When an electromagnetic signal passes through an inhomogenous medium, it is irregularly refracted and diffracted. If the signal is from a point source, various parts of the signal wavefront may recombine at an observing point and, having travelled different distances and thus presenting various phases, will show amplitude variations with time. These are termed scintillations.

Three different plasma media lie between a radio telescope on the Earth and a far distant (point) source: the interstellar medium, the interplanetary medium and the terrestrial ionosphere. Fortunately the frequency spectrum of the scintillations caused by each of these is sufficiently different, that the variations due to each of the three sources can usually be separated.

Region Timescale Critical Source Size
Ionospheric 30 sec 10 arcmin
Interplanetary 1 sec 0.5 arcsec
Interstellar 10s diffractive scattering)
106 (refractive scattering)

These values are for 100 MHz and are taken from Thompson, Moran and Swenson (2001).

Interplanetary scintillations depend on the line of sight density of the interplanetary medium, and may thus be used to monitor the same.

Interplanetary Scintillations

In particular, they may be used to detect the presence of solar coronal mass ejections (CMEs) as they travel outward from the Sun.


The method uses a large VHF (80-300 MHz) array radiotelescope to measure many point radio sources scattered across the sky. Several thousand such radio sources are required to produce a scintillation sky map of the required resolution. The individual steps in the observation and data reduction process are as follows (after Hewish and Duffett-Smith, 1987):

  1. The radio telescope observes each source for about 2 minutes (multiple beams are desirable to allow the observation of more sources per unit time).
  2. For each source the RMS variation of the source flux density (ΔS) in the frequency range 0.1 to 3 Hz is measured over that 2 minute period. The frequency filter is necessary to avoid contamination due to ionospheric scintillation which have periods from 10 seconds to several minutes, and even then strong ionospheric scintillations causing sytem saturation can contaminate the higher frequency IPS spectrum making it useless – it is thus necessary to check the unfiltered data for strong scintillations and reject this data.
  3. An enhancement factor g = ΔS / <S(ε) > is then computed for each observation where <S(ε)> is the long term average value for the particular source at a solar elongation value of ε degrees (ie angular distance from the Sun). This implies the existence of a “background” data base of these average values, from prior measurements. Such a data base will be a function of solar elongation for each of the several thousand sources. Typical g values usually range from about 0.5 to 2.
  4. The line of sight plasma density in the direction of that source is then computed from the relation Np = 9 g2.17 (cm-3).
  5. The observation of many sources over the entire sky then allows a line of sight 2-D plasma density contour map to be constructed. This is referred to as a g map.
  6. Regions of enhanced plasma density (eg CMEs) can be seen on the plasma contour maps.
  7. Computer modelling, using simple density disturbances, is then used in an attempt to recover the 3-D structure of the interplanetary medium from the 2-D contour maps.


Frequencies between about 80 and 300 MHz are typically used for the detection of interplanetary scintillations. Higher frequencies allow the equipment to look closer to the Sun (before strong scintillation from the ambient solar corona and solar wind makes the method unuseable) and also minimise ionospheric scintillation interference. However, source signal strengths decrease with frequency, and the lower signal to noise ratios means fewer sources are useable. Hewish and Duffett-Smith (1987) give a table indicating required array size and minimum useable solar elongation as a function of frequency. A version of this is given below:

Frequency Effective Collecting Area (m2) εM
(MHz) SD = 80 sr-1 SD = 145 sr-1 SD = 800 sr-1 (degrees)
3838,10051,600 145,00050
8214,80020,000 56,00030
1518,90012,000 33,70020
32711,60015,888 43,50012
40814,80020,000 56,00010

SD is the required source density in number of sources per steradian. The lower limit of 80 sources per steradian implies roughly 500 sources over the whole sky visible from the radiotelescope. This is regarded as the absolute minimum number of sources needed. The middle value of 145 per steradian corresponds to a “useful” system, whereas the 800 value is regarded by the authors as a “first class” system. From the above table it would appear that 100 - 150 MHz may be the optimum frequency range for an IPS monitoring system.

The array at Cambridge that Hewish and colleagues used to monitor the IPM consisted of 4096 dipoles at 83 MHz. These were arranged in the form of a north array and a south array, each which consisted of 16 rows of 128 dipoles each. The rows were aligned east-west resulting in a fan beam with its small beamwidth in the east-west direction. In the north-south direction 16 individual beams were monitored by 16 independent receivers. The beams were formed by the product (cross correlation) of the north and south arrays in a typical interferometer configuration. The array was thus used as a multiple beam transit instrument. The scintillation output of each receiver was the sum of the squares of the in-phase and quadrature components of the receiver final detector.


The design presented here is an economical version of the Cambridge array. The economy is achieved by using open-line feeder for each row of dipoles, and using only one receiver. Using such a feeder technique makes the equipment more susceptible to environmental effects, but also enables a reduction in the number of preamplifiers required from 256 ato 32. Limiting the system to one receiver, means that only one beam is available at any one time, thus limiting the one the number of sources that can be simultaneously monitored. As each source must be monitored for about 2 minutes, this limits the number of useful sources to a maximum of 720 per day. There would be, of course, nothing to preclude additional receivers being added at any time should funding be available.

The block diagram for such an instrument is shown below:

IPMA Outline


5.1 Dipole Detail

The major cost of the project will be associated with the 4096 dipole array.

Dipole detail

5.2 Butler Matrix

The two beam formers would most probably be implemented as 16x16 Butler Matrices. These have 16 input ports (the rows of each array, north or south) and 16 output ports, which comprise the 16 beams. One of these beams is selected at a time by a 16x1 RF switch (implemented by 16 diode switches).

A Butler Matrix is formed of quadrature (90o) hybrid couplers and phase delays. A typical implementation is shown below. The phase shifts may be implemented by suitable lengths of coaxial transmission line (noting the velocity factor).

Butler Matrix

5.3 Quadrature Hybrid Couplers

These are four port devices with two inputs and two outputs. The output ports provide equal amplitude signals but with a 90 degree phase difference.

Quadrature hybrid coupler

For high frequencies (ie microwave) the design is normally implemented as stripline on a printed circuit board. At lower frequencies it may be implemented by coaxial transmission line or by lumped LC elements. The latter normally takes the form of pi-networks in each arm.

5.4 Receiver Hybrid Coupler

The hybrid coupler at the input of the receiver is a 180 degree hybrid and not a quadrature hybrid. The 180o hybrid has two inputs and two outputs. The two outputs are the sum and the difference of the two input signals, as required for a standard interferometer.

Note that both types of hybrid coupler are commercially available to cover almost any frequency range.


IPS power spectrum

Figure 6.1: A representative power spectrum of IPS fluctuations. The red line shows the noise floor, the blue bars indicate the rms variation associated with the power spectrum.


Purvis, Alan, PhD Thesis, University of Cambridge, 1981

Butler, Jesse and Ralph Lowe, Beam-Forming Matrix Simplifies Design of Electronically Scanned Antennas, Electronic Design, 9, pp170-173, April 12, 1961.

Hewish, A. & Duffett-Smith, P.J., A New Method of Forecasting Geomagnetic Activity and Proton Showers, Planetary Space Science, 35, pp487f, 1987.

Thompson AR, JM Moran and GW Swenson, Interferometry Synthesis in Radio Astronomy, Wiley (New York, 2001)

ASAAustralian Space Academy