ÎçÒ¹¸£Àû1000¼¯ºÏ

WHEN Explorer I, the first American satellite, was launched in 1958, it carried a Geiger counter to study cosmic rays. To the amazement of scientists on Earth, at certain parts of Explorer’s orbit, the counter gave a read-out of zero, instead of the 30 counts per second that they were expecting.

James Van Allen of the University of Iowa had designed the experiment. Van Allen and his team knew that the Geiger counter was working because at other points in Explorer’s orbit, the counter behaved as expected. Later, Explorer III recorded enough data to unravel the mystery. It turned out that at certain parts of Explorer I’s orbit, radiation so swamped the counter that there was no time for it to recover from recording one impact before the next struck. As a result, the counter gave a reading of zero. Scientists were so surprised that Ernie Ray, a member of Van Allen’s team, exclaimed: ‘My God, space is radioactive!’

The high intensity radiation proved to exist in belts around the Earth, which were dubbed the Van Allen belts. Since then scientists have learned more about radiation in space, but much remains to be clarified. To find out more, the Pentagon and NASA were due to launch a satellite earlier this week called the combined release and radiation effects satellite (CRRES). It was launched on an Atlas Centaur rocket, and is designed to operate for three years.

By radiation in space, scientists and engineers usually mean charged particles of varying energy levels. The particles might be anything from protons and electrons to iron ions. Clearly, spacecraft manufacturers must design their satellites so that they withstand the onslaught of these charged particles. Broadly, we know that a particle with an energy between 0 and 20 kiloelectronvolts can charge the surface of a satellite so that sparks jump between parts of the satellite, causing physical damage. Particles with higher energies cause radiation damage to the memories and software of microprocessors.

Radiation damage might be what is called a single event upset where ionisation flips a memory location from a 0 to a 1, thus corrupting computer programs. Alternatively, an accumulated dose of radiation might cause changes in the current of memories or amplifiers. Such alterations are known as parameter shifts.

The world’s broadcasting, communications, navigation and weather satellites must survive in geostationary orbit, which is roughly the same altitude as the outer edge of the outer Van Allen belt. The Van Allen belts are regions where a higher density of charged particles is trapped in the Earth’s magnetic field than at other regions above and below. These particles bounce back and forth between the northern and southern hemispheres, following the Earth’s closed magnetic field lines (see Figure 1). We now know that there are two main belts. The inner belt contains a higher density of particles, mostly protons, at an average altitude over the equator of 5000 kilometres. The charged particles here have peak energies greater than 30 megaelectronvolts. The outer ring stretches from about 25 000 kilometres over the equator to beyond 36 000 kilometres, and contains mostly electrons with peak energies greater than 1.5 megaelectronvolts.FIG-mg17263701.GIF

The Van Allen belts are only part of the radiation environment. High-energy protons from cosmic rays make up part of the radiation in space. All of this radiation varies over both large and small distances and from minute to minute as well as from year to year. For example, during the years immediately after Solar maximum (defined traditionally as when the most dark spots are visible on the surface of the Sun), the Sun produces more large flares than at other times in its 11-year cycle. And these flares shoot large numbers of high-energy protons through the solar system. Currently, we are at Solar maximum, and radiation from the Sun will be at a peak in the next couple of years. Last year, for the first time ever, Concorde’s radiation alarm sounded. The sensors had picked up high-energy protons, which had penetrated the Earth’s magnetic field lines.

Solar wind sweeps charged particles from the Sun through the Solar System in streams travelling at between 400 and 800 kilometres per second. The particles interact with the magnetic field around the Earth, and in doing so they provide a constantly varying contribution to the radiation environment. We think that the radiation belts trap these charged particles from the solar wind.

The Earth’s magnetic field itself varies in strength, contributing to the complexity of the radiation environment of space. The field is weak in a region, known as the South Atlantic Anomaly, which is south of the equator off the coast of Brazil. Consequently, the magnetic field lines are distorted in this region, and some of the charged particles trapped in relatively large quantities by the radiation belts penetrate further towards the ground than at other places around the Earth.

The problem with the South Atlantic Anomaly is that spacecraft in relatively low orbits can encounter high-energy particles that are excluded elsewhere in their orbit by the Earth’s magnetic field. These particles might damage electronic memories or give a false reading in a scientific instrument. Space agencies must find ways around this problem. Currently, mission scientists running the Rosat X-ray telescope that was launched last month are deciding what to do as it passes through the South Atlantic Anomaly.

Unfortunately, engineers cannot check on the ground how well designed their spacecraft are to withstand radiation, because it is physically impossible to recreate the full radiation environment. The alternative is to simulate the radiation environment using computers. Coupling a computer simulation of the radiation environment with programs simulating the spacecraft’s orbit, engineers can work out the radiation dose the satellite will receive over its lifetime. Satellite designers then select components that can withstand the estimated doses of radiation. As electronic components become smaller, and more and more of them are squashed into the same volume, this task becomes more critical, because one encounter with a charged particle can affect more components.

It is now clear, though, that the models are not adequate. They are based on measurements made during the 1960s and 1970s, with instruments that, today, look relatively primitive. The measurements they made are not sufficiently detailed to give an accurate model of radiation in space. They also overestimate the radiation risk. And, because satellites cost millions of dollars, manufacturers play safe. Often, they specify components that are too ‘radiation hardened’ rather than risk damaging a satellite worth millions of dollars. But the components may cost ten times more than necessary. Clearly, it is important to everybody involved in the space business – university groups building sophisticated scientific experiments, space agencies paying for satellites and contractors building subsystems for satellites – to aim for the most cost-effective designs.

The need to refine computer models and to understand the radiation environment more precisely underlies the CRRES mission. The first task is to gather as much data as possible to feed static models that detail the radiation environment over distance but not time. Eventually, spacecraft may collect enough data to construct dynamic models, showing how the radiation environment changes with time.

CRRES carries instruments to record the energy levels, mass and direction of the flow of charged particles. Each of these parameters is significant to the satellite manufacturer. First, energy levels show what type of damage to expect, and CRRES can measure energy levels over nine orders of magnitude from electronvolts to gigaelectronvolts. Secondly, the mass gives an idea of how much damage a particle will cause because the greater the mass, the greater the ionisation that particle will cause when it hits a component in a satellite. Finally, knowledge about the direction of flow will help satellite manufacturers to decide if and where they should place extra shielding.

Testing technology in space

CRRES is in an elliptical orbit around the Earth. The satellite’s closest approach to Earth is 350 kilometres, and the furthest it travels from Earth 35 800 kilometres. This orbit takes CRRES through the heart of both the inner and outer Van Allen belts. The satellite will also map the edges of the South Atlantic Anomaly.

Besides carrying the instruments to probe the radiation environment, CRRES carries a package of microelectronic devices and individual components using different types of technology so that scientists can see what happens to them in space. Some of the components are no longer state of the art, because the launch of CRRES was delayed by the Challenger disaster and the spacecraft was redesigned for launch on Atlas Centaur. Nevertheless, scientists will gain useful information. In all, NASA and the Department of Defense are testing 65 types of device and 450 individual components. These include memories and microprocessors.

In this way, NASA and the Department of Defense will match specific damage to the satellite’s position in orbit. They will also compare the results in orbit with laboratory tests, where tests are usually made with gamma rays at high dose rates, and they may not represent accurately what happens in space.

The second major aim of the CRRES mission is to study what happens when artificial charged particles are released into the magnetosphere and the ionosphere. The magnetosphere is the region around the Earth where the Earth’s magnetic field dominates the behaviour of charged particles. The boundary between the magnetosphere and the rest of the Solar System is called the magnetopause. The Solar wind flows around the magnetopause. On the side of the Earth where the Solar wind first encounters the Earth’s magnetic field, a bow shock, rather like that around the prow of a ship, forms. At the far side of the Earth, the Solar wind flows by, forming what looks like a wake, leaving a large region where the magnetosphere stretches out like a tail – and is called the magnetotail (see Figure 2). The ionosphere is a region within the magnetosphere, which extends roughly between 60 and 1000 kilometres, where ultraviolet radiation from the Sun ionises atoms and molecules in the upper atmosphere.FIG-mg17263702.GIF

Mission controllers will command the spacecraft to eject canisters of barium, lithium, calcium and strontium. After a delay of 25 minutes, when the canisters are at least 2.5 kilometres from the spacecraft, the canisters will explode, releasing clouds up to 100 kilometres in diameter. Soon after release, the atoms will ionise. Scientists on the ground and in aircraft will view how the clouds of ions move along and around the magnetic field lines. They will measure velocities and temperatures. Eventually, this data could feed into a dynamic computer simulation of radiation in space. One set of releases will be at low altitudes of 350 kilometres, and will clarify how charged particles interact with magnetic fields, the interactions between the upper atmosphere and the ionosphere, the chemistry of the ionosphere and the structure of electric fields at low altitudes.

Releases at high altitude will probe interactions between the magnetosphere and ionosphere and how those interactions affect the stability of particles trapped in the Van Allen radiation belts. Eventually, understanding how stable particles are in the magnetosphere and how that stability is affected by conditions in the Solar wind will help scientists forecast what the radiation environment of space will be.

Another experiment will study how irregularities in the density of charged particles in the ionosphere affect the way that short wavelength radio signals bounce off the ionosphere as well as providing insight as to how variations in density affect the passage of signals through the ionosphere. Besides studying naturally occurring irregularities in the ionosphere, CRRES will introduce a few of its own. An instrument known as a plasma probe will apply a voltage and excite surrounding charged particles. In turn the charged particles excite their neighbours, generating what is called a plasma wave, which eventually dies out. Another instrument on board measures the varying electromagnetic field and observes how the wave dies out, thus obtaining information about the density and temperature of particles of use both to modellers and communications experts. This experiment will operate only on every fourth orbit for about 15 minutes.

British scientists will supply equipment for an experimental package led by the Air Force Geophysics Laboratory near Boston, in the US. The package includes electronic devices that will be exposed to radiation and more than 30 instruments measuring the direction of flow of charged particles, masses, energies and the strength of electric and magnetic fields.

The Rutherford Appleton Laboratory will measure high energy particles, and The Mullard Space Science Laboratory at University College London has provided a sensor that will measure the number of particles at given energies between 10 electronvolts and 30 kiloelectronvolts, and will also show their direction of flow. Raw data from this instrument then passes to equipment provided by the University of Sussex, which measures the frequency of charged particles at given energy levels.

This range of magnetospheric particles is of prime importance, because they are a source for the more energetic radiation belt particles. Particles in this energy range are injected into the radiation belts from the magnetotail when plasma is shot towards the Earth by the explosive process of joining oppositely directed magnetic fields. Understanding this process fully will be vital in the development of a dynamic model of radiation in space and could be useful in cometary physics and nuclear fusion experiments.

The Mullard instrument also works in the appropriate energy range to diagnose spacecraft charging effects. This well known but important problem is caused by the uneven balance of currents to the spacecraft skin. When the spacecraft is in sunlight, the dominant current is normally photoelectrons emitted by the spacecraft surface, and, because the spacecraft has emitted electrons, it has a positive potential with respect to the surrounding charged particles. In eclipse, however, the current balance depends critically on the energy spectra of the ions and electrons of the ambient plasma, and charging of the spacecraft’s skin of up to tens of kilovolts is possible. If the spacecraft skin has non-conductive areas, one part of the surface may charge to a different voltage with respect to another. Sparks between these different regions can cause physical damage as well as spurious electrical commands and interruptions.

These particle measurements are essential in characterising the interaction of a spacecraft with its environment. But are there any examples on real satellites where surface charging or radiation damage has occurred? One example, in which we have been involved during the past 10 years, is the Meteosat weather satellite programme. These satellites, owned by the European Space Agency (ESA), provide photographs for TV weather forecasts. The first of the four Meteosat satellites launched so far suffered from a number of unexplained problems often referred to as satellite ‘anomalies’. These anomalies interfered with the production of weather pictures. Although the rate of the anomalies was quite low, one or two per day, depending on the season, the ESA decided to put equipment to monitor spacecraft charging on the second Meteosat.

The results showed surface charging of the spacecraft, particularly in eclipse, but the anomalies did not correlate with the occurrence of charging. So, the ESA decided to fly a detector of the high energy particles that cause radiation damage rather than surface charging. Such a sensor is on the Meteosat 3 satellite, which was launched in June 1988. The sensor detects electrons at the lower-energy range of penetrating radiation (from 40 to 300 kiloelectronvolts). Data from the sensor told scientists that the anomalies coincide with high flux densities.

We believe that particles with energies from tens to hundreds of kiloelectronvolts cause what is known as deep dielectric charging. If the particle flux remains high, they build up a charge in the insulating sleeve of electric cables. Eventually, the insulation breaks down, causing anomalies. Scientists believe that similar effects caused the complete failure of the US’s East coast weather satellite, GOES 5, in 1984.

Although CRRES is designed to operate for only three years, scientists hope that it will last longer. If it does, the satellite’s operational phase will overlap with those of a suite of satellites due for launch during the 1990s. These satellites will be launched by the world’s space agencies to undertake a variety of similar scientific missions. By comparison between different satellites in different regions of the Earth’s environment, we will get much better information on how the magnetosphere reacts as a whole to, for example, changing conditions in the Solar wind.

The CRRES mission will be crucial for spacecraft design engineers. First it will discover which type of electronic components are safe to use in orbits ranging from low-Earth orbit (an altitude of a few hundred kilometres) to geostationary (36 000 kilometres). Secondly, it will provide a comprehensive data source with which to update our ageing models of radiation in space. And, finally, it will gather information for future dynamic radiation models.