Dr. Marc Rayman’s Deep Space 1 Mission Log Mission Update

Thank you for visiting the Deep Space 1 mission status
information site, now in its fourth year on the list of
most frequently visited sites in the observable universe
for information on this solar system exploration mission.
This message was logged at 5:45 pm Pacific Time on Sunday,
November 11.

After introducing another member of the solar system family
to Earth, Deep Space 1, the little spacecraft that could —
and did! — continues flying contentedly in its orbit
around the Sun. Meanwhile, scientists are analyzing the
fantastically rich harvest of data returned from the
historic encounter with comet Borrelly. More than two years
after the end of its 11-month primary mission, on September
22 DS1 stepped up to its greatest challenge of all with the
elegance and skill of a true master. The encounter
certainly did not go the way I expected — instead,
everything went perfectly!

The scientific analysis of the visible images, infrared
spectra, ion and electron energy and angle spectra, ion
composition measurements, magnetic field measurements, and
plasma wave burst data will go on for quite some time.
After just the initial impressions of a subset of these
data were described in a press conference a few days after
the encounter (see the Deep Space 1 press conference), the
real analysis began. DS1’s images are the only ones in
existence that are detailed enough to allow geological
analysis of the nucleus of a comet. These images are still
being processed to bring out additional details not
discernible in the raw images releases so far. It literally
will take years to mine everything from these data, but
preliminary results will be announced in press releases at
the end of November. They will contain some fascinating and
exciting news, but in order not to steal any of the
impending thunder, let’s focus instead on what happened in
the time between that last two logs (available as a special
discounted gift set at fine establishments throughout the
halo of the Milky Way), as the spacecraft closed in on its
quarry. Following that admittedly somewhat dry material,
we’ll turn back to more of the human experience.

As all corporeal readers know, DS1 has had to thrust with
its ion propulsion system even when we did not want it to
change its course. This served to reduce the consumption of
its conventional chemical propellant. As a result, DS1
thrusted at impulse power back and forth for many months,
tacking its way to Borrelly. The push of the ion engine, as
delicate as it is, adds up to problems for the navigators
who try to predict DS1’s course with the accuracy needed
for the encounter. (During its normal interplanetary
travels, the subtle uncertainties that arise from the
continuous acceleration are too small to be of concern.) As
planned for many months, DS1 stopped firing its ion engine
on September 15 and coasted most of the rest of the way to
the comet. Following almost 15 months of being in powered
flight nearly 100% of the time, the ship was silently
drawing near its destination.

A total of 11 times from August 25 until about 10 hours
before the closest approach on September 22 (still at a
range of over 600,000 kilometers (around 375,000 miles), or
more than 1.5 times the distance between Earth and the
moon), DS1 took images of where it expected Borrelly to be.
These distant views were used to improve estimates of the
location of the comet. Although it has been observed many
times from Earth since its discovery in 1904, as with all
astronomical bodies there are significant limitations in
astronomers’ ability to pin down the orbit. But by
combining data from Earth-based observations of Borrelly
with DS1’s views of the comet as they raced toward their
eagerly awaited appointment, it was possible to get a
better estimate of the comet’s location. These observations
were complex, but, to our great relief, all of them worked
just the way they were supposed to. At first, the comet was
so distant that many images had to be electronically
combined for the comet even to be detected. What appeared
in the images was the coma (the vast cloud that cloaks the
nucleus in gas and dust) and the tail.

As DS1 closed in on Borrelly, these images were used by
navigators to compute course corrections. Before routine
thrusting stopped on September 15, the trajectory was
altered by changing the planned direction and throttle
level of the ion drive. After September 15, the corrections
were accomplished by firing the engine only at specially
selected times and throttle levels.

Nearly every activity on the spacecraft represents an
opportunity for a problem to arise. As just two examples,
when the probe turns, it might have trouble locking to a
new reference star, or any commands sent from Earth might
contain an error. (The extremely small mission control
team, despite the excellence shown during three years of
complex and successful flight, could — gasp! — make a
mistake.) The closer the spacecraft was to the comet, the
more important it was to avoid actions that, if they did
not go as planned, could compromise the encounter. Less
time to recover from problems meant it was more important
than ever to avoid them.

We devised a clever strategy over the summer that made it
likely (but not guaranteed) that most course corrections
would require the spacecraft to thrust while it was in
nearly the same orientation needed for communicating with
Earth. In fact, this worked so well that all but one of the
course corrections required thrusting in exactly that
orientation. That meant that the spacecraft did not have to
execute extra turns and did not have to expend extra
hydrazine. When it came time to modify the trajectory, we
simply told the spacecraft the duration and power level,
and it dutifully and calmly executed the thrusting as
needed. We could communicate with it at the same time,
uploading more of the many files it would need for the
encounter and monitoring its health to be sure no problems
were brewing that might interfere with its chances of
collecting at least some of the data we sought at Borrelly.

One of the reasons that so much effort had been devoted to
saving hydrazine was that during the last day before the
encounter, we expected not to be able to use the ion engine
for course corrections. The ion engine delivers what I’ve
often described as “acceleration with patience,” but with
only hours before closest approach, patience was not a
virtue: the spacecraft needed to point its antenna to Earth
most of the time. But our strategy paid off handsomely, and
we were able to make the final corrections by firing the
engine with no turns at all. As a result, we did not have
to use any extra hydrazine, and no time was lost in using
the reliable and efficient, if leisurely, ion engine.

A few days before the encounter, our old friend PEPE was
activated, its software loaded, and its operation verified.
Its job would be to try to measure the composition,
energies, and directions of the charged particles in the
coma as well as how the comet and the solar wind affect
each other.

Myriad other preparations were conducted in the remaining
few days, including finalizing plans for what to do in a
variety of unplanned circumstances too nerve-wracking to
try to recall if I want to sleep well tonight. In brief,
however, we filled up the available time working as hard as
we could to give the spacecraft its best chances for
success.

Everything went surprisingly smoothly in the days leading
up to Saturday, September 22. I had had myriad concerns
before the encounter. Even casual readers of these logs
know that I did not have high confidence in this daring
undertaking, and many logs over the past few months have
described different aspects of the risk. As I ate breakfast
well before sunrise on Saturday, one of my fears was that
everything would continue to go well but for one mistake,
one oversight, one simple little thing that we should have
done or not done. The encounter would be conducted with 685
stored instructions, containing nearly 4000 parameters,
relying on complex software that had been used for tests
but never for visiting a comet. That represented an
extraordinary number of opportunities for a mistake. I
prepared myself for coming home dejected that night,
expecting to scold myself for missing that one lurking
error. It would be no worse than the entire encounter going
poorly, but just one simple error would make it easier to
devote excessive energy to repeating “If only…” for years
to come.

It was an almost eerily calm day in DS1’s mission control
room, as the spacecraft continued to be well behaved. Most
members of the team had little to do but make sure the
spacecraft was healthy. We did have several critical
decisions to make, based on plans we had worked out
carefully during the preceding year. When the final
pre-encounter images were obtained, we analyzed them and
our earlier ones with a mathematical model of how the
brightness of the scene formed by the combination of the
vast coma plus the tiny nucleus should change as the
distance to the comet diminished. This allowed us to make
our selection for the camera exposure times and to
formulate our final estimate of where the spacecraft should
begin looking for the nucleus.

Some team members suggested we make other, unplanned
changes, but that is always dangerous. It is easy for late
anxiety (augmented with the burritos we had for lunch) to
foster new ideas that, in the absence of calm reflection,
may seem meritorious. But making changes to such an
intricate plan in the final hours requires very careful
consideration, and, by definition, there probably is not
time for that. We did discuss some ideas but opted instead
to trust the more considered judgment we had exercised
earlier and stayed the course.

Shortly before 1:30 pm PDT, signals confirmed the
spacecraft had begun turning. The main antenna would not
point to Earth again until after the encounter (if the
spacecraft survived), and we had only very limited signals
with which to infer its progress. The spacecraft had a
tremendously complicated plan to follow, including locking
to a reference star for a while, then trying to obtain some
views of the nucleus while still more than 85,000
kilometers (53,000 miles) away, then trying to point its
very narrow-view infrared spectrometer at the nucleus. Next
it had to lock to another reference star in a special
location that would provide it information it would need
later in the final encounter. Finally, about 35 minutes
before its closest passage by the nucleus, still 35,000
kilometers (22,000 miles) away, it turned to the
8-kilometer (5-mile) long nucleus for the final time. It
began taking two images per minute in order to try to find
it and lock on so it could track the mysterious core of the
comet. The elaborate choreography continued with many
changes in spacecraft modes and constant measurements by
PEPE and by the reprogrammed ion engine sensors, smelling
and hearing phenomena in the coma as the camera tried to
record the sights.

The indications we had on Earth were that everything was
going well, but that could have been deceptive. If the
signals indicated the spacecraft was having problems, we
could have trusted that. But its belief that it was
tracking the nucleus was not proof that indeed it was;
there were many ways it could fail and not realize it.
Still, it was reassuring that no problems were evident. We
later determined that of the 53 pictures the spacecraft
took, it managed to identify the nucleus in 52 of them. In
one of the pictures, a cosmic ray that struck the
electronic detector in the camera left a track that got
most of the way through the various software guards meant
to eliminate spurious signals. Although it prevented the
system from finding the nucleus in that one image, it did
not disrupt or confuse the attempt to track the nucleus;
rather, the software ultimately discarded the picture,
refusing to be fooled by the deceptive information it
contained.

There was tentative applause at JPL when signals showed
that the spacecraft had traveled half-way through the coma,
completing its closest approach to the nucleus. But it
still had to survive its trip back out of the coma, with
potentially fatal dust impacts and more complex maneuvers.
Finally the spacecraft turned to point its main antenna
back to Earth, and we waited like expectant children
listening as a masterful story teller begins to unfold a
tale of daring, mystery, and adventure.

As the spacecraft regaled us with its spine-tingling
exploits, we gathered around a few of the monitors on which
the pictures would first be displayed. We already had good
reason to believe that the tremendously important PEPE and
ion propulsion system diagnostics sensor data had been
acquired. They would reveal much of great interest about
the comet. But, regardless of our technical or scientific
interests, the roughly 100% human controllers are visual
creatures, and we frankly hoped for a cool picture. Our
goal had been to get a picture in which the nucleus spanned
50 pixels (a pixel is the smallest element of the digital
camera’s view). This would be of great scientific value and
would be good enough to give us a feel for what this
completely unknown body looked like. The images were
returned in a special order, but not in the order in which
they were taken, Still, the first images we saw had been
taken from so far away that the nucleus was still small,
and the scene was dominated by a powerful jet of dust
(these images were from farther away even than the fourth
image. These showed that the comet was an unfamiliar and
strange place indeed.

Why didn’t the dust destroy the spacecraft (see the
September 23rd log)? It appears that our prediction of a
few hundred dust impacts, based on analysis of Earth-based
images of Borrelly, was tricked by this powerful jet. Earth
was too distant for the jet to show up; all that could be
inferred was the total amount of dust in the vicinity of
the comet. But by being concentrated in a jet that DS1 did
not fly through, it left other regions less dangerous.
While we know DS1 was hit, it did not experience enough
blows to suffer damage. Indeed, the only effect of the
encounter we have been able to identify is the appearance
on the spacecraft now of a big grin!

After a few additional distant images were returned, the
first image on theDeep Space 1 Press Releases/Images page
appeared on the monitors at about 5:30 pm, and the real
celebration of the Borrelly encounter began. The spacecraft
had managed to track the nucleus better than we had hoped
and took a picture when it was close enough that the body
was about 170 pixels across — more than 3 times better
than our goal. The purity of the human joy that I shared
with my colleagues there, as cheers and applause erupted in
mission control, is something I will never forget. On
behalf of our curious and noble species, we beheld the
first detailed views not only of a place, but of a kind of
place, never seen before. Our spirits soared to heights
unachievable even with ion propulsion! Your ever-devoted
correspondent, normally of at least average eloquence,
found himself unable to say little more than “I just can’t
believe how incredibly cool this is” every 30 seconds for
the next few hours.

In the two years following the end of the primary mission,
we had made many thousands of difficult decisions,
particularly, but by no means exclusively, because of the
failure of the craft’s star tracker. With the very small
budget for Deep Space 1 (indeed, it is the lowest cost
interplanetary mission NASA has ever conducted), many times
we simply did not have the resources to analyze problems in
as much detail as we might have liked. With a small team
and a very complex mission, too often we found ourselves
having to choose which problems we would penetrate. For the
others, it generally became necessary to go with our best
estimate through a combination of specific and limited
technical information and a strong dose of human judgment.
But what if we had made a wrong choice in which areas to
focus our greatest attention, or what if the less well
considered decisions proved to be wrong in an important
way? Well, in that case, I wouldn’t be writing about the
jubilation that followed a truly flawless encounter.

Coming at a time when so many of us were witness to the
most shocking human actions and were forced to confront our
greatest fears, we felt that we were taking humanity’s
highest ideals to its greatest reaches. More than just an
incremental step, in that one day we made an astronomical
jump forward in our cosmic view. While our grand news may
have been largely lost in the midst of these other
terrestrial events, we were proud to accomplish something
beautiful, surprising, and inspiring on behalf of everyone
who has ever wondered about the universe.

Deep Space 1 completed its primary mission in 1999 and its
extended mission this autumn. So what is left? The
hyperextended mission, of course. Beginning in October, our
attention shifted to retesting many of the technologies
that were the reason DS1 was built and launched. Nine of
the 12 technologies on board are hardware (three were
autonomous software systems), and each is being exercised
more during this phase of the flight, as we return DS1 to
its roots. With a mission that was intended to last 11
months, the opportunity to test these systems after three
years in space (celebrated a few weeks ago with a yummy
cake displaying the gleeful proclamation “3 sweet years!”),
with greater exposure to radiation and other hazards of the
space environment, many large swings in temperature, and
other possible sources of wear, this is an opportunity to
add still more to our understanding of these systems that
are important for reducing the cost and risk of ambitious
space and Earth science missions of the future.

The focus of the hyperextended mission is on the ion
propulsion system, and we are performing many tests to
quantify the effects of its having operated for so long. We
are also testing it in various modes that would have been
too risky or otherwise inappropriate earlier in the
mission. This amazing system has provided the equivalent of
about 4.2 kilometers/second (9400 miles/hour) to the
spacecraft, while consuming less than 70 kilograms (157
pounds) of xenon propellant. The system has accumulated
more than 640 days of thrust time. (The requirement for
“minimum mission success” for DS1 included operating the
ion propulsion system for 200 hours. We have only exceeded
that by a factor of 77; but don’t despair, still more hours
of operation are ahead.) The results of these tests will
represent still greater bonus from the mission as it
continues blazing trails in space exploration.

The hyperextended mission will conclude in December, and
the next log will describe what fate awaits the aged,
wounded, intrepid, and very very happy explorer.

DS1 is now over 68 million kilometers, or 42 million miles,
from its new friend comet Borrelly. As they continue on
their separate ways, we can be sure each will retain a fond
memory of their brief meeting, a special moment of shared
discovery in their very different solar system journeys.

Meanwhile, DS1 and Earth are continuing to get closer in
their individual orbits. Since launch in October 1998, DS1
has completed two orbits of the Sun while Earth has
completed three. By lapping the craft, Earth is now
catching up again. (See the June 30, 2001 log for more on
DS1’s orbit.)

Deep Space 1 is almost 1.2 times as far from Earth as the
Sun is and over 460 times as far as the moon. At this
distance of 177 million kilometers, or 110 million miles,
radio signals, traveling at the universal limit of the
speed of light, take over 19 and a half minutes to make the
round trip.

Thanks again for visiting!

P.S. For those of you who miss listening to these reports,
I apologize for the ongoing unavailability of JPL’s mission
status recordings. Youngsters, who have long enjoyed
visiting The Space Place , now can call in to The Space
Place’s new toll free recording at 866-575-6178.