Changes to various SMU parameters and collimation offsets were made on 20071220. The focus offsets logged since then are reported below, although the weather has did not stabilize until UT20080102.

  RxA
  ---
  UT date  HST    AZ  EL      X      Y      Z      Tmean  Tf-Tb
  ------- -----  ---  --      mm     mm     mm          degC
    23    18:19  303  47    -0.05   0.18   0.10     5.2    -0.4
    23    20:45  414  56      ..     ..   -0.07     3.7    -0.9
    23    03:17  301  32      ..     ..   -0.03     1.0    -0.7
    24    18:18  303  47    -0.16   0.08   0.05     4.4    -0.3
    24    21:00  411  60      ..     ..    0.00     2.4    -0.9
    24    01:18  305  55      ..     ..   -0.12     1.0    -0.8
    27    18:39  302  40    -0.13  -0.18   0.16     3.5    -0.5
    27    05:08  254  70      ..     ..   -0.07     1.0    -0.8
    28    21:38  396  69    -0.08  -0.10  -0.13     4.0    -0.5
    29    18:31  302  40    -0.15  -0.10  -0.08     4.7    -0.4
    29    19:23  418  45     0.04  -0.09  -0.03     4.5    -0.4
    29    23:30  327  70      ..     ..   -0.23     3.5    -0.4
    29    03:07  129  80     0.13   0.06  -0.26     3.4    -0.5
    30    17:53  303  47    -0.16  -0.56   0.13     6.4    -0.4
    30    20:37  346  36      ..     ..    0.01     5.0    -0.5
    30    22:44  351  73      ..     ..   -0.25     4.2    -0.6
    30    01:30   97  59      ..     ..   -0.23     3.6    -0.6
    01    21:43  429  62     0.13   0.06  -0.04     4.9    -0.4
    01    02:15  301  37     0.16   0.27  -0.29     3.9    -0.6
    01    06:30  269  46       ..    ..   -0.23     2.9    -0.6
                            -----  -----  -----
              Mean          -0.03  -0.04  -0.08
              s.d.           0.13   0.23   0.14

  HARP
  ----
  UT date  HST    AZ  EL      X      Y      Z      Tmean  Tf-Tb
  ------- -----  ---  --      mm     mm     mm          degC
    22    19:12  302  38    -0.39   0.01   0.36     3.8    -0.6
    22    20:55  413  57      ..     ..    0.28     2.8    -0.8
    22    05:10  249  74      ..     ..    0.34     1.8    -0.5
    23    23:48  250  37    -0.57   0.20   0.31     2.0    -0.8
    24    00:01  322  68     0.08   0.11   0.23     1.4    -0.8
    27    23:21  335  72    -0.03   0.05   0.18     2.2    -0.6    
    27    02:20  422  62      ..     ..    0.18     1.8    -0.7
    02    19:11  302  29    -0.27  -0.16   0.27     5.6    -0.5
    02    21:00  363  47      ..     ..    0.18     4.9    -0.6
    02    02:00  246  65    -0.33   0.13   0.15     2.6    -0.7
    03    18:45  301  33    -0.28  -0.12   0.35     7.3    -0.5
    03    21:05  361  47      ..     ..    0.18     5.8    -0.7
    03    00:49  207  77     0.08  -0.03  -0.01     4.2    -0.6
    04    18:11  302  39    -0.34  -0.16   0.31     6.9    -0.5
    04    22:56  344  43      ..     ..    0.12     4.9    -0.7
    04    00:46  294  69    -0.00   0.09  -0.04     4.3    -0.6 
                            -----  -----  -----
              Mean          -0.20   0.01   0.21
              s.d.           0.22   0.13   0.12

         Mean HARP - RxA    -0.17   0.05   0.29

Both Z-focus offsets are significant: implying the need to change the SMU parameter Z_ZERO by -0.08mm (i.e.from 24.36 to 24.28mm - done at 11:20 HST 04 Jan 2008). The RxA X- & Y- offsets are not significantly non-zero and the relative (HARP-RxA) values for all foci - see

     /jac_sw/itsroot/src/harpb/XML/instrument_harpb.ent

- are essentially - i.e. within (0.04mm, 0.01mm, 0.01mm) - as before, and are not adjusted. (The similarity of the relative (HARP-RxA) measures to those measured previously confirm the use of the entire dataset, despite the poor weather prior to UT20080102).

Pointing
In the week or so ending UT20080104, 69 RxA data have been logged and are shown below:

Mean collimation offsets (CA,IE) are (-1.4", +3.9"), with rms scatters in (dS,dZ)=(1.9",2.7"). There are no obvious systematics in the residuals and, given the poor weather, the pointing performance is perfectly acceptable.

In the same period, 50 HARP data have been logged and are shown below:

Mean collimation offsets (CA,IE) are (+16.1", +35.0"), which are to be compared with the current relative collimation offsets in /jac_sw/itsroot/src/harpb/XML/instrument_harpb.ent of (+13",+31"). RMS scatters in (dS,dZ) are (1.4", 3.3"). There are no obvious systematics in the residuals and the disparate behaviour in azimuth (good) and elevation (poor) is again noteworthy. Based on these and the previous HARP pointing measures the collimation offsets for HARP are adjusted to (+15",+35") (12:00 HST, 04 Jan 2008).

20080108 - a possible thermal explanation for some part of the poor HARP elevation pointing performance
49 HARP pointing data from UT20080105-08 incl. have collimations (CA,IE)=(15.4",36.0") - confirming the above changes - and rms's in (dS,dZ) of (1.7",3.5") - similar to recent data. The data appear free of structural systematics (i.e. those plotted by TPOINT). Temperature data extracted from the JCMT engarchive reveal the following dependencies of these (49) elevation residuals upon leg temperature:

    del = (+1.20 +- 0.33) * Tmean  - 8.3     s.d.  3.1"      2  < Tmean <   8  
        = (-6.7  +- 4.0 ) * Tdiff  - 8.1     s.d.  3.4"    -1.0 < Tdiff < -0.5

[ Tmean = 0.5*(Tfront + Tback); Tdiff = Tfront - Tback ]

The first trend is significant at approximately the 3.6-sigma level, and suggests the need to change the /jac_sw/itsroot/src/tcs/thi/THI.tide parameter TEMP_MEAN_SLOPE from its current value of 0.0"/deg to -1.2"/deg. 82 RxA data from the period UT 20071222-20080101 suggest no such change :

     del = (+0.21 +- 0.17) * Tmean  + 0.6    s.d.  2.3"     0  < Tmean <   7
         = (+4.4  +- 1.7 ) * Tdiff  + 3.7    s.d.  2.2"   -0.9 < Tdiff < -0.3

implying that the effect is specific to the HARP optical train. [ Azimuth errors for both RxA and HARP - already essentially showing 'nominal' performance - show no significant trends against either Tmean or Tdiff. ]

It is just as hard (for me) to conceive of a thermal-dependency of a component of the HARP optical train having an effect in elevation and not in azimuth as it is to conceive of a misalignment of the K-mirror having an effect in elevation and not in azimuth. Nonetheless, the data presented here support such a theory; someone with a more agile concept of the structures involved may be able to explain . . . .