A talk given at the International Riometer Meeting, Zermatt Resort, Utah, June 22, 2008
First there was an advent of artificial earth satellites (1957) then came an epiphany in Riometers in 1971! In June of that year I was in my office on sabbatical at Bern, Switzerland, when a circulated letter arrived from Ray Heer of the Office of Polar Programs of the US National Science Foundation to state that there was now an unmanned geophysical observatory (UGO) at McMurdo Station, Antarctica, but with no scientific instruments on board. Like an artificial earth satellite, instruments needed to be installed within the capsule and would be provided with a portion of the space and power available together with telemetry lines. The indicated space available was a cuboid about 30x15x15 cms. and the share of power would be 5 watts at12 volts DC for each sensor.
At that moment I took a great leap of faith and decided it would be possible to create a Riometer within these restrictions even though the instruments in service at that time were about 10 times as large and used 10 times the power at line voltage. I quickly contacted (no emails!) John Hargreaves at Lancaster University, and Steffan Maagoe, a Danish engineer working at the University of California, San Diego (UCSD). The main challenge was to reduce the power consumption down to 5w and this could only be done by replacing the thermionic noise diode of the existing machines with some solid state device yet to be invented!
The proposal was written and UCSD was funded rapidly for this work. I returned to San Diego and the task began. After various trials and experiments with possible RF sources our design settled on using the unwanted RF noise generated in the base-emitter junction of an audio frequency transistor. This noise could be amplified in the transistor and used as a comparison source to mimic the received antenna noise in a Riometer. After false starts based on time sharing the avalanche noise in zener diodes the transistor source was perfected with notable contributions from Ian Bird then of the Australian National Antarctic Program (ANARE) in Melbourne. Final construction and packaging design of the complete UCSD Riometer was carried out by Bob Barker and Billy Bird.
By January 1972, the prototype was installed within the UGO at McMurdo. A hastily designed new wide beam antenna was used along with a calibration noise source developed during our work with noise sources. The new Riometer successfully tracked galactic radio noise throughout the first year.
An advantage of the miniature Riometer was that it no longer depended on the delay of heating the cathode of a thermionic diode to a settled level so the loop response times could be reduced. Arbitrarily 0.25sec was chosen since that seemed fast compared with the 1 minute or more for Riometers then in use. Another significant step was to cancel the old concept of an asymmetrical time constant which had been used as an aid to reducing interference, so that both increasing and decreasing absorption could be recorded without distortion.
During the second year of operation another new instrument was installed at Siple station which was specifically created to be a low auroral latitude site conjugate to sites available in eastern Canada. Immediately, new types of auroral absorption events were seen most notably the fast onset (2 sec.) spike events which have been studied since.
At this point the rush to renewed Riometry was on! UCSD could not produce copies requested by colleagues throughout the world so La Jolla Sciences was formed in 1973 to be a commercial production unit in private hands. Since its founding, LJS has supplied 320 Riometers to the world science community along with associated antennas and noise sources. Continuous upgrading has taken place to improve the ruggedness of the design and to incorporate new circuit possibilities as electronic technology progressed.
In 1985, after an advent and an epiphany, came a conversion! In 1968 at UCSD Kenneth Bowles adapted a microwave Butler matrix phasing system for use at VHF to form multiple beams with a large array of antenna elements. Specifically he used a 16x16 element structure switching the beams to follow radio sources across the sky to study solar wind scintillations. It later became clear to a group at UCSD (Sir Ian Axford, John Hargreaves and me) that such an array could be used to measure auroral absorption in each beam continuously by connecting a Riometer to each beam and so produce a map of overhead auroral absorption in real time with the fine resolution given by the 3 degree beams.
A new team was formed to execute this plan and it was proposed to NSF to carry out the work at the University of Maryland. The original P.I.’s were Ted Rosenberg, John Hargreaves and me. The Riometer modification portion was carried out by LJS where the main issue was to create a very fast time constant model so that the number of Riometers required for the beams could be reduced. Each Riometer attached to the array would have to sample many beams sequentially. The time constant was successfully lowered to 0.5mSec. and with this, each of 8 Riometers could sample 8 beams in less than one minute and so provide a dynamic picture of small scale absorption clumps moving through 64 beams of the array.
In the 20 years since the first IRIS was installed at South Pole, about 15 similar systems have been installed world wide, many produced at U. Maryland using LJS Riometers by a team headed by Ted Rosenberg supported by Allan Weatherwax, Dan Detrick, Larry Lutz and John Giganti. Many papers have now been published with results from these studies and many installations continue in use.
At this juncture it is important to consider the future availability of Riometer equipment in the light of continuing interest in both IRIS systems and large networks of individual Riometers monitoring auroral absorption over vast regions separated in latitude and longitude. Regular commercial production of the hardware is not practical because of a small market for limited use instruments which have been developed to function indefinitely. Changes in electronic components have also made for difficult production since replacement devices have to be patched into old circuit structures.
In another article I will present ideas on what needs to be done to continue to make available Riometers for the next 30 years in a non-commercial setting. The first task will be to make a new design so that the latest components can be utilized in a layout which will hopefully be useable over 30 years without serious component problems. Additionally, plans need to be made to service the existing suite of 320 installed Riometers in the field even thought the failure and repair rate is very low.
To conclude and paraphrasing Sir Winston Churchill in his speech at Fulton, Missouri, in 1945, “Riometry is like the Mississippi, it just keeps rolling along and I say ‘Let it roll!'”