Arecibo L-band Feed ArrayArecibo Galaxy Environment Survey (AGES)


The precursor run is finally complete. Here are some initial results that illustrate the success of the observing technique and also areas where we had problems.

Noise behaviour

Noise vs integration time

The points in the graph above were calculated by taking the reduced datacubes and finding the mean value for a large region of sky that didn't contain sources over the whole channel range. The channel range selected was the inner 3696 channels, thus avoiding the increase in noise at the edge of the bandpass. The noise appears to reduce as t-0.49 +/- 0.06 which is in very good agreement with the t-0.5 dependence as theoretically predicted. Following this predicted decrease in noise with integration time we could have reached a final noise level of
0.95 mJy if all 25 nights' observing had been successful.  

New member of the group?

new detection channel map

One of the mose exciting results is a new HI detection within the group (circled above). There are a number of possibilities for the source of this emission: it is a member of the group previously undetected; it is a tidal plume from the interaction of UGC 1171 & UGC 1176; it is an artefact (the result of emission from UGC 1171 being picked up in the coma lobes). The velocity profiles of UGC 1171 and the new detection have different shapes and peak in different channels, suggesting that this is indeed a new detection and not the result of emission being picked up in the sidelobes. Opitcal observations of this new object are still being completed and more HI analysis is ongoing to get the mass of the object.

LIVEDATA and GRIDZILLA do continuum too!

continuum map
Shown above is a continuum map that was produced from 16 nights observing. Data was reduced using the LIVEDATA and GRIDZILLA packages. The continuum reduction is a relatively new feature of LIVEDATA and GRIDZILLA and the early results are promising. This image was compared with the NVSS image of the same region and the positional values are correct to within 1 arcmin (1 pixel in the final ALFA data).

Wandering Birdie

On some nights we experienced an unusual source of RFI. While the RFI was quite narrow, usually occupying a few channels, it would shift frequencies in a quasi-periodic fashion contaminating up to 60 MHz of the band. If one wasn't bad enough there was a harmonoic that would also rear its ugly head. Below are two graphs showing the noise in each channel of the spectrometer, for two identical regions of sky taken over two consecutive nights. The first night's data contained the birdy, the second night we were birdie-free. The effect on the quality of the final data is quite astonishing.
channel noise 22/11/2004channel noise 23/11/2004

Click here to see a movie of the 'wandering birdie'. It manifests itself as a ghostly image of the ALFA footprint projected on to the sky. Each frame of the movie is an image of the observing region from each individual velocity (or frequency) channel in the datacube. As we progress through the movie frequency is increasing. The footprint appears to move across the image, reflecting the sky's movement over ALFA.

Since the end of the observations we have been informed that the birdie has been traced to some equipment in the focus cabin. Once this RFI has been shielded this should not prove to be a significant problem for subsequent observing runs.