Search

Radio aurora detected at Prince Albert advances farther south with increasing magnetic activity. The latitude vs. magnetic activity relation for radio aurora is compared with similar relations for ionospheric currents. The effect of aspect sensitivity is discussed.

Meteor velocities can be determined from the diffraction characteristics of radar echoes from under-dense trails, but may be in error if atmospheric diffusion is neglected. Error factors are presented for circumstances when the diffusion coefficient is known. By a suitable selection of measurements beginning with the first minimum, the error in velocity may be held to about 1%.

The amplitudes of meteor echoes recorded on high-power radar equipment at Springhill Meteor Observatory have been measured for 57 samples of about 500 meteors each, obtained at various times of day on at least 1 day per month. The number of echoes as a function of echo amplitude is represented by a power law and the exponent in this relation calculated for each sample. Within the limits of error, no strong diurnal or seasonal variations in this exponent have been found. Similarly, there are no significant changes in the exponent when the data are divided into groups according to the height at which the echoes occurred and also according to the distance from the radar to meteor trail.Calibration of the equipment and the methods employed in obtaining and reducing the data are treated in detail. Theory relating the number distribution of echo amplitudes to the distribution of meteoroid masses is reviewed and assessed critically.A best value for the average mass index, s, of the sporadic meteor population in the range from 7th to 10th magnitude is s = 2.3 ± 0.1. The dispersion of individual sample values of s is about ± 0.3.

The determination of the prevailing wind in the 80–100 km region of the upper atmosphere by a new method, involving the simultaneous use of a CW doppler radar at 30.02 Mc/sec and three-station pulsed radars at about the same frequencies, is presented in this paper. This method involves the determination of the exact location of each observed meteor train and the component of the velocity of its horizontal drift in the direction of the azimuth from Ottawa. A 40-minute period during the Geminid shower on the night of Dec. 10/11, 1948, has been selected for this investigation. Theory for the analysis of the body doppler records is briefly outlined. The prevailing wind speed obtained from the body doppler frequencies (fd) is 54 m/sec. The observed linear variation in the average fd with time, in the case of each meteor, has been explained as caused by the effective point of reflection drifting along its train towards the maximum echo duration level. Periodic fluctuations of fd of the order of 1–3 c.p.s., on the average, have also been observed. The above two phenomena can be explained from a postulate of atmospheric turbulence on a scale of about 1 km or above.

The determination of the probability distribution for the magnitude of a radar echo from randomly distributed scatterers is a standard problem of the "random walk" type. We are concerned with the information to be gained about the scatterers from observation of the echoes, and the methods whereby this information may be extended. The method of averaging independent signals is investigated, and compared with the method of continuous averaging, which is found to be somewhat better. By appropriate variation of frequency of the radiation, it is found possible to obtain effectively independent determinations from a given region, and hence to derive more precise information about it. Finally, the continuous variation of signal with time is studied, and a simple picture of the development of the signal proposed. The Markoff method is used throughout.

A peak in 48.2-Mc/sec echo occurrence is observed at Saskatoon corresponding in time to the period of breakup of quiet arcs into active rayed structures. This is also the time of most frequent occurrence of characteristic "curl" forms in the aurora. If, as Gartlein suggests, the "curl" forms are formed by instabilities in a sheet beam then the primary particles are positively charged.

Short-lived bursts of 10 cm. radiation from auroral displays have been received by radar equipment. The sources of continuous radiation from a partially ionized medium are briefly discussed. From a knowledge of the constants of the equipment used it is deduced that the power density at the receiver is at least 2 × 10−10 watts per (meter)2, and it seems that the most likely source of this radiation is a plasma oscillation of the ionized volume associated with the auroral display. If this is so, the electron density in at least localized regions must be of the order 1011 per cm.3

A frequency-modulated, continuous-wave radar seems to be particularly well suited to the observation of auroral ionization, since it provides both range information and a Doppler indication of radial motion. An experimental equipment of this type has been operated for a few months near Ottawa. The system parameters are briefly considered and the radar observations are compared with similar measurements made with higher-powered pulse systems.

By assuming that auroral radio reflections are produced by volume scattering in clouds of ionization having the same spatial configuration as the visible auroral structures, and by taking into account the radar pulse duration and antenna beam width, it is possible to predict the probability of echo occurrence as a function of range and azimuth. This echo distribution is quite similar to that observed experimentally even when "aspect sensitivity" of the individual scatterer is neglected. Unfortunately, the optical evidence is not sufficiently extensive to permit precise calculations to be made, nor the radio evidence to permit detailed comparisons, but previous estimates of the shape of the scattering structures that have been based on the azimuthal echo distribution without regard to the factors discussed here are likely to be seriously in error.

Correlation studies of the radar echo occurrence rate from aurora in half-hourly intervals at Ottawa, S and Sd components of the horizontal magnetic field H at Agincourt, and the disturbance diurnal variation of earth current at Crow River have been carried out. Short-time variations in auroral echo strength and moderate perturbations in H have also been correlated. The auroral echo occurrence rate seems to have a diurnal variation characteristic similar to that of H. It is found that auroral activity always precedes magnetic activity. The variation of the delay time between the two phenomena shows a local time dependence, being practically constant and quite small (0–15 minutes) before local midnight and increasing afterwards. This variation of the delay time appears to have a connection with reports of others regarding reversal of the direction of auroral ionization drift from west to east somewhere around midnight with subsequent magnetic perturbations which change from positive to negative.No definite conclusion could be reached regarding the relationship of earth current to other factors because of a practically random variation of cross-correlation coefficients from month to month.

A backscatter, pulse-type, patrol radar has recorded meteor echoes continuously near Ottawa for the years 1958 to 1962 inclusive. Five-year mean hourly echo rates for each day in the year are presented for all echoes counted, a total of some seven million, and for echoes of duration greater than or equal to 8 seconds. The diurnal vaiations in hourly echo rates throughout the year are given as five-year means in five-day groups. (Author)

In relating the frequency distribution of the durations of radar echoes from meteor trails to the mass distribution of the meteoroid flux, two major effects must be taken into account. Because bright meteors tend to be observable by radar from any angle, the rate of long-duration echoes is enhanced relative to short-duration echoes. This effect is particularly pronounced for the Ottawa radar which has wide-beam antennas. On the other hand, the effect of electron attachment is to suppress the long-duration echo count, particularly for low-velocity meteors, which produce ionized trails at lower altitudes.The influence of both these effects on the observational results from the Ottawa meteor radar is examined quantitatively.

A study of the fine structure of air refractivity in the lower troposphere has been continued through the spring and summer of 1962 by the observation of weak radio reflections from clear air (radar angels), with a vertically directed radar of special design operating at 6770 Mc/s. These observations have been limited to the layer of frictional influence, within 1500 meters of the surface. The interpretation suggests that the reflecting centers are broad strata whose refractivity contrasts weakly with that of their environment, whose vertical depths are no more than a few centimeters, and which are either flat over horizontal distances of at least several meters or concave downwards with radii of curvature somewhat less than their height above ground. The incidence of transitory reflections generally follows a regular distribution in the vertical, with a maximum at 300 meters; the form of this distribution is modified by the intrusion of weather fronts, thermals, and other clear-weather structures. The transitory reflecting stratum is cut off from its generating source early in its life history, to be dispersed into its environment by molecular or eddy diffusion. The total incidence of transitory angels fluctuates quasi-periodically in time, with a period of about 10 minutes; it is suggested that this periodicity is due to the influence of internal gravity waves in the atmosphere. In contrast, persistent reflections are associated with a more stable environment; the maximum incidence is at the lowest heights observed, with a gradual decrease towards higher levels. Their relationship to clear-weather structure is less certain than for the transitory reflections. The persistent reflecting stratum must be replenished continuously by the generating source during its lifetime to offset diffusion into the environment.

Data from a Canadian v.h.f. radar network which operated during the IGY were analyzed to determine the aspect sensitivity of 48-MHz radio aurora. Two radars of this network simultaneously viewed a region of high-visual and radio-auroral activity. This paper is concerned with echoes which were obtained over a two-year period from auroral ionization simultaneously observed by these radars. The general properties of these simultaneous echoes were determined and are compared, where possible, with the corresponding properties of the overall set of echoes seen by each radar when looking independently at the same region. This comparison indicates that the mechanism responsible for simultaneous echoes is very probably the same as that which is responsible for the majority of radar echoes seen individually by the two radars. The spatial and temporal correlation between the simultaneous echoes and visual aurora is presented. The average aspect sensitivity of these echoes (48 MHz), which were obtained from simultaneous observations at off-perpendicular angles near 5° and near 20°, was found to be 1 dB per degree.

In the winter of 1948–49 a series of measurements was made at Ottawa in an attempt to verify, by experiment, the theoretical computations of J. W. Ryde for the intensity of backscattering of 3 cm. radio waves by snowflakes. The radiation backscattered from falling snow was measured, together with the actual mass of snow per cubic meter of atmosphere. Whenever possible the size distribution of the snowflakes was determined. Owing to the small number of snow-storms studied, firm conclusions cannot be drawn. However, the results obtained are in agreement with those predicted by Ryde, and provide useful practical information.

The photoelectric meteor detector consists of 19 type 931-A photomultiplier tubes arranged to cover the entire sky. Signals from the photocells are repeatedly sampled in time sequence and displayed on a cathode ray tube. Photoelectric meteor signals, radar meteor echoes, and a marker light operated by visual observers are photographed by a continuous-film camera.Preliminary tests indicated that the average sensitivity of the photoelectric detector to meteors was about one visual magnitude inferior to that of the human eye. Approximate meteor positions and qualitative light intensities could be obtained from the photoelectric record. Timing comparisons between the photoelectric and the radar records could be made to 0.05 sec. Methods of improving the sensitivity and other operational characteristics are discussed.

Radar backscatter observations of auroral ionization have been made from Ottawa on two similar systems operating at 488 and 944 Mc/s. These systems, as nearly scaled as was practicable, were operated almost continuously during the period from May 1959 to April 1961, inclusive.Narrow-beam transmitting and receiving antennas were used. They were each fixed in position to intercept the ionosphere at auroral heights in the region where the orthogonality condition for specular reflection is most nearly satisfied. A magnetometer recorded the magnetic disturbance at ground level at a point under the echoing region. Measurements were made of the time, amplitude, range, and frequency spectrum of the auroral echoes on each of the two frequencies, at both the transmitted and orthogonal polarizations.The data obtained indicate a double-peaked night time maximum in the occurrence of auroral echoes. A slight but distinct minimum appears just before local midnight. Equinoctial maxima are observed in the seasonal variation. The probability of occurrence of echoes increases rapidly with increasing magnetic disturbance. The average Doppler frequency shift of the echoes exhibits a systematic diurnal variation. There is a reversal in the direction of apparent motions, coincident with a change in sign of the average magnetic disturbance, and occurring at about the same time as the night time minimum in occurrence. From the two frequency data, echo cross section is found to vary as about the seventh power of the wavelength. There is no distinct evidence of depolarization or rotation of the plane of polarization of the echoes.Observations of the Doppler frequency shifts are interpreted in terms of the motions of auroral ionization. The wavelength dependence of auroral echoes is used to infer certain properties of the scattering medium.

Observations made with crossed-polarization radar system systems do not support the suggestion that the ionized meteor trail may act as a strong filter–polarizer of the incident radio wave. Experiments have been carried out to determine the variation of normal meteor echo rates with transmitter power, antenna gain, and radio wave length, and all confirm Lovell's scattering formula, provided that account is taken of the effective broadening of the scattering pattern of the meteor trail with increasing wave length. The limiting sensitivity of the 9.22 m. 200 kw. radar is determined to be about 9th magnitude. During a strong visual shower the observed increase in visual rates and low-power radar rates, compared to high-power radar rates, is explained by assuming that the magnitude distribution of the shower meteors differs from the normal nonshower distribution.

A comparison of simultaneous data on the visible aurora, the earth's magnetic field variation, vertical and oblique ionosonde echoes, and very-high-frequency radar echoes has been made. Long-range high-frequency and very-high-frequency radio echoes do not appear to correlate individually. Reflections are observed, however, on both frequency ranges coincident with the appearance of low elevation auroral arcs, which are associated with magnetic bays.

Measurements are presented on some 500 radar meteor-head echoes observed during the 1961 Perseid meteor shower. Mean absolute powers of head echoes and of the associated enduring echoes are given as a function of the angular position of the meteor along its trajectory. The detailed amplitude variation of three head echoes is discussed in relation to various scattering mechanisms. It is shown that a solar-controlled parameter of the atmosphere has a strong influence on head echoes.

The durations of radar echoes from meteors have been observed simultaneously on 9.22 m. and 5.35 m., and also on 9.22 m. and 2.83 m. The ratio of durations on two wave lengths decreases with increasing duration by a factor of two over the observed range, deviating significantly from the accepted square law of wave length. Plotting the log of the ratio against the log of the duration yields two straight lines of different slopes, one in the short-duration range and the other applying to the longer echoes. General empirical formulae are developed to predict the echo duration on one radio equipment in terms of the duration of the same echo recorded by another apparatus of different sensitivity and wave length.

For a meteor trail, rotational motion caused by wind gradients leads to errors in the decay constant of the radar echo. Two cases are assessed. In the first, large wind shears produce gross motion of the reflection point along a trail in which the electron density varies exponentially. Secondly, smaller motions on a short, finite trail are examined. These models cannot account for all of the observed scatter in decay-time measurements.

A systematic motion of 48.2 Mc/sec echoes associated with aurora is found at Saskatoon. The motion is towards the west before midnight and towards the east after midnight, the mean velocity in either direction showing a statistical relation to variations in the earth's magnetic field. No correlation is found between individual echo velocity and magnetic disturbance, and no period of zero velocity is observed. There is also evidence of an ordered relation between motion in the north–south direction and disturbances in the earth's magnetic field.

The rate of decay of the radio signal reflected from an underdense meteor trail is usually considered to be a simple function of trail height. In fact a large scatter is found in decay rates measured by means of radar or backscatter observations and this dispersion persists for forward-scatter measurements, where the decay rates can be measured with greater precision. The scatter persists even for measurements made at different radio frequencies on the same trail. The results of calculations are presented which indicate that the observed degree of variation in decay rates could result from nonuniformities in the trail ionization.

The radar echo from randomly distributed scatterers fluctuates in intensity about a mean equal to the sum of the intensities contributed by each particle. If k independent intensity values are averaged, the r.m.s. deviation of the fluctuation is reduced by a factor √k. Counting the fraction of signals falling above one or more thresholds is a practically simpler, though somewhat less rapid, alternative to averaging. As a further alternative, one can superimpose many independent sweeps of amplitude or signal level on an A-scope, and so observe the most probable value.In range interval l, a single sweep contains 2l/h independent data, where h is the pulse length. The intensity averaged under the continuous trace is slightly more accurate than the average of the independent data.Pulses sent through the same scatterers yield identical returns until the scatterers have had time to reshuffle themselves. Successive pulses give independent echoes, however, if their frequencies differ by about 1/τ, τ being the pulse duration. Echoes from the same pulse received at separate antennas are similarly independent.Several (k) independent echo values are helpful in measuring steady echoes against a noise background. For fluctuating echoes, k must be large, regardless of the relative strengths of echo and background. Large values of k in addition materially enhance the detectability of fluctuating echoes weak compared to the noise.

Approximately 9 years of Ottawa fixed-antenna 48 MHz auroral radar records have been examined for cases of periodically varying radio aurora following ssc's and si's. The occurrence, period, and time delay of onset of long period pulsating radio aurora (DPEl, and FPEl,) are found to be related to the amplitude of the ssc which precedes it. These relationships do not always hold for DPEl, and FPEl following si's. A discussion of the possibility of an association between the radio aurora and hydro-magnetic waves generated by the ssc or si is presented.

Meteor echo durations were observed simultaneously with two radar systems having a power sensitivity ratio of 33 db. An experimental relation between echo duration and system sensitivity is obtained which shows that the duration of meteor echoes varies slowly, but significantly, with system parameters. Curves are furnished to correct observed durations to zenithal durations, applicable to any system using a horizontal half-wave dipole antenna.

A simple, low-power 50 Mc./s. radar is described which has been designed for automatic recording of radio reflections from auroral ionization. The system features high sensitivity with good reliability. Photographic records are taken in the form of a continuous film strip displaying range and time. A complementary recording system employs circuitry which cancels interference and noise signals but produces an output signal proportional to the integrated echo intensity. The display in this latter case is in the form of a chart record from which data are immediately available.

The distribution in range and azimuth of auroral echoes obtained with an ultra-high frequency radar at Prince Albert is examined. Such distributions may be controlled (i) by the effect of volume integration in the auroral configurations, (ii) by the geographical distribution of aurora, and (iii) by aspect sensitivity of the auroral scatterers. With the radar parameters chosen for this experiment, (i) can be neglected. Of the remaining two effects, (ii) appears to be more important in the period immediately before local midnight and (iii) in the period immediately following local midnight.

The specifications and the component parts of a 20-kW meteor radar working at about 33 Mc/sec are described. In the design of this equipment special attention was given to long-term use and constancy of performance. Serviceability characteristics are noted for 8 years of continuous use.

A study of radar angels at 6800 Mc/s has been carried out at London, Canada, over a period of 1 year. Examination of the clear-air reflections from heights between 300 and 1500 meters above the vertically directed radar has shown that persistent angels occur for surface air temperatures between 30° and 50° F, and that transitory angels are present for surface temperatures above 20° F that are outside of this range. The maximum incidence of angels increases with type of air mass in the order cA, mA, mP, mT, although the most common duration is approximately the same in these air masses. If the angel is due to reflection at a horizontal flat stratum, the power reflection coefficient of the stratum lies between 10−16 and 10−14 for transitory angels and is as large as 10−11 for persistent angels. It is shown that transitory angels may arise through eddy mixing in the lower troposphere when refractivity contrasts of about one part per million occur, but that the reflecting stratum must be no more than a few centimeters in depth. It appears that persistent angels are associated with a high concentration of moisture at the ground, but their physical origin is not clear.

The concept of cross section as applied to a semi-infinite scattering body seems to require some clarification. The need for careful formulation of the problem arises because of the simultaneous occurrence of two characteristic lengths tending to infinity: the range from the radar to the target, and the size of the target. The infinite range assumption in the definition of the cross section allows the incident wave to be approximated as a plane wave in the case of a finite scatterer. For a semi-infinite body, it is customary to retain the plane-wave incidence, and introduce ad hoc arguments to dispose of the awkwardness due to the infinite extent of the scatterer. A return to the basic definition of a cross section, and examination of its motivation, lead here to an unequivocal formulation for the cross section of a semi-infinite body. Its consequences are pursued in the physical optics approximation. In particular, the nose-on backscattering from a body of revolution is exhibited, and results are computed for the paraboloid and the cone (which turn out to agree with the traditional ones). The broadside backscattering from a cylinder is also calculated, and the difference in this case between mono-static backscattering and the return in the backward direction from an incident plane wave is discussed.

Auroral radar echoes at very high frequency received at antennas sensitive to linearly polarized components in directions symmetrical to the plane of polarization of the vertically polarized transmitted wave and the ground plane indicate that the received wave, in addition to an unpolarized component, contains a linear component of polarization tilted in the general direction of the earth's magnetic field lines.

Total electron content measurements of the earth-moon interspace were made by measuring the Faraday rotation of the returned echoes from the moon at Prince Albert, Saskatchewan (Lat. 53 degrees 13 minutes, Long. 105 degrees 56 minutes), during the July 20, 1963, solar eclipse. The radar operates at 448 Mc/s with a maximum average power of about 150 kW and uses an 84-ft paraboloidal antenna with a beamwidth of about 2.5 degrees. Transmissions were made using horizontal linear polarization.(Author)