The engineering prototype Mark IV has been taking "good" data since 010829. The the camera has been pointed at roughly 7 degrees north and has been taking data in a mode where the sky is followed as long as possible. This results in a sequence of 56 ea. 100 second exposures for a run. The camera is then returned to the home position and the sequence is repeated as long as darkness allows. This results in a maximum of 5 fields in the winter, and somewhat more than two in the summer.
This note takes a first look at one of these fields. The field was taken on 011220 and is centered at roughly 7h 18m and 7N.
Processing the 56 V and I frames through the pipeline produced 218895 measurements. The pipeline produces measurements only when it finds the star in the matching V and I frames. Measurements of the same star were grouped together which produced 10621 stars. These stars were measured from once to the maximum possible 56 times.
Sigma was computed for the set of measurements for each star. The first plot is magnitude vs sigma for all of the V data as Figure 1.
Next the flagged data is removed. The pipeline adds flags where stars are saturated or are near a bad pixel or the edge of the frame. This cuts the number of measurements to 202640 and the number of stars stars to 10478. This data is plotted as Figure 2.
It can be seen that this removes many of the stray points near the brightest stars as should be expected since many of them are saturated or nearly so. Next the 99 points added by the pipeline are removed as well as any other magnitudes beyond magnitude 15. This cuts the number of measurements to 202507 and the number of stars to 10422.
There are very few measurements removed by this cut. It mainly serves to remove wild swings from the sigma computed for the measurements of a star. Next it is required that each star be measured at least 10 times. This reduces the number of stars to 5762. I did not compute the effect on the number of measurements, but it is relatively small. The stars measured less than 10 times represent a small fraction of the total measurements. These measurements are plotted as Figure 4.
This really cleans up the measurements "below the line" where single measurements produce a sigma of zero and there is the possibility of a very small sigma. It also removes a lot of stars at the faint end. Again this is to be expected. Probably the cut could as well be placed at 4. We will try smaller values for the next data. It is comforting to see a straight line on the log plot. Next it is wondered what other cuts could improve the data. The center 25% of the sky coverage was extracted and compared to the data for the whole image. There was no obvious difference. This is comforting. Next the middle 30% of the time was extracted. If there is a drift over time then there might be less scatter. Since there was a maximum of 17 points in this data, the number of hits cut was set at 6. Through an oversight, this data does not have the flagged data removed. This is plotted as Figure 5.
While not so obvious, this reduced the scatter somewhat. To bring out the difference the time limited data is plotted with all the data as Figure 6.
It can be seen that the time limited data distribution is somewhat below the total data. It should be noted that this may not be drift with time. The telescope is mounted on a deck with many walls in view. The telescope moves through 35 degrees while taking the run. The first few frames lose some light from the obstruction of a roof line. Throughout the run, there are different reflecting surfaces in view. There are also trees in sight. While the telescope is fitted with long lens shields to attempt to reduce these effects, it is not known how effective these efforts are. For completeness, the I filter equivalent of figure 4 is presented as Figure 7.
A note is not complete without presenting a few interesting stars. Stars with a large sigma were plotted. About 10 out of the 5762 stars that made the final cut appeared to be possible short period variables. This is not a very sophisticated test, so better methods may produce more candidates. Four stars are presented in the order of processing. Figure 8 is a2570. Note that the stars are numbered from 1 to 10621 on the list before the cuts.
The next interesting star is a4881. This is plotted as Figure 9.
Next we plot a4881 with a random "constant" star a378. Star a4881 is not some artifact, since most of the stars are constant as can be seen from this comparison. There is some small trend that can be seen in a378. This has been observed for all the data. There is room for improvement in the analysis. The pair of stars are plotted as Figure 10.
Two more stars complete the first tour of the engineering data. These are plotted as Figure 11, and Figure 12.
The four stars above were entered into Simbad. Only a7147 seemed to have a match. This star matches CCDM JD7621+0747AB. Simbad lists a V magnitude of 9.56. Tass produced a mean measurement of 9.54. The other stars are quite a bit fainter, which may explain why they were not found.