Re: Baseline Optimization Memo

Hi Again,

The detector size is fixed in my study. But this is not the point. The baseline
optimization (not the absolute sensitivity) can be determined independently of any
assumptions about the detector size, GWatts, or years, by looking at the ratio of
systematic to statistical error at a particular baseline. This fixes the systematic
error and allows the statistical error to vary with 1/r^2 as you shift the baseline.
The optimization as a function of systematic error must be done independantly for the
rate and rate+shape analyses. You can't mix them as you have impled. Clearly the
rate+shape wins as the rate analysis becomes systematics limited (this is the well
known conclusion of the Huber et al.), but it does not have lower systematics. The
difference is that it is not sensitive to the systematics in the same way.

If this note does't resolve the confusion then perhaps you should give me a call.

-Jon

Josh R Klein wrote:

> Right---all I am saying is that minimizing the absolute systematic
> uncertainty is a bigger win than optimizing for the ratio to the statistics.
> As it happens, I think these argue the same way in our case, if you believe
> that the rate+shape analysis has inherently smaller systematic uncertainties,
> since then you want to be closer anyway. The bottom line is that I think we
> should just pick the detector size and fix it, and then re-do your study with
> statistics and systematics treated independently.
Received on Fri May 14 15:12:58 2004