SummaryData from 25937 star acquisitions have been analyzed to document any offset between ACA observed magnitude and calculated AGASC catalog magnitude. This is a re-visitation of the analysis done in January 2001. Stars having hard coded B-V colors of 0.7000 were eliminated from the study as were stars with B-V = 1.5000. Only positive OBC identifications were used. There is a noticeable mean offset of -0.1 magnitudes that is a result of under-prediction at B-V colors greater than 1. Because the algorithm amplifies errors as magnitude increases, larger residuals are seen at fainter magnitudes. There is no appreciable offset in magnitude as a function of time. Analysis of the distribution of magnitude differences shows that they are non-Gaussian. The absolute value of magnitude difference is distributed as follows:
Magnitude difference versus color, magnitude, and dateThe plots below show magnitude difference (AGASC magnitude minus observed magnitude) as a function of color (B-V from the AGASC), AGASC magnitude, and date. The key feature here is the trend in magnitude error as a function of color. Remember that the B-V color is truncated at 1.5.
Distribution of magnitude differenceBelow is the distribution of all magnitude differences with absolute separation less than 0.4 (the 99% criteria). The curves are best fit Gaussian for data with magnitude differences (absolute) of 0.1, 0.2, 0.3, and 0.4.
Recalibration of MAG_ACA predictionsThe AGASC 1.5 conversion from V Band magnitude to predicted ACA magnitude is given by the following polynomial (for Tycho 1 & 2 stars):
MAG_ACA = V_MAG + C0 + C1*COLOR + C2*COLOR^2 + C3*COLOR^3
with coefficients as listed in the table below. This calibration was based on 1939 Tycho-1 stars. Using the new, larger dataset we can improve the predictions (especially for colors greater than 1.0) by fitting COLOR2 data from observed Tycho-2 stars. See plot and coefficients below.
DiscussionThe operational impact of using the current AGASC 1.5 color correction is currently unknown. It is clear that we are predicting brighter magnitudes then are seen for the reddest observed stars. A study of the acquisition rate of stars with B-V > 1.0 is probably is probably the best way to determine if AGASC 1.6 is warranted.
Aspect Information main page
Comments or questions: Aspect Help