Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 77, 1948-49
This text is also available in PDF
(594 KB) Opens in new window
– 85 –

Some Problems And Techniques Involved In The Radio-Meteorological Investigation Conducted In Canterbury.

Summary: The phenomenon of superrefraction results in the trapping of ultra-high frequency electromagnetic radiation close to the earth's surface. It is caused by abnormal variations of temperature and humidity in the lower atmosphere. resulting in a sufficiently sharp lapse rate of refractive index to cause lays to have a curvature greater than that of the earth.

These conditions arise when warm dry air passing over a cooler sea surface is modified thereby, and an investigation is being conducted in Canterbury to discover:—


The relationship of the distributions of refractive index in the atmosphere with the distribution of energy from a transmitter situated at various levels.


The laws governing the modification of the air mass.


A forecasting technique for determining the onset and intensity of superrefraction conditions from general synoptic data.

The organization and general techniques adopted to achieve these results are discussed.

Introduction: Superrefraction or anomalous propagation is a phenomenon associated with electromagnetic propagation in the troposphere, which has very marked and important effects at the lower wave-lengths used in radar techniques.

– 86 –

With the comparatively small degree of bending of rays in a normal atmosphere, one might expect to find only diffracted energy at these wave-lengths beyond the horizon of a transmitter, but under superrefraction conditions strong fields are found close to the surface far beyond the horizon, very much greater than could be explained by diffraction alone.

One is led to the conclusion that a strongly reflecting or refracting layer exists at some distance above the earth's surface. The ionosphere can be disregarded, as the frequencies with which we are dealing are high enough for radiation to pass straight through it. A cloud might suggest itself as a reflecting layer, but superrefracion is frequently observed in a perfectly clear atmosphere, so we are forced to assume that a strongly refracting layer at some level in the atmosphere is responsible for the phenomenon.

We must first examine the normal distribution of refractive index in the atmosphere, determine what factors cause it to vary, and discover the variations under superrefraction conditions to see if they can account for the phenomenon.

Refraction in a Standard Atmosphere: If μ is the refractive index of the atmosphere, it is easily shown that, for small angles of elevation, the downward curvature of rays is given quite closely by − dμ/dh or μ′ where μ′ denotes the lapse rate of μ in the atmosphere.1

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Near ground level the refractive index of the atmosphere at height h, down to wave-length of a centimeter or so, is given by2. μn = 79/106T [P − PC/621 + 4800Pq/621T] + 1 (1)

where P = total pressure in millibars

q = specific humidity in grams per kilogram

and T = absolute temperature.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Differentiating with respect to height, this reduces at NTP roughly to— μ′ = 5 × 10−6 (. 072P′ + 1.6q′ − .11t′)

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

where t denotes degrees Fahrenheit and P′, t′, and q′ denote the lapse rate of pressure, temperature, and specific humidity. But the curvature of the earth is 5 × 10−6 radians per hundred feet, so the above is equivalent to μ′ = Ke (.24 + 1.6q′ − .11t′) (2)

where Ke is the curvature of the earth and q′ and t′ are expressed in grams per kilogram and degrees Fahrenheit per hundred feet. The term .24 arises from the normal lapse rate of pressure in the lower atmosphere.

In a standard (i.e., thoroughly mixed) atmosphere, we have

q′ = .04 gms/kgm/100 feet.

t′ = .5°F/100 feet.

so μ′ = ¼ Ke.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Thus the downward curvature of rays due to refraction in a standard atmosphere amounts to about one quarter of the earth's curvature. The treatment of curved rays over a curved earth is mathematically complicated, and in calculations on the subject the curvature of the earth is adjusted so that rays are straight— i.e., the earth is made to have a radius of 4/3 its natural value. The horizon on this new earth is known as the “radio” or “radar” horizon, and in a standard atmosphere one would expect only the weak diffraction field beyond this horizon.

Picture icon

Fig. 1—Path of Rays in Standard Atmosphere over Earth of Curvature ¾ Ke.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Modified Refractive Index: It is convenient for many purposes to consider the earth as flat, and apply the earth's curvature to rays. With the fiat-earth treatment a modified refractive is introduced, denoted by M, such that M′ = μ′ −; Ke (3)

where M′ denotes the lapse rate of M. In a standard atmosphere μ′ equals ¼ Ke so M′ is negative, or the gradient of M is positive with height, and rays have an upward curvature equal to ¾ Ke.

– 87 –
Picture icon

Fig. 2—The Distribution of Refractive Index and Path of Rays in a Standard Atmosphere on the Curved and Flat-earth Concepts.

The path of rays leaving a transmitter at various angles in a standard atmosphere is shown in Fig. 2, on the curved and flat-earth concepts.

Abnormal Variation of Refractive Index in the Atmosphere giving rise to Superrefraction: Under certain conditions, however, lapse rates of humidity and temperature are far from normal, and the downward curvature of rays may be several times greater than that of the earth. From equation (2) above it can be seen that a lapse rate of humidity greater than normal and a lapse rate of temperature less than normal or negative lapse rate of temperature (i.e., temperature inversion), will give rise to a greater value of μ′. An actual case will illustrate this point.

On the 24th February of this year, when a north-west wind was blowing offshore in Mid-Canterbury, average lapse rates of temperature and humidity in the first 135 ft. at a point 8 ½ miles off the coast were

t′ = —4.3°F/100 ft.

q′ = 2.4 gm/kgm/100 ft.

which gives μ′ = 4.6 Ke and M′ = 3.6 Ke in this region.

Above this level, lapse rates were normal, so the average distribution of modified refractive index and the behaviour of rays from a transmitter situated at, say, 50 ft. is shown in Fig. 3.

Picture icon

Fig. 3—The Behaviour of Rays Propagated within a Surface Duct.

Rays leaving the transmitter at small angles of elevation are bent towards the earth's surface, where they are reflected, and then by a process of continuous refraction they are once more returned to the surface, where they are again reflected. This process will continue as long as the given distribution of modified refractive index exists, and the energy associated with these rays will be concentrated close to the surface of the earth and penetrate well beyond the radiohorizon. Rays propagated at other than small angles of elevation will, however, penetrate into the region where the lapse rate of M is normal, and will not be returned to the earth's surface.

When a negative gradient of modified refractive index occurs we have what is known as a “modified index inversion” or “radio duct.” Rays are said to be “trapped” within the duct, which is sometimes known as a “trapping layer.” An M inversion can exist only over a limited height interval, however—thus in the above case it extended from the surface up to 135 ft., the gradient being normal above this height. This type of duct is known as a “simple surface duct.” “Elevated” ducts exist, as shown in Fig. 4, where the lapse rate of M is normal above and below the inversion.

From equation (3) it is seen that if μ′ = Ke then M′ = O (i.e., M is constant with height) and rays have the same curvature as that of the earth. This condition is shown in Fig. 5. If M′ > O, rays have a curvature greater than the earth. From this the usefulness of the modified refractive index concept will be readily appreciated.

– 88 –
Picture icon

Fig. 4—Elevated Duct.

Picture icon

Fig. 5—Behaviour of Rays when M is constant with Height.

The case cited in Fig. 3 is simpler than that obtained in practise. A given distribution of modified index with height is seldom maintained over any great distance. In fact, at the same time as the case above, at a point fifty miles offshore, the average lapse rates of temperature and humidity were such that μ′ = 1.85 Ke and the M inversion extended up to 350 ft. from the surface. The initial properties of an air mass, the distribution of wind velocity and direction in it, and the nature of the surface over which the air mass is moving, all influence the present and future distributions of refractive index over a path. The problem of discovering the behaviour of radiation in practise is obviously a complex one.

Conditions giving rise to Superrefraction and the Importance of the Problem: We have seen that superrefraction occurs to a varying degree when the lapse rate of humidity is greater than normal and the lapse rate of temperature is less than normal. These conditions are brought about when warm dry air overlies cool damp air, and can occur in various ways. One of the most important is the advection of warm dry air over a relatively cool sea; the lower layers are cooled and moisture evaporates into them, bringing about the conditions for a simple surface duct to form.

The phenomenon of superrefraction makes itself apparent on radars by giving greatly extended ranges on targets at low levels, and this is a very important effect especially in war time. It became vital during the recent war to discover the mechanism of the phenomenon, especially under advection conditions, and although a great deal of quantitative radio information was obtained, it was of little use owing largely to the number of variables in the atmosphere over a path near the surfce. A site was required where radio-ducts could develop in a steady meteorological situation, and where the distribution of modified refractive index and the field strength from transmitters situated near the surface could be studied in detail.

The Fohn Wind in Canterbury and the Launching of the Canterbury Project: When a moving air-mass encounters an obstacle such as a mountain range it is forced upwards, and in so doing much of the moisture condenses and is precipitated, the latent heat of condensation warming the air. On the lee side of the obstacle the air descends, and streams out at lower levels as a hot dry wind. The north-west wind as developed in Canterbury, New Zealand, is of this type, known as a Fohn wind, and blows at all seasons of the year up to several days at a time.

Fig. 6 is a map of the centre portion of the South Island of New Zealand In Mid-Canterbury, at Ashburton, the plain is about thirty miles wide from the foot-hills to the sea, with a gentle slope clear of obstructions.

The general direction of wind development is approximately at right angles to the coast-line, which is practically straight and runs in a north-east–south-west direction. The area offshore is completely clear of any obstructions, such as islands, to complicate the picture, and conditions are ideal for a simple surface duct to form in a north-west situation. There is every chance that a detailed study of the problem of superrefraction can be made in a steady and simple meteorological situation, with the minimum number of complicating factors. A long-term investigation, known as the “Canterbury Project” has been inaugurated; operations were begun towards the end of September. 1946. and will continue until some time towards the end of this year.

– 89 –
Picture icon

Fig 6 Map of Centre Portion of South Island of New Zealand Showing Area of Operations

The Problem: Briefly, the problem is threefold:


To correlate the distribution of modified refractive index with the distribution of field strength in and above a radio-duct, from transmitters on different frequencies situated at various levels.


To study the modification of an air-mass as it moves offshore and formulate the rules governing the changes.


To develop if possible a forecasting technique for the onset, intensity and duration of superrefraction at various frequencies under advection conditions, from the general synoptic situation, or from a detailed low level study of the air structure at some convenient point inland.

The first section of the problem requires ideally an instantaneous picture of the conditions of the atmosphere and the distribution of field strength from transmitters on different wave-lengths over a path from the coast to some distance out to sea, of the order or 100 miles, from the surface up to a few thousand feet. This is impossible to obtain in practise, but in a steady meteorological situation fast-moving sources of measurement should be able to obtain a fairly complete and reliable picture. The second and third sections of the problem require successive detailed measurements of conditions in a moving air-mass, up to some 1,500 to 2.000 ft., as it passes over land and out to sea. Again a steady meteorological situation is highly desirable, since it allows some freedom in the timing and placing of the various observations.

Mobile sources of measurement are essential to obtain full advantage of a given set of conditions without an unwieldy amount of equipment.

General Techniques Adopted in Canterbury: Fig. 6 shows the area of operations. The line runs in a north-west-south-east direction across the Canterbury Plains just north of the path of the Ashburton River, and out to sea for about 100 miles.

– 90 –

Radar transmitters on 3 cm., 10 cm., and 3 m. are situated at the coast about 25 ft. above mean sea-level, while a second station on 3 and 10 cm. is situated a little to the north at a height of about 85 ft. above mean sea-level. Anson aircraft, on loan from R.N.Z.A.F., are fitted with receivers on these three channels which amplify signals from their respective ground stations and provide pulses to trigger associated transmitters on the same frequencies, which reradiate pulses of constant power. The signals are accurately measured on receivers at the ground stations. A trawler is fitted with similar equipment. Under north-west conditions the aircraft and trawler work in the region offshore, and with suitable manipulation, their positions always being known, a picture of the distribution of field strength in two dimensions along a path approximately at right angles to the shore is obtained. A channel on 50 cm. is being installed to obtain further data on the influence of a given set of conditions on propagation at different frequencies. The aircraft is also fitted with a wet-and-dry-bulb aircraft psychrometer, type Mark VI B, which measures temperature and humidity.

Three heavy trucks on land and the trawler at sea are fitted with the American-wired sonde equipment which, by means of raising and lowering special elements with the aid of kite or balloon, measures the temperature and humidity of the lowest 1,500 ft. or so of the atmosphere.3 A small cup-type anemometer is similarly used to give wind profiles in this region, and pilot balloon ascents at Ashburton provide further data above the reach of this equipment. Three fixed ground meteorological stations across the plains supply continuous records of temperature, humidity, pressure, and wind velocity and direction. The trawler is also fitted with a marine barometer to supply further pressure data out to sea. Radiosonde data on the properties of the air-mass before it crosses the mountain chain is obtainable from the meteorological station at Hokitika.

By suitably positioning the trucks on the coast and inland, and the ship and aircraft out to sea, the structure and modification of the air-mass as it moves across the land and offshore may be obtained.

All observations are controlled and co-ordinated from a central point at the Headquarters, Ashburton Aerodrome. The meteorological sounding trucks and the trawler are linked to Headquarters by radio-telephone, with the aircraft linked on a second channel and the ground radar stations by land-line. Information concerning sharp gradients in the atmosphere are frequently passed to the controller of operations at Headquarters from all meteorological sounding teams so that he may have at all times as complete a picture as possible of the whole situation, and may consequently direct activities so that the maximum amount of useful information may be obtained.

The aircraft, being the fastest moving source of measurement, covers as much ground as possible during the period of the flight (around two hours), and usually follows the type of course shown in Fig. 7.

Picture icon

Fig. 7 Cross Sections of Area of Operations Showing Participating Units.

The sounding trucks are situated with one always at the coast and others at variable distances inland according to conditions. The modification of the air-mass close inshore is very rapid, the gradients in this region being sometimes too sharp to be covered successfully by the aircraft, so to date the ship has been employed mostly in this region on kite and balloon soundings. Soundings by the trucks fill in the picture over land. Radar observations are taken on the outer legs of the aircraft flight, readings being made at the ground stations every ten seconds. Aircraft psychrometer observations are made every twenty seconds throughout the flight, and at the trucks and trawler observations are made every 6 to 50 ft. or so according to the sharpness of the gradients.

– 91 –

It can be seen with this organization and technique that, under steady meteorological conditions, data can be acquired which should lead to the solution of the problems. How far they are successful can be shown only by the examination of results, some of which are described in another paper.4


1. Booker, H. G. The Theory of Anomalous Propagation in the Troposphere and its Relation to Waveguides and Diffraction. T.R.E. Report, No. T1447, 12th April, 1943.

2. Katz, I., and Austin, J. M. Qualitative Survey of Meteorological Factors affecting Microwave Propagation. Radiation Laboratory, Massachusetts Institute of Technology, Report No. 488, 1st June, 1944.

3. Anderson, P. A., Barker, C. L., Fitzsommons, K. E., and Stephenson, S. I. The Captive Radiosonde and Wired Sonde Technique for Detailed Low Level Meteorological Sounding. Department of Physics, Washington State College, N.D.R.C. Project No. P.D.R.C.—647, Contract No. OEMst—728, Report No. 3, 4th October, 1943.

4. Davies, H., A Preliminary Study of some Results obtained in the Radio Meteorological Investigation conducted in Canterbury. This volume, pp. 78–85.