 |
||||||||||
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
|
 |
|
|
|
|
|
|
|
|
|
|
SETI
- Allen Telescope Array
- Optical SETI
- Interstellar Message Construction
- Project Phoenix 1995-2004
- Background Science
- History of SETI
- Center for SETI Research Team
- Other people doing SETI
- Hollywood's View of SETI
Past Observing Campaigns:
Arecibo, Puerto Rico
Diary November 1999
Nov. 28, 1999
The first week of observing has passed in a flash, a flash of radio frequency interference (RFI) that is. We've been experimenting with new techniques to try to observe more of the spectrum without being overwhelmed by signals from terrestrial technology. We've had some success but not as much as we had hoped. In some parts of the spectrum it will simply be impossible to search with our current signal processing techniques. We're learning lessons for the New Search System and 1 Hectare Telescope projects.
Actually, it's not fair to call these signals interference since the majority are broadcasting within their assigned bands. Most radio astronomy experiments operate within "protected" frequency bands such as the frequency range around 1420 MHz, the frequency emitted by hydrogen atoms. Since ET shouldn't be expected to know about Earth's radio regulations, we search as many frequencies as possible.
Sometimes we encounter strong radar signals that have interesting modulation patterns. The screen image below was produced by a strong radar. Our test signal is barely visible as a vertical line within the complex structure of the radar signal. Jill Tarter named this the "tire tracks" pattern. This is an apt name since it looks like the radar has "run over" our data.
When viewed on the display of the power across 10 MHz, the power and width of the radar signal is clearly visible in the image below. When this radar is transmitting, it will drown out any other signal from 1237 to 1243 MHz.
Dec. 9
It's Thursday, and it's a day off for Project Phoenix.
Tonight, we won't get our usual 12-hour stint on the telescope, and in principle the astronomers and engineers could grab a car, head for Condado Beach, and squeeze sand through their toes.
But they won't and they aren't. Peter Backus is in the control room, looking over the data logs from last night's run. Kent Cullers is down in the Visiting Scientists' Quarters, working on a paper. The engineers, John Ross, Kevin Day, and Tom Kilsdonk, are fiddling with the Phoenix hardware as well as trying to establish a video link back to Mountain View. I'm writing these lines
The dedication to the job at hand seems to have had a salubrious effect on our observing. Compared to the March/April run, the observations of the past few weeks have been as smooth as a lounge lizard's pickup line.
For the past four days, we've gotten control of the Arecibo telescope an hour or so early. Startup procedures, which in March were plagued by software hangups, repeated microcode downloads, and difficulties in communicating with the telescope itself, now proceed without a hitch. Some of this improvement is due to the efforts of Phil Perillat, a computer programmer at the Observatory who has integrated most of the startup procedures on the Arecibo side of the experiment (i.e., on all of the Phoenix observing chain, including the telescope drive, that involves standard Observatory hardware). In addition, we've learned from the difficulties of the March run, and have allowed additional time to get the system up to speed before observing begins.
And frankly, there's another point. The amount of terrestrial interference (or RFI, as the radio engineers like to call it) here is thicker than cold, pea soup. In previous runs, the Phoenix system has waded relentlessly through this forest of static, a practice that's frustratingly slow. Since March, modifications to the software have allowed particular bad bands on the dial to be circumvented. The result is that between 1,200 and 1750 MHz (so-called L Band), Phoenix now skips about 30% of the frequencies. This is a regrettable price to pay for the growing onus of interference, but it allows the observing in the relatively clear bands to proceed far faster.
Tomorrow I'll be giving a talk on SETI to the local staff, and as payment for my efforts will be treated to an exotic lunch at El Observatorio - a new restaurant located about a mile down the road, specializing in crushed and fried plantains. After that, it's back to the control room, and another night of searching the night sky.
Dec. 13
On the last day of observations, we bring you a preview of the Project Phoenix at Arecibo webcam, dubbed the SETICam. See what the scientist and engineers are doing in real time. We hope to bring you continuous live coverage at the next observations in February 2000.
Project Phoenix's Fall run has been like a boat ride down a placid river. With hardly a ripple in the system, we've watched the celestial scenery pass by slowly, all the while scouring the river banks for tell-tale signs of intelligent life.
Surprises can lurk 'round the next bend, of course. On Saturday night, we slammed into a small cascade.
It was nearly 6:00 pm, and Tom Kilsdonk and I were setting up the system for observations. Among other things, this involves configuring the software for two-site (Arecibo/Jodrell Bank) mode, restarting the FUDDs, and getting Jodrell Bank on the horn to tell them we're going to take control of their lovely telescope. All of that was done by 5:55. The next step was to start up some Arecibo software that allows the Phoenix computers to steer both the telescope and the various, local receiver systems.
The Arecibo software came to life, but for some reason the Phoenix computers couldn't get in touch. A small communications problem, in other words. We've come all this way to search for a communication link that might be 100 light-years long or more, but got hung up on getting bits across the room.
Tom and I tried all the usual fixes, including the mildly Draconian one of restarting the Phoenix system. Nothing helped. Phoenix wasn't talking to Arecibo. Feeling the situation was now in extremis, we played our ultimate card: we woke up Peter Backus.
Peter ambled into the control room shortly, and looked over the situation. The problem, both he and Tom agreed, was that a backup tape drive had been replaced that afternoon, and this entailed re-booting Mycroft, the principal Phoenix control computer. In so doing, we presumably had managed to rupture connection to the Arecibo LAN.
In the end, this was a job for Superwoman. We paged Lee Henrikson, somewhere in California. Lee is the Institute's network guru. She and Peter were on the phone for about 20 minutes, and in a triumph of remote diagnosis, Lee was able to find a small error in Mycroft's startup script.
The error was fixed, and the system booted up like a champ. We were back on the air.
It was a relatively small drama in a relatively trouble-free journey. By 8:00 pm, we were once again stepping up the band every five minutes, checking out nearby star systems for E.T.'s presence. We were covering the spectrum at the rate of hundreds of millions of channels per hour. We continued to stare intently at the river bank, hoping to find the object of our search.