The following images were made from AO125. The images shown below are made with natural weighting. The beam size is 58" by 51" at PA=-55 degrees.
**NEW** Click here for a postscript image showing all continuum sources (labelled) in the VLA primary beam. Click here for positions and flux densities of these sources (primary beam corrected). The fits file is here
DATA SET 2 (AW536), April 2000, 2.68 hours on source, HI spectral line in C array:
The following images were made from AW536. The images shown below are made with natural weighting. The beam size is 17.3" by 16.3" at PA=-15 degrees.
DATA SET 3 (AB292), Sept 1984, 4.15 hours on source, D array: This data set was reduced but the noise (and possibly other errors) was considerably higher (as expected, given the date of observations) than the previous data sets. Combining these data with the previous D array data resulted in a higher, rather than lower noise level, so the data were not used.
DATA SET 4 (AJ125), Aug 1985, 7.40 hours on source, C array: This data set was reduced but the noise (and possibly other errors) was considerably higher (as expected, given the date of observations) than the previous data sets. Combining these data with the previous C array data resulted in a higher, rather than lower noise level, so the data were not used.
DATA SET 5 (AQ11), Sept 1996, 0.45 hours on source, D array & April 1996, 1.30 hours on source, C array: After some reductions were done, I realized that these two data sets were incorrectly set up with the central velocity of the galaxy set to the end of the observing band. Thus, only half of the galaxy is represented in the data set. The data had to be ignored.
OTHER: There were a few other spectral line data sets in the direction of NGC2903, but at the wrong frequency.
COMBINATION OF DATA SET 1 (AO125, D) AND DATA SET 2 (AW536, C): The central positions were slightly offset from each other so UVFIX was used to align them. The velocities were adjusted to the same velocity definition (one was lsr, the other heliocentric). Also the D array data used no on-line Hanning smoothing whereas the C array data did, so the D array data were Hanning smoothed so that the channel width=resolution was virtually identical. The resulting channels were then found to be shifted by 1.98 km/s with respect to each other. These data (with a channel width of about 10 km/s) were then combined as is. So the resulting velocity resolution is really a bit worse than 10 km/s. A self-cal of all channels using the continuum map improved the results. A number of cubes were then made from the combined-array data. I show moment maps of different weightings below. Cubes can be obtained from http://www.astro.queensu.ca/~irwin/pub/ngc2903 for anyone who wishes to look at the data in more detail.
Last revised March 25, 2004