The 19 pointing residuals plotted in the report of 20080117 have raw rms scatters in (az,el) of (2.2",2.6"), with a moderately obvious trend of the azimuth residual as a function of (E-2K). Without one outlier in elevation rms scatters are (2.3",1.8"). Those trends that do exist may formally require solutions using trigonometric functions of (2K) or (E-2K), but there may also be simpler (empirical) linear functions that could prove useful. Such linear trends are described here for completeness and are examined for their impact.

The data are those residuals plotted against (E-2K). Straight line fits to all of them, or subsets thereof are:

    daz = -1.29 + (+0.021 +- 0.008)*(E-2K)   s.d. = 1.9   N=19
    del = -0.15 + (-0.013 +- 0.010)*(E-2K)   s.d. = 2.5   N=19

    daz = -1.25 + (+0.026 +- 0.008)*(E-2K)   s.d. = 1.8   N=18
    del = -0.06 + (+0.000 +- 0.008)*(E-2K)   s.d. = 1.8   N=18

The impact is a substantial reduction in the azimuth scatter, with the azimuth fit in the N=18 case exceeding the 3-sigma significance level. Without the outlier the elevation residuals show no systematic variation with (E-2K). The residuals are all plotted below (left) as a function of IMG_ANG (i.e. M, 2K), as shown previously, and (right) after correction by the N=18 solution above:

The residuals show no further trends vis-a-vis IMG_ANG:

    daz = -0.07 + (-0.003 +- 0.008)*IMG_ANG   s.d. = 1.8   N=18
    del = -0.19 + (-0.005 +- 0.008)*IMG_ANG   s.d. = 1.8