The previous analysis was of data taken during the first half of September. This analysis is of two large-ish (N=22, 20) datasets taken with SCUBA immediately afterwards (UT20040919, 20) and of accummulated data (N=21, 61 ,67) taken since then with SCUBA (in the period UT20040921-23), RxA (UT20040923-30), and RxB (UT20040922-29), respectively.

  FE    No.data   del -vs- temperature                   Temp. range
       (UT200409+)
 -----  -------   -------------------------------------  -----------
 SCUBA     22     del = (-3.4  +- 1.8 ) * (T_f - T_b)    -0.3 to +0.8
                  del = (-1.09 +- 0.17) * (T_f + T_b)/2    6  to  12
 SCUBA     20     del = (-5.8  +- 2.6 ) * (T_f - T_b)    -0.3 to +0.5
                  del = (-1.38 +- 0.17) * (T_f + T_b)/2    5  to  11
 SCUBA     21     del = (-1.4  +- 0.9 ) * (T_f - T_b)    -0.5 to +1.5
                  del = (-0.36 +- 0.20) * (T_f + T_b)/2    3  to  11
  RxA      61     del = (-2.3  +- 1.2 ) * (T_F - T_b)    -0.3 to +0.7
                  del = (-0.77 +- 0.16) * (T_f + T_b)/2    3  to  11
  RxB      67     del = (-0.0  +- 0.7 ) * (T_f - T_b)    -0.7 to +1.3
                  del = (-0.86 +- 0.09) * (T_f + T_b)/2    2  to  10

T_f = mean temperature of 4 probes on the front legs, etc.

1. While previously the value of TEMP_MEAN_SLOPE of -1.0"/deg worked fine for SCUBA, results from UT20040919/20 suggest that an extra -1.0"/deg is needed. This is not borne out by data from the following nights, and makes me somewhat suspicious of a sign error.

2. The previous result also suggested that RxA and RxB needed values for TEMP_MEAN_SLOPE about 0.5 units more negative than did SCUBA, and this is possibly maintained here if the UT19/20 data are ignored and the remaining data are considered contemporaneous.

The previous report concluded with this question:

Do we need a focus-station-dependent version of TEMP_MEAN_SLOPE !?

I feel that the case is not yet made.

However, it does seem necessary to make the value more negative (maybe -1.5"/deg) than it is currently (-1.0"/deg), even though the previous value seemed well determined. This was done at 17:25 HST on 30 Sep 2004.