Thermal

  FE  No.of data   fit                              Temp range
  --  ----------  ----------------------------     -----------
   A     133      del = (+1.01 +- 0.73)*T_diff     -0.6 to +1.0
                        (+0.21 +- 0.12)*T_mean       1  to  10

   B     273      del = (+1.85 +- 0.44)*T_diff     -0.9 to +0.9
                        (+0.05 +- 0.06)*T_mean       2  to  13

   W      25      del = (-1.72 +- 0.95)*T_diff     -0.7 to +0.3
                        (-0.37 +_ 0.15)*T_mean       4  to  10

There's nothing significant and consistent across the FEs to justify changing the values of TEMP_SLOPE or MEAN_TEMP_SLOPE. The RxB data show possible systematics in the daz-vs-el plots (upper row, centre panel) - as did last month's results, seen in this light - but, again, this is supported only weakly by the RxA data. The general applicability of the current pointing model (that from 15 December 2004) was discussed recently and it was concluded that a heterodyne-based mode could be created if needed. We continue to wait for SCUBA to return to service in order to more fully quantify, by experiment, these apparent, elevation-dependent pointing systematics.

Focus

      FE  axis  No.obs.    mean   s.d.  Notes ('significant' trends)
                           (mm)   (mm)
    ----- ----  -------  ------- ------ -----------------------------
     RxA    Z      57     -0.25   0.08
            X      37     +0.19   0.24   1
            Y      37     +0.03   0.10

     RxB    Z     111     -0.23   0.09   2
            X      66     +0.18   0.22   3,4
            Y      67     +0.11   0.16

     RxW_C  Z       9     -0.03   0.10 
            X       8     -0.08   0.23
            Y       8     +0.39   0.23 

Notes

1. RxA_X = (-0.72 + 0.15 )*cos(el)
2. RxB_Z = (+0.010 + 0.003)*T_leg     temperature range +2 to 13
3. RxB_X = (-0.57 + 0.09 )*cos(el)
4. RxB_X = (+0.032 + 0.008)*T_leg     temperature range +2 to 13

It is noted that the corrections for elevation trends in Z seen last month seem to work OK, although a zero-point weak may be needed.
Trends 1 and 3 above seem newly significant at the 3-sigma significance level.
However, while the negative progression seen in the RxB data over the last 6 months has indeed reached significance for the first time :

  YYMM   RxB_X =    ( slope +- error ) * cos(el)    No.data
  ----               ------    -----                -------
  0410               -0.17     0.23                   29
  0411               +0.05     0.21                   13
  0412               -0.22     0.25                   30
  0501               -0.28     0.11                   37
  0502               -0.35     0.20                   41
  0503               -0.41     0.14                   48
  0504               -0.57     0.09(**)               66

the RxA data in the same period might have triggered a more sensitive alarm:

  YYMM   RxA_X =    ( slope +- error ) * cos(el)    No.data
  ----               ------    -----                -------
  0410               -0.66     0.16(**)               54
  0411               -0.48     0.11(**)               49
  0412               -0.62     0.21                   39
  0501               +0.07     0.14                    9
  0502               -0.05     0.22                   51
  0503               -0.64     0.23                   23
  0504               -0.72     0.15(**)               37

Only in this last month have both results passed (**) the 3-sigma threshhold together.

SCUBA : There was no indication during January of any trend in the SCUBA X-focus measures :

   SCUBA_X = (-0.02 +- 0.13) * cos(el)

so I shall not change the relevant SMU.ifl parameter, XA, until relative SCUBA-heterodyne pointing and focussing measures can be made (see above), and/or until the new SMU controller is released (soon).