Tuesday 28 January began with a morning alarm call at 0645 Central Standard Time and the news that, during the night, there had been two delays in countdown for the launch of mission 51-L, the 鈥淭eacher in Space鈥� mission. One had involved a problem in loading fuel; the other concerned the unpleasantly low temperatures at the Cape. A crew was busy at the pad removing icicles. Launch would be two hours later than scheduled, but everything looked good for a go at 1038.

Most of the press regulars were on hand when I arrived at Johnson Space Center鈥檚 media building at 0900. I caught up on overnight news of the countdown, and went on the air at about 1000. We were in the routine 10-minute hold at T minus 20 minutes. I was slightly surprised that there was no hint of extending the hold as the icicle removal team was still at work on the pad, but they did indeed leave by the time the count was resumed. Everything ran smoothly to the last scheduled hold at T minus 9 minutes at 1019. We came out of that hold at 1029 and, at that point, I shut off my microphone and let Hugh Harris, NASA鈥檚 public affairs commentator at the Kennedy Space Center in Florida, do the talking. Having covered 16 previous launches with the National Space Institute鈥檚 Dial-A-Shuttle service, which provides a telephone commentary throughout each flight, I knew that he could talk almost continuously.

My colleague Pete Kappesser joined me at the Dial-A-Shuttle console and we sat back to watch the launch on the monitors in front of us. We had a smooth lift-off, on time at 1038. The programme to put the orbiter on the correct trajectory was started on cue. Houston was controlling, and Capcom Jim Weatherbee gave the expected call 鈥淕o at throttle up鈥�, which acknowledges that all three main engines are performing well and that they can go up to 104 per cent power.

One minute and 8 seconds after the launch Commander Dick Scobee replied in acknowledgement: 鈥淩oger, go at throttle up.鈥� It was the last the world heard from him.

We continued to follow the pictures on the monitor. I felt vaguely uncomfortable about the far solid rocket booster. Was it sputtering a bit, or were the two boosters not performing identically? It was a brief, gut feeling, filed in an instant.

At a 鈥渕ission-elapsed time鈥� of 1 minute and 13 seconds there was a crackle on the air-to-ground communications channel. In the moment it had taken for my uneasiness to register, something had gone terribly wrong.

Catastrophe seems frozen in time. The solid rocket boosters had separated prematurely: they were not due to separate until 2 minutes 7 seconds into the flight, but the mission-elapsed clock in front of me registered 1 minute 14 seconds. In mission control, the moment became unreal because Steve Nesbitt, the public affairs officer, was watching his data consoles and not the pictures on his monitor. 鈥�1 minute 15 seconds鈥� he intoned, 鈥渧elocity 2900 feet per second, altitude 9 nautical miles, down range distance 7 nautical miles.鈥� (In my mind I鈥檓 saying, 鈥淪teve, you鈥檙e not seeing, look, look!鈥�) I grasped at the hope that a premature booster separation was all there was to it, that the orbiter was still going up. But both boosters were beginning to behave erratically, one veering sharply, the other zigzagging wildly. They should have been falling away symmetrically as the orbiter forged on.

Reality began to register. I could not distinguish the familiar three main engines lighting away from me. Beyond the smoke was a ball of fire. It was expanding. Three of us in the media room recognised the awful truth. In unison, we breathed a barely audible 鈥渘o鈥�. As the endless moment continued, confirmation came as bolts of smoke shot out of the lighted mass and the debris began to fall.

All this time (about 15 seconds) mission control said nothing. My brief was to convey factual information: my audience could only listen. If nothing went out over the air for 20 seconds, an alarm would sound in the telephone company鈥檚 headquarters in Kansas City. Surely Nesbitt would say something? I took a deep breath, glanced nervously at Pete Kappesser, and switched on my microphone: 鈥淪omething has gone terribly wrong. We have a disaster.鈥� I pulled the microphone back and watched, disbelieving. The camera was feasting on the vapour plumes 鈥� usually a gorgeous sight. The camera started moving rapidly, tracking individual pieces of debris. Seconds later, Nesbitt鈥檚 voice, subdued, controlled, came over the line from mission control: 鈥淔light controllers here looking very carefully at the situation.鈥� There was a pause 鈥� he had to be sure of his facts 鈥� and then: 鈥淥bviously a major malfunction. We have no downlink.鈥� A longer pause聟 鈥淲e have a report from the flight dynamics officer that the vehicle has exploded.鈥�

The camera continued to rove the horizon. It zoomed in on an extended pall of smoke over the sea 鈥� the point where most debris had landed. Nesbitt came over the line again talking about control teams discussing what rescue vessels could do. We sensed the mood in mission control: the computers had given no instant answers.

The newsroom was stunned. Shortly after 1100, when falling debris was still keeping recovery vessels out of the area, I looked around for the first time. Most people there had been crying. Several of them had applied to go up in the shuttle, as the first representative of the media, later this year. We had all had several opportunities to meet and talk with the crew. I was to assist in following up Christa McAuliffe鈥檚 flight, talking to teachers and schools. In our small way, we felt part of the mission. We had lost friends and colleagues.

The moment of catastrophe had seemed, and for me will always seem, to last for ages. In fact, it was quickly over. NASA personnel started moving partitions and bringing in tables and chairs. People from AT&T, the telephone company, started laying cables. Soon, journalists began arriving in droves, pulled in from energy conferences, political meetings or whatever. Few appeared to know much about the shuttle programme; most had no idea of who and how many had been on board. We were subjected to an endless stream of seemingly trivial questions, only to see our replies taken away and yelled into telephones. Subdued grieving had given way to frenetic newsmaking. And all the while, the mission-elapsed clock on the monitor ticked through the hours, ironically reminding us of what should have been happening.

National broadcasting dwelt on the incident all day long, reshowing the video of the launch and the explosion. Pundits speculated endlessly as to the cause of the disaster. NASA would not speculate. In the Johnson Space Center, NASA鈥檚 television did not replay the launch or the explosion once.

I respect NASA鈥檚 refusal to speculate. Christa McAuliffe鈥檚 message to the schoolchildren of the US was the new frontier is out in space. Hopefully, at the end of the day, when the investigation is complete, we will have some of the answers we crave for now. The space programme will forge ahead once again. That is what the crew would have wished.

First published in New 杏吧原创on 6 February 1986

50 Years of New 杏吧原创: The Best Articles 鈥� find many more in our exclusive Special Report. You can also have your say on what you think was the biggest scientific advance of the last 50 years, in our

The first pictures arrived back from Magellan as it neared the end of
its second test orbit on 16 August. NASA released the pictures last week.
Once the craft begins mapping the surface of Venus in earnest, it will image
swathes of the planet in strips 25 kilometres wide and about 16 000 kilometres
long. The craft will map the Venusian surface at resolutions of between
250 and 600 metres, which will give ten times more detail than previous
probes to visit the planet.

Despite the success so far, the early problems had the mission controllers
biting their nails. They lost contact with Magellan twice during manoeuvres
and test runs around the planet.

The spacecraft first orbited Venus on 10 August after a 15-month cruise
across 150 million kilometres of space 鈥� it was released from the space
shuttle Atlantis on 4 May 1989. The imaging team first lost contact with
the probe towards the end of the second mapping test orbit on 16 August,
after the spacecraft had returned its first pictures.

Mission engineers recontacted Magellan 14 hours later and found that
three distinct failures had occurred. The first of these failures disrupted
equipment that helps the two major on-board computers to keep continuous
check on one another鈥檚 functioning. Mission controllers suspect that a direct
hit on the probe by a cosmic ray particle caused the fault. To compensate
for the failure the main computer triggered the craft鈥檚 backup positional
control and its auxiliary memory.

This auxiliary memory failed as well, contributing to the second failure.
To check its position, the probe was meant to lock on to the star Sirius.
Instead, it locked onto another star, so when the probe looked for Earth
in its expected position relative to Sirius, it was not there.

The spacecraft then automatically commenced a 鈥榮afeing鈥� procedure, turning
its solar panels towards the Sun to avoid depleting battery power and waiting
for Earth to make contact. The spacecraft actually reestablished contact
itself by a process called 鈥榗oning鈥�, in which the probe swivels on its axis,
sweeping the skies with signals in the hope that some would reach Earth.

Controllers finally got a firm lock on the spacecraft on 17 August,
but they lost contact again on 21 August. This time, the signal faded steadily,
suggesting that the spacecraft was slowly turning away from Earth. Subsequent
instructions to resume coning were accepted by the spacecraft. A firm lock
was reestablished on 22 August.

Magellan is currently in orbit around Venus with a radio antenna pointed
to Earth. Two further commands to the spacecraft have disabled a part of
the fault protection system to stop it rotating away from Earth contact
again.