We have seen a couple of new RxA pointing models in use over the last couple of months. This report is an analysis of the subsequent data. The first new model was installed on the 5th March, this model was then superceded on 3rd April.

Period: 5th March - 3rd April (UT)

Without any adjustment for nightly collimation differences, they behave as follows:

   Dataset   FE     No.    CA  ± dS       IE  ±  dZ
                             (daz)          (del)
   200903    A     139    +1.8 ±  2.0    +1.1 ± 2.7 

   200903    H     178   +19.5 ±  2.1   +40.9 ± 3.8
 

These are displayed below; RxA on the left and HARP on the right:

RxA is showing a normal distribution whilst HARP continues to display the characteristic systematics in both dS and dZ with elevation. HARP performance is particularly poor in dZ with an rms residual of 3.8".

Period: 4th April - 27th April (UT)

Without any adjustment for nightly collimation differences, they behave as follows:

   Dataset   FE     No.    CA  ± dS       IE  ±  dZ
                             (daz)          (del)
   200903    A     163    +0.3 ±  1.8    +0.4 ± 2.6 

   200903    H     187   +19.7 ±  1.9   +39.4 ± 3.0
 

These are displayed below; RxA on the left and HARP on the right:

Whilst the installation of the new model has only slightly improved the RxA performance, the HARP performance has shown significant improvement with residuals of (dS,dZ) of 1.9, 3.0. Slight systematics remain however which are likely due to as yet uncorrected for k-mirror terms.

Focus

 Month   FE    Axis   N      mean     ±             Notes
 -----   ----  ----  ---    ------- ------  ---------------------
  Mar    RxA     Z    27     0.25    0.15      -3 < Tmean < 10
                 X    16    -0.11    0.13
                 Y    16    -0.01    0.09              
 
  Mar    HARP    Z    55     0.32    0.17      -2 < Tmean < 10
                 X    22    -0.35    0.27     
                 Y    21    -0.17    0.09 

  Apr    RxA     Z    90     0.28    0.17      -1 < Tmean < 11
                 X    28    -0.05    0.30
                 Y    24     0.05    0.17              
 
  Apr    HARP    Z    41     0.28    0.17      -2 < Tmean < 11
                 X    14    -0.36    0.15     
                 Y    15    -0.17    0.08 

 Mar:  HARP_Z  = (+0.58 ± 0.11)*sin(el) s.d. 0.14
 Mar:  RXA_Z   = (+0.40 ± 0.16)*sin(el) s.d. 0.16
 Apr:  HARP_Z  = (+0.39 ± 0.16)*sin(el) s.d. 0.15
 Apr:  RXA_Z   = (+0.40 ± 0.08)*sin(el) s.d. 0.14

Click below for plots of

• 2009 Mar RxA Tmean-vs-Z
• 2009 Mar HARP Tmean-vs-Z
• 2009 Apr RxA Tmean-vs-Z
• 2009 Apr HARP Tmean-vs-Z

Downward trends in the UT-vs-Z plots remain apperent, especially in the RxA focus data. This correlates with the trend seen in the mean temperature which, as is shown below, continues to drop throughout the night.

• 2009 Mar RxA UT-vs-Z
• 2009 Mar HARP UT-vs-Z
• 2009 Apr RxA UT-vs-Z
• 2009 Apr HARP UT-vs-Z

The large dataset from April provides a good sample with which to look at temperature trends. The plot on the left shows mean back leg temperature against time; the temperature starts relatively high at ~8° C then starts to drop once the sun sets at ~04:30 (UT). A downward trend is then apparent with the leg only beginning to stabilize for the last few hours of the night.
The plot on the right gives Tdiff (front leg temperature - back leg temperature) against time. Here again before sunset the front leg temp is higher, its position near the doors meaning it is more exposed to the sun and warmer outside air temperature. However once the sun sets the front leg cools more rapidly and Tdiff goes negative. This stabilises by approximately 9pm (07:00 UT) with the back leg remaining slightly warmer than the the front for the remainder of the night. Airflow with in the building appears insufficient to bring the two legs into thermal equilibrium.