Why launch in the middle of the night? Blame Johannes Kepler and Isaac Newton and the engineers who designed the Chandra X-ray Observatory.
It starts in orbital mechanics, a sometimes obscure but always essential part of space travel. The field is rooted in the earliest days of modern astronomy, when Kepler discovered that planets moved in well-regulated elliptical paths and Newton discovered that gravity is universal. Newton also invented the calculus that became the foundation of the complex math used in planning satellite orbits.
But we can't always use it to avoid working late at night.
Chandra's night launch time was set by several factors, starting with the need to protect the telescope's science instruments from damage in the Van Allen Radiation Belts.
To avoid most of the belts, Chandra would have to be placed in a highly elliptical orbit. That makes tidal forces a bigger concern than they would be for a satellite in low Earth orbit, explained Dr. Jonathan McDowell, a scientist with the Chandra Science Center in Cambridge, Mass.
"It was not picked for any astronomical reason. It was picked for orbital lifetime as determined by lunar and solar perturbations," he said. Highly eccentric orbits will have their perigee (point closest to the Earth) shifted up and down by tidal forces from the Sun and Moon. "The trick is to choose an orbital plane where you ensure a long orbit."
McDowell said engineers at TRW calculated the best combination to give Chandra a long orbit. NASA's baseline mission is 5 years, and the spacecraft has resources to last 10 years. TRW engineers worked out an orbit that should last 30 years. It will start out ranging from 10,000 to 140,000 km above the Earth (6,200 x 86,800 mi), almost a third of the way to the Moon at apogee (the high point). The orbit will be inclined 28.45 degrees to the equator, the same as the latitude of the launch site, Kennedy Space Center. The selected orbit is the result of tradeoffs involving Chandra's weight and rocket power available from the Space Shuttle, Inertial Upper Stage (a 2-stage solid rocket), and the observatory's onboard Integral Propulsion System.
Chandra's orbit has a period of 64.2 hours, with about 15 hours of that inside the Van Allen Radiation Belts, forcing operators to shut down the science instruments while the observatory is below 60,000 km (24,000 mi) altitude. Essential chores like repointing the observatory and moving gratings into and out of the light path that would disrupt science observations will be scheduled for the passes through the radiation belts.
Large orbits are nearly fixed in inertial space. Just as the Earth and the rest of the planets have orbits that keep virtually the same orientation relative to the galaxy, with only gradual changes over long periods, Chandra's orbit will hold the same orientation.
That also means that when Chandra is near apogee, the Earth will always be in the same section of Chandra's sky. Because Chandra can't look within 20 deg. of the Earth (an angle about 40 times larger than the size of the Moon as we see it from Earth), the same small chunk of sky will be largely off limits for a long time.
"It's part of the price you pay for being in a deep orbit," McDowell said.
The obscured zone covers an area from 4 to 11 hours right ascension and -10 to -50 deg. declination, roughly opposite the galactic center. It's not a total blackout, just a sharp enough reduction in observing time in that area that little science of value would be achieved.
Fortunately, McDowell noted, that part of the sky has a few choice targets. McDowell noted that planners for Europe's Infrared Space Observatory actually did have a choice, a tough one, when they opted for an eccentric orbit. They had to give up the galactic center or the Orion nebula, both areas of remarkable activity in infrared light. They chose the galactic center, but ISO lasted long enough for its orbit to shift and let them view Orion.
How does all this lead to a launch at around midnight?
Simple. Imagine Chandra's final orbit as an ellipse moving with the Earth around the Sun. Even as the Earth spins, the ellipse stays pointed in the same direction. Launch and all the intermediate steps to get to the final orbit all lie within this ellipse.
An additional factor, McDowell explained, is that Chandra has strict power requirements, so it must be oriented to expose its solar arrays properly during maneuvers to its final orbit. Furthermore, the upper stages have a horizon sensor that must look in a certain direction. These two bits of spacecraft geometry are what finally set Chandra's launch time relative to the stars, not human work schedules.
For a launch on July 20, 1999, Kennedy Space Center rotated into position on the ellipse's plane at 12:36 a.m. EDT. For every day of delay, McDowell noted, Chandra's launch moves up by 4 minutes. If Chandra was launched around Jan. 20, launch would be in the afternoon 12:36 p.m.
Similar needs drive launch times for other satellites, such as weather satellites that have to cross the equator at the same time each morning and afternoon, or communications satellites that have to arrive at a certain place over the equator.
These aren't the only pointing considerations for Chandra. It can't look within 45 degrees of the sun or 6 degrees of the Moon, lest their brightness confuse the visible light aspect cameras that point Chandra at the observing target. But these are temporary blockages that change during the orbit and the year. If the target a scientist wants to observe is blocked by the Sun now, his turn will come in a couple of months when Earth's orbit carries it and Chandra into a more favorable position. At any given time, up to 85 percent of the sky will be available for observations.
Even so, the blocked area eventually will become available. The same forces that make tides rise and fall will slowly shift Chandra's orbit, so that the blocked area is clear after a few years - but at the expense of blocking something else.
Those forces will also bring Chandra home. McDowell said that after the perigee is shifted higher by solar perturbations, it will be shifted lower so that by 2029, give or take a few years - and long after it has ceased operations - Chandra will become a shooting star blazing through Earth's atmosphere.