Observation of Meteoroid Impacts by Space-Based Sensors
Mercury, November/December 1998 Table of Contents
Edward Tagliaferri
The Sun was already up as the two fishermen made their way out of the harbor from the tiny island of Kosrae, in Micronesia, on a Tuesday morning. It was going to be a beautiful day. By mid-morning, the day was bright and clear as only the South Pacific can be. Suddenly, without sound or warning, the day got brighter -- as bright as if there were now two Suns in the sky. A brilliant flash lit up the Sounth Pacific sky with more energy than the atomic bombs that destroyed the Japanese cities of Hiroshima and Nagasaki in World War II. The fishermen, who felt as though the detonation had gone off over their heads, were actually more than 300 kilometers from the location of the blast.
What had happened was that an asteroid, a rock only about ten meters in diameter, traveling more than twenty kilometers per second, had struck. As it entered the Earth's atmosphere, it quickly encountered aerodynamic pressures which exceeded its structural strength and had literally exploded with an energy equivalent to fifty to seventy kilotons of TNT.
Frightened, the two fellows looked up and saw the huge track across the sky over their heads, like the contrail of a giant jet. They quickly gathered up their gear and headed to port, where they were to subsequently report what they had seen to the Minister of the Interior as something "awesome and frightening."
Unknown to the fishermen, they were not alone. High above them in space, the United States's Silent Sentinels, infrared sensors aboard satellites operated by the Department of Defense, had also detected the event and had instantly flashed the message to waiting operators on the ground. Simultaneously, visible wavelength sensors, part of the U.S.'s space-based Nuclear Detonation Detection System (NDS), had detected the flash and sent the report to an operations center.
For several frustrating weeks the Interior Minister would make inquiries around the world asking if any one could explain what had happened. It would be some time before the answer could be told.
Orbit of an Earth Crosser. Members of the Apollo asteroid group have orbits that cross Earth's, illustrated here in a view from above the ecliptic, yet those orbits may be inclined to that of the Earth. Nonetheless, a significant portion of the Earth Crossers will eventually strike Earth; energy of the impactors is usually deposited in Earth's atmosphere. Illustration by Donna Wolke and adapted from one courtesy of author.
This event actually happened.
While most asteroids are in relatively benign orbits around the Sun between Mars and Jupiter (in the so-called Asteroid Belt), every so often something happens to perturb an orbit to send an asteroid careening in towards the Sun. It could be a collision with another asteroid or the perturbing effect of Jupiter. We are not quite certain what all the causes of these perturbations are, but the result is that a stable, nearly circular orbit between Mars and Jupiter becomes highly elliptical, with a perihelion inside the Earth's orbit and an aphelion at or outside Earth's orbit.
An object with these new orbital characteristics has just become an "Earth Crosser," crossing the Earth's orbit twice each time it goes around the Sun. This is presented in the accompanying figure, which shows the geometry as seen from above the plane of the Earth's orbit. Of course, what is not apparent from the figure is that the orbits of the Earth Crossers can be inclined to that of the Earth; even if these asteroids appear to cross the Earth's orbit when the Earth is there, there could still be a significant gap between the two if the Earth Crosser's orbit is above or below the Earth's at the time of crossing. Eventually, a significant fraction of the Earth Crossers will impact the Earth. When one does hit, the energy it carries is usually deposited high in the atmosphere when the object detonates.
That is exactly what happened at 9:38 a.m. on Tuesday, 1 February 1994, over the South Pacific. This asteroid actually broke into several pieces, with one large piece detonating about 34 km above the surface, and a second, much larger piece detonating at an altitude of approximately 21 km. The data collected and recorded by the space-based sensors permitted a detailed analysis to be performed which gave the trajectory of the asteroid, its velocity, the energy it delivered, the locations of the detonations in three dimensions, and, by careful modeling to reproduce the observed lightcurve, an estimate of the object's size and make-up. The U.S. Departments of Defense and Energy, recognizing the scientific value of the information, permitted publication of the results of the analysis in the February 1995 issue of the Journal of Geophysical Research - Planets.
Terrestrial Asteroid Impacts
Objects from space are continually hitting the Earth. However, the Earth's atmosphere is a wonderful shield, and stony objects smaller than about sixty meters in diameter rarely reach the surface intact. Most of the objects hitting the Earth are very small and burn up completely while still high in the atmosphere. One can go out on any clear night of the year and see the flashes of light produced as they burn up. Such objects, popularly called "falling stars," have been the object of wonder and of the poet's pen for millennia. Larger objects, say bigger than one meter in diameter, behave just as the 1 February 1994 object. They remain intact until they encounter significant aerodynamic pressure, and then they break apart. Large pieces of stone with some structural strength will experience surface heating and ablation, but the interaction with the atmosphere happens so quickly that the main body remains cold. Thus, one often hears reports that "the object was warm to the touch when first found, but then became cold and frost formed on its surface." When the aerodynamic pressure on the entering body exceeds it structural strength, the object undergoes what can only be described as an explosive disintegration high in the atmosphere. The effective area of the body then increases dramatically, with a simultaneous dramatic increase in the transfer of energy to the atmosphere. The result is what is often referred to as a "bolide" or "fireball," as the rapidly heated air and vaporized material from the object create a shock wave and a brilliant flash of light. After the detonation of the meteoroid, the remaining pieces fall to the ground, becoming meteorites.
Meteoroids composed of iron are much stronger than the stony variety and can reach the ground intact even if only a few meters in diameter. Iron objects also reach thermal equilibrium very quickly and remain hot to the touch after hitting the ground. Such small iron objects and stony objects larger than about sixty meters in diameter can reach the ground intact. Upon impact, the kinetic energy of the object is so large that a major fraction of the object is vaporized, along with material in the area of the impact, and a crater is created. At the same time, a large amount of the in situ soil and rock is ejected from the crater, which subsequently settles back to Earth around the crater. For example, a nickel-iron meteoroid about 50 meters in diameter is suspected of having created Meteor Crater in Arizona. Searches of the crater for the main body of the object have never turned up anything but small pieces.
Even a small impactor leaves a big hole. About 50,000 years ago an object estimated to have been 50 meters in size struck the Earth in what is now Arizona. Traveling at 18 km/sec, the massive object hit the ground and produced a blast equivalent to the detonation of 20 megatons of TNT. What is left today from the impact is a crater over 1.5 kilometers across and 170 meters deep. Image courtesy of Meteor Crater, Northern Arizona, USA.
As the impacting object gets larger, the consequences of impact become increasingly more severe. Earthquakes, shock waves, and heat can cause death and destruction up to several hundred kilometers from the impact site. It has been estimated, for example, that an object several hundred meters in diameter impacting at about 20 km/sec would cause a magnitude 12 earthquake, something never experienced in recorded history. Such an object impacting in the ocean, as in the recent movie "Deep Impact," would create a tsunami more than a kilometer high, moving out from the impact point at almost 800 kilometers per hour (the movie actually understated the destruction from the impact, completely ignoring the effects of the tsunami on the U.K. and Europe). Larger objects can carry so much kinetic energy that the effects of the impacts are no longer local, but become global.
It has been hypothesized that the major extinction of life that occurred at the Cretaceous/Tertiary (K/T) boundary 65 million years ago was caused by the impact of an asteroid about ten kilometers in diameter. The object vaporized on impact, creating a crater 200 km to 300 km across and ejecting trillions of tons of soil, rock, and soot into the atmosphere. The larger pieces rapidly fell back to Earth, causing the atmosphere to become incandescent and starting fires on a global scale. As the cloud of soot and dust remaining in the atmosphere began to circulate around the globe, it blocked sunlight from reaching the surface. Eventually the planet was shrouded by a cloud of dust so thick that the surface was plunged into darkness for a period of months or years. What followed was a mass extinction, in that the plants died first, then the plant eaters, then those who ate the plant eaters. Depending on how one counts, up to seventy percent of the species then living, both on land and in the sea, became extinct, including the dinosaurs.
The end of an age. At the geological border between the Cretaceous and Tertiary periods, evidence exists to support the hypothesis that an object ten kilometers in size struck the Earth. The impact and subsequent effects led to the extinction of the dinosaurs and a large fraction of other terrestrial life. Painting by Don Davis and courtesy of NASA.
It is clear from the above discussion that the rate at which objects impact the Earth and the range of sizes one can expect to encounter are things we would like to know. Estimates have been made of the expected impact rate of objects as a function of their size. The late Gene Shoemaker, for example, used the craters on the Moon and the Earth to estimate the rate of impacting objects as a function of their size. An object the size of the one that impacted on 1 February 1994 can be expected about once every ten years or so. An object with an energy equivalent to that of the atomic bombs of World War II is expected to strike the Earth once a year. Yet one seldom hears of such large explosions happening, let alone such events occurring once a year. Why is this?
The answer is that 3/4 of the Earth's surface is covered with water, and a significant fraction of the land area is uninhabited. Hence, there is no one there to observe the impacts, and so no one to report them.
The View From Above
It is easy to imagine that, without the satellite observations, the February 1994 event would have been relegated to an interesting anecdote from a couple of fishermen. However, the satellites were there and the observations enabled us to perform a rather detailed analysis of the event.
Over the past twenty years or so, a number of meteoroid impacts have been detected by space-based sensors. Infrared sensors on spacecraft operated by the U.S. Department of Defense have detected and recorded over 400 such impacts since 1972. These sensors were designed for military missions, and, therefore, it is not possible to describe them in detail in the open literature. At this point in time, the only description permissible is that the satellites have scanning sensors which operate in the short wavelength infrared. The number of sensors and their deployment is such that essentially the entire world is observed by them 24 hours a day, 7 days a week. However, this is not a perfect world in which we live, and the system has limitations.
In operation, the sensors are connected in real time to very large, very fast computers. These computers are able to scan the sensor data and, using extremely complex algorithms, extract details of events of interest to the operators. The remainder of the data is recorded and saved for a short period of time. If after that time, there is no call for saving the tapes, they are recycled. The problem lies in the phrase "events of interest to the operators."
Meteoroid impacts are generally not of interest, and, therefore, the data are generally not recorded for sensor-detected impacts. The exceptions are impacts that release large amounts of energy (e.g., large objects like that which produced the February 1994 event) or that are in areas of the globe where any detections are of possible interest; in those cases the data are recorded for future analysis. The result is that not all detected events are recorded.
The first recorded event was an Earth grazer that "impacted" on 10 August 1972. This object entered the Earth's atmosphere over Utah, traversed several thousand kilometers in the atmosphere on a northerly heading, and left the atmosphere over Canada. The ground track of this object is shown in the related figure. Later analysis of this event showed that the object was an Apollo asteroid about ten meters in diameter that was traveling about 15 km/sec when it entered the atmosphere. It was first detected by satellite at an altitude of about 73 km, tracked as it descended to about 53 km, and then tracked as it climbed back out of the atmosphere. The event garnered national interest because it was a very bright, daylight fireball seen by hundreds of people on the ground. There were many still and moving pictures taken of the object as it tracked across the Grand Tetons, and media attention was high. This object is still in an Earth-crossing orbit around the Sun and passed close to the Earth again in August 1997.
Path of a bright fireball observed over the northwestern U.S. and western Canada on 10 August 1972. Analysis has revealed that the object, which passed into the atmosphere and eventually headed back out into outer space, is an Apollo asteroid ten meters in size. Illustration by Donna Wolke and adapted from one courtesy of author.
The next events don't appear in the sensor record until 1975, when the operators became aware that the data are of value, and events began to be recorded more often. These were still not events of operational interest, and the number of recorded events in the early years reflects that fact. In 1975 only six events were recorded. Over the first seventeen years, the number of events recorded averaged about eight per year, with some years having no events recorded at all. The number of events recorded per year has steadily increased until, over the last three years or so, about fifty events per year have been recorded. The 336 events recorded between 1972 and July of 1997 are shown in the accompanying figure. As can be seen, the distribution of events is more or less uniform over the globe. In the figure, the open circles represent daytime events (i.e., events which occurred in local daylight), and the solid circles represent nighttime detection's.
A different type of impactor. Only about 25 micrometers in size, the tiny meteoroid (bright spot at top of "swoosh") left a one-millimeter-long trail as it penetrated exposed gel on the European Recoverable Carrier, a spacecraft deployed by the Space Shuttle. At high velocities, even small objects can carry much kinetic energy. Image courtesy of ESA and NASA.
The nature of the sensors and the manner in which they are operated cause a slight bias towards daytime detections. That is because, for large objects, there is often a dust cloud produced by the detonation of the object. When the geometry is right, sunglint off of these dust clouds can be detected, revealing an impact even if the impact itself is missed. At night, either the event is detected, or it is missed.
Visible Wavelength Detectors
In recent years, space-based, visible wavelength detectors operated by the U.S. Department of Energy as part of the Nuclear Detonation Detection System (NDS) have come on line. These sensors have demonstrated that they too can detect the detonation of relatively large meteoroids in the atmosphere. The visible wavelength sensors are not as sensitive as the IR sensors discussed above and so detect fewer events. However, these sensors are calibrated and, hence, can give a direct indication of how much energy is deposited in the atmosphere by the detonation of an object. They are instrumented such that they do not detect objects which do not detonate, and, therefore, they would not have detected the Earth grazer of August 1972, for example. They also give poor location information. For comparison, the IR sensors are not well calibrated but give excellent location information. In addition, the IR sensors can often collect enough data to give a velocity for the object.
Given the velocity and the energy deposited, it is possible to deduce a rough mass of the impacting object. By assuming a mass density, it is also possible to estimate the object's size. For example, theorists were able to model the 1 February 1994 impactor and deduced that it was most likely a stony object. The energy of the detonation, estimated from data from the visible wavelength sensors, was equivalent to between fifty and seventy kilotons of TNT. The velocity determined from the IR sensor data was about 23 km/sec, from which it was possible to deduce a size of roughly ten meters diameter. The IR sensors also showed that, although the object broke up into several pieces, there were two pieces that deposited the bulk of the energy: The smaller fragment detonated at an altitude of about 34 kilometers; the larger one, at about 21 kilometers.
From the small sample of visible wavelength sensor events we have, it appears that the objects we are detecting with the IR sensors range in size from about one meter in diameter and up. The largest recorded event to date is that on 1 February 1994; the responsible object is estimated to be about ten meters in diameter (although we could be as much as a factor of 2 off at either end of the scale).
Geographic location of meteoroid impacts detected by space-based infrared sensors from August 1972 through July 1997. Solid circles represent night-time detections, and open circles represent impacts detected during daytime. Plot courtesy of author.
Looking Up, Looking Down
The utility of space-based sensors has been demonstrated. They provide global coverage, are not affected by weather, and, for events in the detectable size range, it is often possible to extract enough information from their data to determine the velocity, trajectory, and energy deposited by an impacting extraterrestrial object. Further, analysis of those data also permit us to estimate the size and composition of the impactor. Moreover, the data have often been of sufficient scope and quality to permit the pre-impact orbit of the object to be determined.
What happens when big things hit planets. In July 1994, fragments of Comet Shoemaker-Levy 9 struck Jupiter's atmosphere. In this 2.34 micron image obtained by Australian National University's 2.3 meter telescope at Siding Spring, we see the fireball that resulted from Fragment G. Image courtesy of Dr. Peter McGregor & Mark Allen, Research School of Astronomy and Astrophysics, Australian National University.
An unexpected turn of events is that the combination of ground-based observations with space-based observations has yielded much more information than either set of observations alone. And as our communication with the astronomical community improves, we hope to develop a productive alliance which exploits both.
The sensor systems discussed in this article are limited in capability. The restrictions under which they must operate constrain the timeliness of dissemination of the data and sometimes constrain the types and amounts of information which can be released. Nevertheless, the satellites provide a unique and extremely valuable source of data from which we can validate and refine the models of the flux of objects in the vicinity of the Earth. These models, in