CHECK OF JOHN'S IP OPTION IN SHARP_COMBINE

Giles Novak, January 16, 2008


(1) Work done in May 2007:

In my DG Tau memo of May 2007, I determined an i.p. correction "by hand", 
i.e. without using sharp_combine.  I obtained this estimate as follows:

First, I assumed the following i.p. parameters:

instrumental: (q_0, u_0) = (0.3, 0.0) %

telescope (q_t, u_t) = (q_t', u_t') = (0.3, 0.0) %

Then I crudely corrected for parallactic angle and elevation, using 
estimates for these quantities taken from the excel logbook.

The result I obtained for the i.p. correction using this "by hand" method 
was:

(delta-q, delta-u) = ( +0.09 , +0.10 ) %

The reason this is so small is that the parallactic angle and elevation 
change a lot during the observations, so the i.p. values, originally of 
order 0.3%, get washed out.


(2) Work done today:

Today I analyzed the same DG Tau data with sharp_combine (version 4.21) on 
zamin, using the same instrumental polarization parameters as given above.

The result is: (q, u) = (-0.51, 1.17) %

With no i.p. correction, I obtain (q, u) = (-0.61, -1.29) %

(This is consistent with what I reported in the DG Tau memo, though in 
that memo I rounded to the nearest tenth of a percent.)

So the i.p. correction as determined by John's code has the following 
result:

(delta-q, delta-u) = ( +0.10 , +0.12 ) %

This is consistent within two hundredths of a percent with the value for 
the i.p correction that I determined "by hand" as given above.

I also verified that the i.p. correction has no effect on the errors in q 
and u.

I think that these tests are reasonably conclusive, so we can use the i.p. 
correction with confidence.  Note that the specific values for i.p. 
parameters to be used (for all 350 micron data except Jan 2006) are posted 
on John's memo of July 3, 2007 (last bullet).