-to-pulse basis, first from one antenna andthen the next in a cyclic fashion. After June 2000 thesystem was upgraded to include five receivers, so thateach meteor receiving antenna was connected to aseparate receiver, allowing higher data acquisitionrates and therefore better signal-to-noise ratios. Datareceived are interrogated using interferometric tech-niques, as will be described in sections 2.4.3 and 2.5.

In “ionospheric mode,” signal is both transmittedand received from the four Yagi antennas on thetowers to the northeast of the main array, which arecombined to form a moderately narrow beam ofhorizontal half width of �8�, pointing toward thesouthwest. The half width in the vertical direction isnominally �40�.

We will now describe the arrays in more detail. A

block diagram showing the interplay between thevarious antennas is shown in Figure 3. The design ofthe EPCO VHF main radar antenna field followsclosely that of the University of Western OntarioVHF atmospheric radar (CLOVAR), as reported byHocking [1997]. As noted, the main array for MT(mesosphere-troposphere) work takes the form of across (Figure 2a), with basic antenna units beinggroups of 4-x-2-element Yagis at the corners of asquare with sides of length one half of a wavelength.These so-called “quartets” are then staggered withinthe arms of the basic cross in a manner identical tothat of the CLOVAR system [Hocking, 1997, Figure1, section 3.1]. The cross is 17 wavelengths from oneside to the other. In contrast to Hocking [1997], nospecial ground plane has been constructed for thisarray, and the reflector of the two-element Yagiantennas is relied upon to minimize radiation downinto the ground. Within the cross each quartet isindependently connected to the main transmit-re-ceive building using separate lengths of coaxial cable.These cables all have equal lengths (21 wavelengths),making interferometry experiments flexible and easy.The antennas near the center of the array are fed withlow-loss cable (Andrews 1/2” heliax), while thosetoward the outer sections are fed with higher-losscables (Belden 9913). Antennas in between these twoextremes are fed with a combination of the two cabletypes. Each cable comprises a section of heliax and asection of 9913 cable, organized in such a way that thepercentage of 9913 cable increases as the antennasbeing fed are farther and farther removed from thecenter. The electrical phase length of the cables is thesame in all cases. This variation in the relative lengthsof 9913 coaxial cable and half-inch heliax provides apower taper across the aperture. The cable loss to thecenter of the array is 1.75 dB, and the losses to theouter antennas are typically 3.5 dB. The mean lossacross the whole system is �3.0 dB.

The ionospheric towers are fairly standard. Theionospheric antennas are horizontally polarized Yagiantennas tilted at �15� above vertical, mounted atopof the towers, and their height above the ground is�1.75 wavelengths. No further discussion will beentertained about the construction of these.

We now move to a discussion about the meteorantennas. The receiver antennas will be consideredfirst. The coordinates for these antennas were chosento minimize ambiguities. As seen in Figure 2a, thereare three quartets quote close together, and then afourth displaced slightly to the southwest. Without


this last antenna quartet it would be impossible tounambiguously locate any meteors. With only threereceiving antennas there are often several possiblelocations of meteors in the sky which will give thesame phase differences between three antennas. Thiscan only be avoided by keeping the antenna spacingto be less than a half wavelength, and this introducesproblems with antenna coupling. If spacings largerthan a half wavelength are used (as we have), thenextra antennas are need to remove such ambiguities.The choice of the position of the fourth antenna set isimportant: Certain positions do not remove the am-biguities at all, while other locations are very effectiveat discriminating between ambiguous direction possi-bilities. The choice which we have used (as in Figure2a) is effective at removing ambiguities for zenith

angles of less than 60�. Ambiguities exist beyond thisangle, but the transmitter power is weak at these lowelevations. In addition, our software permits us todetermine if any meteor has potential problems withambiguities, and if such potential exists, these mete-ors are not used in subsequent analysis. Alternativechoices of antenna coordinates can be used: Forexample, Hocking et al. [2001] used five receivingantennas to permit much more complete removal ofambiguities.

The meteor transmitter antennas have a somewhatunusual design, which is a variation of the designpresented by Hocking and Thayaparan [1997]. Theantenna quartets are spaced apart by 1.5 wavelengthsdiagonally (1.06 wavelengths along and across thearray) and are fed with cables which are phased in

Figure 3. Flowchart showing the interactions between the Resolute Bay control computer and thevarious antenna and beam arrangements available with the system.


such a way that the phase difference between succes-sive quartets is 180�. The net result is that thetransmitter beam is divided into four main beamswhich are broadly pointing to the north, south, east,and west, with a zenithal tilt of typically 20�–30�. Thisarrangement allows the beams to point in the zenithaldirections at which meteors are most common andtherefore optimizes the system for wind measure-ments. While it was found that this arrangement wasnot optimum for the CLOVAR radar (in that case,because of cable losses to the antennas), the ResoluteBay antennas are fed directly from the transmitter viaAndrews 7/8” heliax cable, and therefore losses arevery small. Pulses of radio waves are transmitted fromthis array and are scattered by meteor trails present inthe atmosphere. Some of the signal returns to theground, which then excites the receiver antennas inthe meteor Rx antennas, thereby producing a signalfrom these antennas. The signals produced by thesefour antennas are received separately, and comparedby cross-correlative interferometric methods, in orderto determine the location of each meteor [e.g., Hock-ing and Thayaparan, 1997]. From knowledge of therange and angular position of the meteor trail itsposition may be precisely determined.

Figure 2b shows a density plot of the number ofmeteors received by the meteor radar over an entiresummer, as a function of angular position. The whitedashed lines show the expected positions of the maintransmit lobes, based on numerical modeling. Thecontours are shown at �12 dB and �24 dB levels.When comparing the model and experimental results,it needs to be recognized that meteors are detectedwith much higher efficiency when they occur at lowelevation angles, which will tend to push the locationsof the maxima out to larger zenith angles relative tothe predictions of the model. This is exactly what isseen in the figure; the maxima are generally in thecorrect location but occur at larger zenith angles thanthe prediction. It should be emphasized that althoughthe transmitter array produces these four broadbeams, this knowledge is not important in actuallydetermining the meteor locations. The interferomet-ric procedures which we use ensure that meteors canbe correctly located irrespective of the polar diagramof the transmitter beam pattern.

2.2. Ground EffectsOne interesting and important aspect of the posi-

tioning of the main array is that it is positioned on theside of a hill, which is necessary because of the choice

of site of the EPCO building. Because of the hugeexpense which would have been involved in levelingthe area we have chosen to simply install the antennafield without regard for antenna height and then goback and survey the antenna heights after this instal-lation. It was found that the total vertical distancebetween the lowest and highest antennas was 7.82 m.The height of each antenna relative to a suitablereference was then determined to an accuracy of �2cm. Following this, phasing cables have been cut tocompensate for the differences in antenna heights.For example, cables higher up the hill have been fedwith extra cable, in order to delay the signal reachingthem, while antennas farther down the hill have beenfed with shorter cables. This process ensures thatwhen the array is fed in its “vertical beam mode,” ahorizontal plane wave is formed above the arraywhich moves vertically upward. Thus the array be-haves as if it is electronically flat. On the basis of oursurvey we estimate that the maximum deviations ofthis array from being electronically flat are of theorder of 2–4 cm and the main beam is vertical tobetter than �0.2�. This procedure also ensures thatthe radar properly corrects for the sloping groundalignment upon reception as well. It also properlycorrects for the land alignment when the radar is usedto transmit in nonvertical directions, as, for example,is the case when the main beam is pointed at 10.9� offvertical. It is important to note, however, that thebeam should not be pointed further than �15� offvertical, since at larger angles the correction effects ofthese extra cables introduce second-order effectswhich begin to distort the beam. It should also benoted that the four quartets which are used formeteor reception (i.e., the meteor Rx antennas inFigure 2a) were adjusted in height so that they wereall in the same horizontal plane, since these were tobe used to measure meteors out to 70� from offvertical.

Figure 4 shows plots of the mean powers receivedas a function of height and time with the radar,showing that we get reasonable power levels returnedwhen using the system in Doppler mode. However,further tests are needed to ensure that the radar isbehaving properly. One of the most convincing wayshas been by comparing radar measurements of tro-pospheric winds to those determined from a nearbyradiosonde. Detailed results are presented by Hock-ing [this issue (a)]. Radial velocities measured onbeams aligned to the north and south showed mirrorsymmetry, as did measurements on beams to the east


and west. Temporal variations matched the radio-sondes very well, and magnitudes showed good agree-ment, except for an expected effect due to the scat-terer anisotropy [e.g., see Hocking et al., 1990; Tsudaet al., 1986; Hooper and Thomas, 1995]. An example isshown in Figure 5 of this paper, and this will be

discussed in section 2.5. In addition, we have moni-tored the noise levels recorded with the narrowbeams. This is dominantly galactic noise, as has beenverified by disconnecting the antennas from the re-ceiver; a decrease in noise level of well over 10 dB isclearly evident. The galactic noise shows a temporal

Figure 4. Contour plots of the power as a function of height and time recorded in the frequency bandbetween �0.8 and 0.8 Hz, for (top) a vertical beam and (bottom) a beam pointed at 10.9� off zenith tothe “west.” Values where the contour shading is at its lowest level generally indicate regions where nodata were saved to disk (see text for details).

Figure 5. (a) Scatterplots of radar-derived wind directions compared to radiosonde-derived directions.The directions shown are the directions from which the wind has come, measured clockwise from truenorth. (b) Scatterplots of radar-derived wind magnitudes compared to radiosonde-derived magnitudes.In each case the parameters g0x and g0y represent least squares fits of one variable against the other,where g0x refers to the regression of y on x (i.e., assumes zero error in the radiosonde data) and g0yrefers to the regression of x on y (i.e., assumes zero error in the radar data). In both cases there were612 points.


variation over the course of the day, but the variationis less than 1.5 dB, and there are no major VHF radiosources which pass through the beam. Furthermore,the noise level shows a variation of no more than 0.5dB between beams, indicating that the beams all havesimilar gain characteristics. The noise is strongest onthe east and southward looking beams, which may beeither a geophysical effect or simply due to variationsin beam characteristics, but either way, the differ-ences are very modest. We are therefore confidentthat the beams are being properly formed, despite thefact that the array is not physically flat, and all beamshave similar gains and efficiencies.

2.3. ElectronicsWe now turn to discussion of the transmitter and

receiver systems. Table 1 shows a brief tabular sum-mary of the main features of the new radar. Thereceiving and digital acquisition system (RDAS) wasmanufactured by Genesis Software of Australia, andsupplied through ATRAD (Australia). The transmit-ter was partly built in-house, with supply of the poweramplifier modules, combiners, and transmit-receiveswitches by Tomco (Australia) and ATRAD.

The whole system is controlled by a personalcomputer operating under FreeBSD UNIX and has alarge number of options in regard to selection ofpulse repetition frequencies, pulse codes, beam point-ing, and system modes. Signal processing is similar inmany ways to the procedures outlined by Hocking[1997] and will be described in more detail in section2.5. The transmitter is a solid state system producing

12 kW of peak power, with a maximum duty cycle of8%. It is most commonly run at �5% duty cycle.Losses through the cables to the antennas (as de-scribed in section 2.1), and the TR switch and otherrelays, mean that the radiated power is only �4–5 kWpeak. The transmitter is capable of pulse lengths ofbetween 150 m and several kilometers, and pulsecoding is available. In addition, there are also variousother modes not normally available with standardmesosphere-stratosphere-troposphere (MST) radars,including the availability of pulse-pair experimentsfor ionospheric experiments. While Table 1 summa-rizes the system features, we will now add to thisinformation with a more detailed discussion of thesystem logistics.

2.4. Operations LogisticsFigure 3 shows a flowchart demonstrating the

interactions between the various subelements of thesystem. The different antenna arrays which can beselected are shown. The operation of the system iscarried out using a mixture of routine programs andspecial campaign experiments. Some of these modeswill now be described.2.4.1. Tropospheric mode. The tropospheric modesare some of the most commonly used modes. Pulsesare transmitted and received through the main beamof the array, formed by phasing all of the antennas ofthe main cross together. The beam can be steered toseveral directions, and in the most common arrange-ment the beam is steered successively from vertical to10.9� off zenith to the “north,” then “east,” then

Table 1. System Parameters for the EPCO VHF radar

Parameter Value/Description

Modes of operation (1) mesosphere-troposphere (MT) (up to nine beam directions can be selected undercomputer control) (2) meteor studies (3) ionospheric studies, and (4) specializedinterferometric studies (modes 1–3 can be selected under computer control over aphone line)

Frequency 51.5 MHzTransmitter peak power 12 kWMaximum duty cycle 8%Typical duty cycle 5%Peak power-aperture product 2.4 � 107 Wm�2

Typical mean power-aperture product 1.2 � 106 Wm�2

Pulse repetition frequency 16–16000 HzPulse codes various codes selected under computer control

variable: 150 m to several kilometersPulse lengthMain antenna array cross structure; width equals 100 m, and collecting area is approximately 1500 m2

Beam half-power half width (main array) one way: 2�Beam steerability nine beam directions available under computer control (vertical, north, east, south, and

west, plus four yet to be added)Off-vertical beam directions 10.9� off zenithPurpose of additional antennas (1) meteor studies and (2) coherent E region and backscatter studies


“south,” then “west.” In reality, these beams arerotated by 19� anticlockwise from true north, south,east, and west, and are aligned along the arms of thearray, as shown in Figure 2a. However, we will referto them as “north,” “east,” “south,” and “west,” withthe quotations indicating that they are not true geo-graphic north, south, east, and west.

Although pulse coding is possible, we have foundthat the highest data rate is achieved by transmittinga monopulse of 5 �s duration (750 m effective length)at a pulse repetition frequency (PRF) of 10,344 Hz.The pulse is smoothed and tapered on the ends inorder to minimize the production of higher-orderfrequency harmonics. The use of such a high PRFproduces serious range aliasing (aliasing range of 14.5km), but because of the low power of the system,almost all of our tropospheric echoes usually comefrom below 8 km. The only measurable echoes fromabove 8 km are the polar mesosphere summer echoes(PMSE) in summer, and this choice of PRF forcesthese echoes to be seen at an effective range of above8 km, so that they do not contaminate the tropo-spheric data. Data are sampled at 750 m steps. Othermodes which utilize pulse coding are available but arenot often used. This is because the use of a mono-pulse of effective length 750 m, at a PRF of 10,344Hz, results in a duty cycle of 5%, and this is as high asthe system duty cycle is allowed to reach in unat-tended mode. Other tropospheric modes exist, how-ever, including those using various types of pulsecoding and higher resolution.2.4.2. PMSE mode. Polar mesosphere summer ech-oes (PMSE) have been an important area of study forover 10 years, because they clearly reveal some veryunusual physics which is operative in the polar meso-sphere at 80–90 km altitude [e.g., Kelley et al., 1987;Czechowsky et al., 1988; Cho and Kelley, 1993; Rottger,1994; Cho et al., 1996]. Therefore we have developeda special mode which is used to search for andinterrogate these echoes. In this mode, just as for thetropospheric mode, radio waves are transmitted from,and received by, the main beam of the array, which issuccessively pointed to the vertical, then 10.9� offvertical to the north, then east, south, and west. APRF of 1200 Hz is used, together with an 8 bitcomplementary code. The effective pulse resolution is750 m. However, other modes exist for special exper-iments, including modes which use 10 bit complemen-tary codes, and different PRFs. Normally, coherentintegration is performed over 16 successive points, so

that the temporal resolution is 0.013 s, and thespectral folding frequency is 37.5 Hz.

In this PMSE mode we have also chosen a some-what unusual filter for the receiver. Traditionally, afilter width of �200 kHz should be used for a pulselength of 5 �s. However, through both experimentand theoretical determinations we have determinedthat a filter full width of 125 kHz produces a bettersignal-to-noise response than does the wider one. It istrue that the effective pulse width is increased, andsome of the returned signal is lost (with a reduction of�1.8 dB in peak power and a loss of 0.9 dB in powerwhen integrated across the pulse), but the noisereduction is even greater than the signal loss, and thesignal-to-noise ratio is optimized. The pulse length ofthe received signal increases to �6.1 �s after passingthrough this filter (effective pulse length of 920 m).2.4.3. Meteor mode. This is one of the most com-monly used modes. It involves running the systemusing a 5 bit Barker code and a PRF of 750 Hz withtwo-point coherent integration. An all-sky approachis used, transmitting simultaneously into four moder-ately broad beams (as described in section 2.1) di-rected approximately to the north, south, east, andwest. The backscattered signal is then recorded byfour separate antennas strategically placed near thecenter of the array, as noted (Figure 2b). The signalsfrom the receiving antennas are then phase-comparedvia interferometric algorithms to determine the loca-tion of any meteor in the sky. At the same time,meteor decay times and radial velocities are deter-mined, thereby allowing these data to be used collec-tively at a later stage to determine wind vectors andtemperatures at these mesospheric heights.2.4.4. Ionospheric mode. In this case, pulses of radiowaves are transmitted from four 4-element Yagiantennas aligned at 15� above horizontal, mountedatop four large “ionospheric towers,” and pointingtoward the southwest (Figure 1b). These towers arealso used for reception. A variety of pulse schemesand PRFs are available, including 10 km monopulses,pulse pairs, and pulse codes of various types, and thePRFs vary from 95 Hz to 4800 Hz. Unlike the othermodes, raw data are streamed directly to disk, so thisfills the disk in a few hours.2.4.5. Common modes. Often programs are devel-oped which involve interleaving of the various exper-iments in sections 2.4.1–2.4.4. The most commonusually runs in meteor mode for 5 min, then runs intropospheric mode for �1 min with one choice ofbeam, then returns to meteor mode for 5 min, then


returns to tropospheric mode with a different beamdirection, and so forth. Other combinations includeoptions which alternate between PMSE mode andmeteor mode, PMSE and tropospheric modes, and soforth. The time required to change the beam config-uration is of the order of milliseconds.

When using such modes, care needs to be taken indata interpretation. The beam only completes a fullcycle of four off-vertical beams and one vertical beamevery 30 min, and there can clearly be temporalvariations of the wind field in that time. Thereforethese data are best used to provide averages over timeframes of 1–2 hours and more. The vertical velocity ismeasured but cannot normally be used to help definethe true horizontal winds because of its spatial andtemporal separation from the off-vertical measure-ments. It is more normal to assume that the verticalwind is zero, which is normally a suitable approxima-tion over periods of an hour or more.

2.5. Signal ProcessingThe signal processing employed for the tropo-

spheric and PMSE modes with this system followsclosely that described by Hocking [1997]. Data arerecorded to a buffer for typically 40 s and thentransferred to computer disk for subsequent analysis.This subsequent analysis takes place at the same timethat a new batch of data are being recorded, so thatrecording and analysis continue in parallel. Someon-line coherent integration takes place, but it is keptto small numbers (in most cases, 64-point integrationsare used), so that the temporal resolution whentransmitting at 10,344 Hz PRF is 0.0062 s.

In the analysis phase, spectra of the received signalare first calculated from the raw data. The spectra arethen stored to file for subsequent analysis. The fold-ing frequency for the spectrum is 80.8 Hz, so thespectra are only calculated between �80.8 and 80.8Hz. In the second stage of the analysis the centralregion between about �2.6 and 2.6 Hz is diagnosed,with the program searching for spectral peaks in thisband. These limits of �2.6 and 2.6 Hz are less thanthe limits used for the CLOVAR radar (where �4and 4 Hz are used), and this is necessitated by theexistence of frequent, persistent auroral interference.By reducing the allowable bandwidth we improve thelikelihood of rejecting auroral echoes. This does havethe negative effect of limiting the maximum usefulwinds which we can measure, and the radar cannotmeasure winds if the horizontal component along thebeam exceeds 32 m s�1. However, studies of the

winds expected at Resolute Bay from earlier radio-sonde studies over several years have shown thatlarger winds are moderately rare, at least below �6km altitude.

The spectra are analyzed on-line in a variety ofways. In the simplest calculations the integratedpower in various spectral frequency bands is deter-mined. The power in the bands �2.6 to 2.6 Hz, �2 to2 Hz, �0.8 to 0.8 Hz, and �0.08 to 0.08 Hz is foundand stored. Figures 4a and 4b show examples of theintegrated power in the band �0.8 to 0.8 Hz, for datarecorded in the period August 20–24, 1999, for botha vertical and an off-vertical beam. In order topreserve disk capacity, not all data are stored. Onlydata which show some evidence of containing aspectral peak somewhere in the band between �2.6and 2.6 Hz are stored to disk. (This procedurereduces disk storage requirements by typically a fac-tor of 30–50. An option does exist to store all powersto disk, and this is often used for PMSE studies, andfor special occasions, but as a rule this more space-efficient strategy is adopted.) Hence it should benoted that data in Figures 4a and 4b which are at thelowest contour level generally correspond to caseswhere no data were stored to disk. The apparentcutoff in signal at 30 dB (digital storage units) is nota true indicator of the maximum height at whichsignal above the noise level can be received, but itdoes represent a limit above which no useful windscan be extracted. It will be seen that useful winds canbe obtained to a maximum height between typically 3and 7 km. The lowest height at which useful windinformation can be obtained is 1.75 km. The upperlimit depends on the atmospheric backscatter crosssections, which depend, in turn, on the atmospherichumidity gradients and strengths of turbulence [e.g.,Hocking and Mu, 1997; Tsuda et al., 1988]. In thearctic, humidities are often very low, which is why theupper radar echo height limit is low. The greatestheights at which useful echoes can be obtained is afunction of season, being generally more in summerand less in winter (when humidity is low). Typically,the maximum heights recorded on the vertical beamin summer are 5–7 km, while the smallest values forthe maximum height occur in January and Februaryand are of the order of 3–4 km. For off-vertical beamsthe summer maxima are at �4–5 km, and the wintermaxima are at �3–4 km. These results are presentedin greater detail by Hocking [this issue (a)].

The next stage of analysis is to perform a moredetailed study of the spectra. As also described by


Hocking [1997], the spectra are first examined in thewings, where they are searched for spectral peakscorresponding to atmospheric backscatter. If no no-ticeable peaks are found in these outer regions (be-yond �0.8 and 0.8 Hz), the central part of thespectrum (region between �0.8 and 0.8 Hz) is inves-tigated. Spectral searches are then done here, butonly after special procedures are applied to removethe near-DC components of the spectrum. This isdone by using polynomial fitting in the time domainand spectral notch filtering in the frequency domain[Hocking, 1997]. Once dominant spectral peaks areidentified in any region, information about them,including their peak powers and spectral widths, arestored to disk. These data can then be used forfurther subsequent analysis at a later time. Datastored include weighted moments around the spectralpeak and parameters which result from fitting aGaussian function to the pertinent spectral region.

The system has one weakness in regard to slowlyoscillating components. In contrast to the CLOVARradar, the transmitter at Resolute Bay does not runcontinuously, but is turned off while data are trans-ferred to the digitizer, which typically happens aboutonce per minute, and takes typically 10–20 s. As aresult, the transmitter modules cool during this pe-riod. Then, when the transmitter once again turns on,it starts to heat up. Although the exact mechanism isnot clear, the result is that the signal received justafter the main pulse is transmitted tends to oscillatein time with a period of a few seconds, resulting in amodulation of the signal at close ranges. The effectinvariably diminishes within �10 s of the transmitterturning on, but it is a nuisance. To partly bypass thisproblem, data are only analyzed for times at least 15 safter the transmitter turns on, but even then there areslow drifts which forbid the calculation of verticalvelocities, and small radial velocities. As a result theradar is not capable of detecting radial velocities ofless than �0.4 m s�1 (frequency offset of 0.13 Hz).Thus the system cannot measure winds well if thewind is perpendicular to one of the radar beams andcannot reliably measure vertical winds. Future up-grade plans will address this issue.

Because the radar uses four off-vertical beams, thedata can be further checked after data collection. Thisprocedure is especially important for removal ofauroral echoes. Data recorded on off-vertical beamswhich are aligned at 180� relative to each other shouldshow mirror symmetry relative to each other, and ifone beam determines a radial velocity of, say, 1.6 m

s�1, then the other beam should record �1.6 m s�1

[e.g., see Hocking, 1997, Figure 11]. This test, togetherwith outlier rejection [see Hocking, 1997, Figure 9](please note that the scaling of the ordinate in Figure9 of that paper is in error, and should vary from �6 to�4), is used very successfully to remove auroralechoes. Auroral echoes tend to come in from direc-tions which are not along the main beams, but fromdiscrete directions through the sidelobes. Thus au-roral echoes tend to have the same sign for the radialvelocities when recorded using either beam direction,in contrast to atmospheric echoes, which typicallyshow opposite signs for the radial velocities in the twooppositely directed beams. Periods where the twobeams show the same radial velocities for extendedperiods are rejected as times of auroral interference.Application of quality control filters of this typetherefore help ensure that the data truly representatmospheric wind motions.

Despite these various interference effects, and sys-tem limitations, the radar does produce reliable windmeasurements. A radiosonde site is situated only 4km from the radar, and comparisons between theradar winds and the radiosonde winds have beenmade over a full year. Figure 5 shows scatterplots ofthe 3-hourly radar winds compared to the radiosondewinds, using data from the height range 1.75–3.25 km(612 points). The wind directions are generally ingood agreement (regression coefficient 0.975), as arethe magnitudes (regression coefficient 0.857). Hock-ing [this issue (a)] has performed a more extensivecomparison between the radar and radiosonde windsand has concluded that any discrepancies in the plotsin Figures 5a and 5b are most often due to thetemporal and spatial separation of the sondes and theradar, which can arise because of the fact that thesondes drift with the wind. That same paper alsostudies the aspect sensitivity of the radio wave scat-terers in some detail.

As an additional point, it will be noted that thereseem to be fewer occurrences in Figure 5a of windswith orientations of 70�, 160�, 250�, and 340�. This is,in fact, an artifact and arises because these happen tobe occasions when the wind was blowing parallel toone of the arms of the array. Thus one pair of theradar beams will be orientated perpendicular to thewind and therefore measure very small speeds. Theinability of the radar to measure radial velocities ofless than 0.4 m s�1 has already been noted, and thislimitation means that when the winds are alignedalong one arm of the radar, the beams perpendicular


to this alignment will record no useful spectra. Thuswinds orientated 19� anticlockwise from true north,east, south, and west will be deficient in informationabout the wind speeds in one of the orthogonaldirections, meaning that it will not be possible todetermine a wind speed and direction in this case.

For meteor mode the processing follows Hockingand Thayaparan [1997] in relation to wind determi-nations. Meteors are located via interferometry andrange determination, and then the winds are foundusing an all-sky fit to the radial velocities withintypically 2-hour periods. Temperatures can also bededuced by utilizing the meteor lifetimes as deter-mined by the radar, and as described by Hocking[1999]. Lines of best fit are determined from graphs oflog (inverse decay times) versus height, and the slopesof these lines may then be used to estimate thetemperature at the height of maximum meteor countrates, which is typically 88 km in summer and 86 kmin winter at this site for this radar frequency.

3. Some Specific Recent ObservationsWhile many of the more important observations

will be reported in other articles in this issue, it ispertinent here to summarize some key recent find-ings. These cases demonstrate the system capabilitiesand also help to define future experiments for thesystem.

3.1. Polar Mesosphere Summer EchoesOne important reason for the construction of this

radar was for studies of polar mesosphere summerechoes (PMSE). These echoes have intrinsic interestsbecause of their unusual nature but are also interest-ing as a possible indicator of global warming. There-fore we consider that an important aspect of ourresearch is to monitor the strengths and frequenciesof occurrence of PMSE over the lifetime of the radar.The fact that it was possible to detect these scattererswith the radar was an important point, and since theearliest detections in 1997 we have consistently mon-itored their occurrence. Data are recorded on all fiveradar beams on all of these occasions.

Figure 6 shows one of the strongest occurrences ofthese echoes, on June 18, 1998. The echoes are shownfor all five beams. At the height of the PMSE layersthe beam and a horizontal surface intersect in an(almost) circular region of approximate radius 2 km(to half power), so the received power is an integralover this region and over a depth equal to the pulselength. The off-vertical intersection regions are sep-

arated by �16 km from that of the vertical beam. It isof interest to compare the structures on the variousbeams. It is often true that the powers received on theoff-vertical beams are less than those for the verticalbeam, and this is clearly true here. However, it is alsoclear that some of the differences between the beamsare due to spatial variability. For example, the burstof activity at 86–90 km from 0930 to 1100 UT isclearly strongest on the vertical beam, but is also quitestrong on the westward beam. It is much weaker onthe north, east, and west beams. The burst of activityin the same time period at a lower height (82–85 km)is clearly very strong on the vertical beam, but almostnonexistent on all the off-vertical beams, indicatingstrong aspect sensitivity in this case. Thus variabilitybetween the beams is clearly a function of bothdegree of vertical anisotropy of the scatterers [e.g.,Czechowsky et al., 1988] and spatial variability. Theability to diagnose these echoes on five such beamssimultaneously will be an important capability of thisradar as our studies proceed. Here, we simply wish toemphasize the complex interplay between spatialvariability and aspect sensitivity.

The data shown for June 18 are somewhat atypical,in that the echoes persisted for most of the day. Morecommonly, the echoes are most prevalent between1200 and 0000 UT. Examples are shown in Figure 7.More detailed discussions about the times of occur-rence are given by Huaman et al. [this issue].

3.2. Daily Temperature VariationsThe tendency for PMSE to occur preferentially

between 1200 and 0000 UT (0600 and 1800 local time(LT)) requires an explanation. We have tested severalpossibilities, but one part of the explanation appearsto lie with the daily temperatures. Hocking [1999] hasshown how meteor decay times may be used in acollective manner to infer temperatures at the heightof peak meteor count rates, as a function of month ofthe year. We have now applied the same procedure ina slightly different manner. We have grouped allmeteors for the months of June, July, and August, forboth 1998 and 1999, into 12 bins, according to time ofday. The first bin includes all meteors which occurredbetween 0000:00 and 0159:59 UT, the second binincludes meteors which occurred between 0200:00and 0359:59 UT, etc. Then, for each of these bins wehave determined a mean “typical” temperature forthe height region between �85 and 92 km (since theheight of peak meteor count rates at a frequency of51.5 MHz in summer is �88 km). In this way, we can


deduce the “typical” diurnal variation of tempera-tures during the summer. As noted by Hocking [1999],there is some sensitivity in our determinations to themean temperature gradient, but during summer thetides at Resolute Bay tend to have very long wave-lengths, so we do not expect tidal variability of themean gradient to affect our determinations. Evidencethat the vertical wavelengths are long at these lati-tudes (greater than 100 km, and often the waves areevanescent) can be shown using meteor wind mea-

surements at Resolute Bay [Hocking, this issue (b)],by previous experimental evidence using MF radars[Manson et al., 1999], and by modeling studies involv-ing the global scale wave model [e.g., see Hagan et al.,1999; Manson et al., 1999, and references therein].

The lower graph in Figure 7 shows the results ofour temperature determinations. As additional evi-dence, we have repeated the process using only datafor the summer of 1998, in this case using 4 hour databins, and these results are also shown in Figure 7. In

Figure 6. Contour plots of the backscattered signal strengths (digital units) for all five beams on June18, 1998. The scalings are the same for all graphs.


both cases, there appears to be a minimum in tem-perature between 1500 and 0000 UT, which broadlymatches the period of increased PMSE activity. Wetherefore surmise that the colder temperatures at thistime of day play an important role in contributing toPMSE occurrence.

There is one possible further complicating factorwhich might affect these determinations. If the mete-

ors form in the PMSE layers, where ion mobility issupposed to be very low, it is possible that thediffusion coefficients measured by the meteor radarmight be anomalously small, simply because of thepresence of the PMSE themselves. However, we haveinvestigated this possibility and feel that it is notaffecting our measurements. First, we examined theheight region around 85 km, where the PMSE are

Figure 7. The upper three graphs show height-time diagrams of recorded power for 3 typical days. Thetendency for PMSE to occur in the afternoon sector (UT) is demonstrated. The lower graph showscomposite-day time variations of the temperature for summer during 1998 (shaded squares) and1998–1999 (solid circles).


most common, for anomalous meteor decay times,and none were found. Second, because measurementssuggest that the Schmidt number should be of theorder of 100 or more [e.g., Cho et al., 1996], this wouldcause any diffusion times to increase by a factor of100. Times which might normally be 0.1 s wouldbecome 10 s and also would be distorted in this timeframe by the presence of turbulence and wind shear.The radar software is very careful to select onlyunderdense meteors, with timescales of less than 2 s,so such long-lived trails, if indeed they exist, would befiltered out by the software. We therefore have con-siderable confidence that temperatures are indeedcolder in the period described.

3.3. Wind MeasurementsAs noted, the radar is also capable of wind mea-

surements in the region between 80 and 98 kmaltitude. This capability is important not only forstudies of the general dynamical field over ResoluteBay but also for support for other special studies likePMSE investigations. We have therefore ensured thatthe meteor radar mode operates during most of theyear, interleaved between tropospheric modes. Onoccasions we have performed dedicated PMSE stud-ies, during which the meteor radar mode has beenturned off, but as a rule the meteor mode is keptoperative. As a result, we have an acceptance rate of2-hourly winds at 86 km altitude of 86% throughoutthe entire year. This permits us to study dynamicalmotions throughout all seasons. In the early period ofoperation of the radar the acceptance rates weretypically 600 or more meteors per day in summer andwere somewhat lower in winter, when the ratedropped to �300 per day. A recent upgrade to thesystem has included the addition of four extra receiv-ers, so that each meteor antenna can feed signalseparately into a different receiver. In addition,sources of external noise have been located andremoved, and since June 2000 the more commonmeteor count rates are of the order of 1500–2500meteors per day. Dynamical studies of wind motionsare reported by Hocking [this issue (b)], but we willdiscuss just a few examples here.

During a period of strong PMSE in July 2000 thesignal-to-noise ratio was of sufficient quality that itwas possible to produce reliable measurements ofradial velocities using the narrow beam mode of thearray. At the same time the meteor system continuedto record at periods interlaced between the Dopplerrecords. Thus near-simultaneous winds were available

by two quite different techniques. Figure 8 shows theresults of a comparison between the two data sets.The PMSE data were averaged over two separateone-week periods, and the meteor winds were deter-mined over the full two-week period. The PMSEwinds were recorded at a resolution of 750 m, as seenin Figure 8, but we have also averaged the PMSEwinds to produce averages over a vertical extent of �3km, in order to better compare Doppler winds to themeteor ones. The PMSE 3 km averages are shown bythe stars in Figure 8, and the meteor winds are shownby the triangles. Agreement between the two methodsis quite good. Only one point showed a large differ-ence between the two techniques, and that is thevalue of the zonal wind at 85 km altitude. Even herethe winds are in the same direction, and at least partof the difference can be ascribed to the fact that thePMSE winds are biased to times of day when thePMSE echoes are strongest, while the meteor windsare biased more to early morning when meteor countrates are stronger. The other data points show goodagreement, particularly in the meridional compo-nents.

The above comparison is important in that itdemonstrates that both Doppler methods and meteormethods are valid procedures to use to determinewinds at heights below 90 km at Resolute Bay. This isan important result, because it disproves speculationthat electric fields at polar latitudes could bias meteorwind measurements (analagously to the proposal thatelectric fields at the equator could affect meteor windmeasurements there [Oppenheim et al., 2000]).

Figure 9 gives an example of the types of resultsproduced by spectral studies. In this case, the zonalmeteor winds have been ascribed to a “real” compo-nent, and the meridional winds have been ascribed toa “quadrature (imaginary)” component, and thenthese complex vectors have been Fourier trans-formed. In the example shown, 20 day segments havebeen used: one from January and one from July. Theresultant so-called rotary spectrum shows severalimportant peaks, including tidal oscillations, plane-tary-scale oscillations, and possibly some free modes.It is also evident that planetary wave activity is muchhigher in winter than in summer. Tidal oscillationstend in the main to rotate clockwise with time. Moredetails are given by Hocking [this issue (b)], whoconsider seasonal variability of tidal amplitudes andphases, planetary waves, and free modes of oscillationin more detail.


4. Astronomical StudiesBecause the system detects meteors with good

efficiency, the radar is also suited to some studies ofan astronomical nature. Studies of this type havetaken on new prominence in recent years, with in-creased concern about damage to sensitive spacecraftelectronics by meteor “hits.” Optical methods arerestricted to nighttime conditions only, which ex-cludes most of the summer at Resolute Bay. Radiostudies of meteors were severely downgraded duringthe 1970s and 1980s, and new stations now need to bebuilt. The radar at Resolute Bay is one of only a veryfew which can perform astronomical studies of me-teor streams at high northern latitudes. This uniqueposition means that many of the streams seen fromResolute Bay are circumpolar and so can be observedfor 24 hours per day. The fact that the radar records

meteor information continuously throughout theyear, at this unique site, gives it an important role toplay in future meteor astronomical studies. Meteorstudies also help to verify the general functionality ofthe system.

One especially useful function which the radar canperform is determination of the coordinates of theradiants of meteor streams. This procedure relies onthe fact that meteors can generally only be detectedby radar if they are aligned perpendicular to thevector which points from the radar to the meteor. Asa result, it is possible to define a “great circle” in thesky, centered around the radar, which defines a subsetof possible directions from which the meteor couldhave come. The exact location of the meteor radiantat this point in time is unknown, except that it liesupon this great circle. However, when this process is

Figure 8. Comparison of winds measured by Doppler radar mode and meteor mode during a periodof strong PMSE activity in July 2000. (a) Zonal winds. (b) Meridional winds.


repeated for many meteors, and the great circles areadjusted in position to compensate for the rotation ofthe Earth as time passes, then many of these greatcircles start to overlap at a common point. This pointtherefore defines the radiant, assuming that a signif-icant number of the meteors were produced by acommon shower. This process has been tested inmany cases and seems very good at determiningsource radiants on those occasions when showersexist. It is a variant on the method proposed by Jones

and Morton [1982]. An example of the results pro-duced from application of this process is shown inFigure 10, where the procedure has been applied fordata from December 12 to 13, 1998. A strong peak ata right ascension (RA) of 7 hours and 30 min, and adeclination of 32�, is apparent, and this is due to theGeminid meteor shower. According to Norton [1973]the correct radiant is at RA � 7 hours and 28 min,and declination equals 32�.

The ability to perform source mapping in this way

Figure 9. Sample 20 day rotary spectra of winds for typical summer (solid line) and winter (shading)months. Positions of the tidal harmonics (if present) are also indicated, as well as the position of the 2day wave. Dashed lines on the arrows indicate that these tidal modes do not appear to exist in this figure,while solid lines indicate that there is some evidence for a peak. A possible 5 day wave may also bepresent, but is not highlighted.

Figure 10. Results of meteor shower source determinations for December 12–13, 1998. The Geminidsare clearly located.


is important for several reasons. First, it serves as avery rigorous check on the system alignment andphasing. We have already confirmed that the system iscorrectly orientated by comparing our troposphericwinds with radiosonde winds, and these meteor radi-ant determinations serve as further confirmation ofcorrect alignment. The fact that we can locate thisshower source accurately also indicates that there areno phase errors between the various meteor receiverantennas and that our meteors are being correctlylocated. This therefore serves as a check that ourmeteor winds will not have any biases due to meteorlocation errors. Other radiant sources have beenconfirmed at other times of the year, including theQuadrantids in January, the Zeta Perseids in June,and the Daytime Sextantids in September. Thesedeterminations demonstrate that the meteor radar isfunctioning correctly all year round.

The ability to determine sources is also importantfrom an astronomical point of view, especially whennew streams are detected, as happens from time totime. An example of the importance of this processwith another similar radar was reported by Arlt et al.[1999], who showed how a meteor radar was able todetermine the source of an outburst of the JuneBootid meteor shower. By also deducing the meteorentrance speeds (not currently available with theResolute Bay radar) those authors were able todetermine the orbit of the stream and thereby asso-ciate it with the comet 7P/Pons-Winnecke. The abilityof the Resolute Bay radar to determine orbits in thisway can play a useful role in other future meteorshower studies. This is especially true because of thehigh latitude of the site, which permits it to observemany of the northern streams for 24 hours per day.

5. ConclusionsThe Resolute Bay VHF radar has been described.

Unique features about the radar have been high-lighted, including its flexibility for use in differentareas of research. The radar may be used for studiesof meteors, atmospheric winds and temperatures,tropospheric dynamics, and ionospheric physics. Theneed to situate the radar on ground which had aconsiderable slope has been discussed, and the cor-rection procedures which have been used to compen-sate for this effect have been described. The applica-tion of phase corrections in this way to compensatefor the sloping ground is very new in the area of VHFwind profiler technology. Sample results from the

radar, including important new results, have beenpresented. Key studies have included a detectedcorrelation between diurnal temperature variabilityand the occurrence of PMSE, comparisons of windsby both meteor and Doppler methods (and demon-stration of the value of using interleaved Doppler andmeteor modes of operation), and studies of thespatial distribution of PMSE using multiple beamdirections. Another important result which has beenillustrated is an increased planetary wave activity inwinter relative to summer at Resolute Bay.

Acknowledgments. The support of Wayne Ketchabaw,Peter Duenk, Harry Chen, Steven Nangae, and BabuhKalluk during the construction phase of the arrays is verymuch appreciated. Bruce Johnson is acknowledged for hisspecial efforts in regard to development of the transmitter.Support from the staff of Narwhal is also acknowledged.We are also grateful to SRI International, and especiallyJohn Kelly and Joe Cyr, for their support of the radarduring the construction phase. This project was supportedby a grant from the Natural Sciences and EngineeringResearch Council (NSERC) of Canada.

ReferencesArlt, R., J. Rendtel, P. Brown, V. Velkov, W. K. Hocking,

and J. Jones, The 1998 outburst and history of the JuneBootid meteor shower, Mon. Not. R. Astron. Soc., 308,887–896, 1999.

Cho, J. Y. N., and M. C. Kelley, Polar mesosphere summerradar echoes: observations and current theories, Rev.Geophys., 31(3), 243–265, 1993.

Cho, J. Y. N., C. M. Alcala, M. C. Kelley, and W. E. Swartz,Further effects of charged aerosols on summer meso-spheric radar scatter, J. Atmos. Terr. Phys., 58, 661–672,1996.

Czechowsky, P., I. M. Reid, and R. Ruster, VHF radarmeasurements of the aspect sensitivity of the summerpolar mesopause echoes over Andenes (69�N, 16�E),Norway, Geophys. Res. Lett., 15, 1259–1262, 1988.

Hagan, M. E., M. D. Burrage, J. M. Forbes, J. Hackney,W. J. Randel, and X. Zhang, GSWM-98: Results formigrating solar tides, J. Geophys. Res., 104, 6813–6828,1999.

Hocking, W. K., System design, signal-processing proce-dures, and preliminary results for the Canadian (London,Ontario) VHF Atmospheric Radar, Radio Sci., 32, 687–706, 1997.

Hocking, W. K., Temperatures using radar-meteor decaytimes, Geophys. Res. Lett., 26, 3297–3300, 1999.

Hocking, W. K., VHF tropospheric scatter anisotropy atResolute Bay and its implications for tropospheric radar-derived wind accuracies, Radio Sci., this issue (a).


Hocking, W. K., Middle atmosphere dynamical studies atResolute Bay over a full representative year: Mean winds,tides, and special oscillations, Radio Sci., this issue (b).

Hocking, W. K., and K. L. Mu, Upper and middle tropo-spheric kinetic energy dissipation rates from measure-ments of Cn

2 : Review of theories, in-situ investigations,and experimental studies using the Buckland Park atmo-spheric radar in Australia, J. Atmos. Terr. Phys., 59,1779–1803, 1997.

Hocking, W. K., and T. Thayaparan, Simultaneous andcolocated observation of winds and tides by MF andmeteor radars over London, Canada (43�N, 81�W), dur-ing 1994–1996, Radio Sci., 32, 833–865, 1997.

Hocking, W. K., S. Fukao, T. Tsuda, M. Yamamoto, T.Sato, and S. Kato, Aspect sensitivity of stratospheric VHFradio wave scatterers, particularly above 15-km altitude,Radio Sci., 25, 613–627, 1990.

Hocking, W. K., B. Fuller, and B. Vandepeer, Real-timedetermination of meteor-related parameters utilizingmodern digital technology, J. Atmos. Sol. Terr. Phys., 63,155–169, 2001.

Hooper, D., and L. Thomas, Aspect sensitivity of VHFscatterers in the troposphere and stratosphere from com-parisons of powers in off-vertical beams, J. Atmos. Terr.Phys., 57, 655–663, 1995.

Huaman, M., W. K. Hocking, and M. C. Kelley, Polarmesosphere summer echo studies at 51.5 MHz at Reso-lute Bay, Canada: Comparison with Poker Flat results,Radio Sci., this issue.

Jones, J., and J. D. Morton, A method for imaging radiometeor radiant distributions, Mon. Not. R. Astron. Soc.,200, 281, 1982.

Kelley, M. C., D. T. Farley, and J. Rottger, The effect ofcluster ions on anomalous VHF backscatter from thesummer polar mesosphere, Geophys. Res. Lett., 14, 1031–1034, 1987.

Manson, A. H., et al., Seasonal variations of the semi-diurnal and diurnal tides in the MLT: Multi-year MF

radar observations from 2 to 70�N, and the GSWM tidalmodel, J. Atmos. Sol. Terr. Phys., 61, 809–828, 1999.

Norton, A. P., Norton’s Star Atlas, 153 pp., Gall and Inglis,Edinburgh, Scotland, 1973.

Oppenheim, M. M., A. F. vom Endt, and L. P. Dyrud,Electrodynamics of meteor trail evolution in the equato-rial E-region ionosphere, Geophys. Res. Lett., 27, 3173–3176, 2000.

Rottger, J., Polar mesosphere summer echoes: Dynamicsand aeronomy of the mesosphere, Adv. Space Res., 14(9),123–137, 1994.

Tsuda, T., T. Sato, K. Hirose, S. Fukao, and S. Kato, MUradar observations of the aspect sensitivity of backscat-tered VHF echo power in the troposphere and lowerstratosphere, Radio Sci., 21, 971–980, 1986.

Tsuda, T., P. T. May, T. Sato, S. Kato, and S. Fukao,Simultaneous observations of reflection echoes and re-fractive index gradient in the troposphere and lowerstratosphere, Radio Sci., 23, 655–665, 1988.

W. O. J. Brown, Atmospheric Technology Division, Na-tional Center for Atmospheric Research, P. O. Box 3000,Boulder, CO 80307. ([email protected])

W. K. Hocking, D. Moorcroft, and J.-P. St. Maurice,Department of Physics and Astronomy, University of West-ern Ontario, Richmond Street NTH, London, Ontario,Canada N6A 3K7. (whock ing@ju l ian .uwo .ca ;[email protected]; [email protected])

M. Kelley, School of Electrical Engineering, CornellUniversity, 318 Rhodes Hall, Ithaca, NY 14853-5401.([email protected])

R. Rogers, Department of Atmospheric and OceanicSciences, McGill University, 805 Sherbrook Street West,Montreal, Quebec, Canada H3A 2K6. ([email protected])

(Received May 1, 2000; revised May 4, 2001;accepted May 8, 2001.)


1858

Documents

Resolute Bay VHF radar: A multipurpose tool for studies of ...physics.usask.ca/~jean/hocking 2001 2000RS001005.pdf · Resolute Bay VHF radar: A multipurpose tool for studies of tropospheric