*SNAPIN* reads all of the 10 chord-integrated electron density
measurements provided by the vertically-viewing MIRI diagnostic.
A *proper* normalization of the Thomson scattering profile
measurement of *n*_{e}(*R*) to one of these chords would iteratively

- Calculate the
*n*_{e}(*r*) by smoothing the original Thomson*n*_{e}(*R*) and mapping to minor radius according to the chosen algorithm; - Integrate
*n*_{e}(*r*) vertically along a selected MIRI viewing chord (i.e. a given major radius) to determine the Thomson value of ; - Normalize the
*n*_{e}(*r*) profile to give the value of measured by MIRI at this radius.

However, up through release 6.0 of *SNAP* this algorithm is *not*
used. Instead, irrespective of which MIRI chord is selected as
the one to be used for normalization, it is treated as a *
horizontal* line integral measurement of . The Thomson
scattering measurement is then normalized so that its horizontal
integral of equals the MIRI value. Note that the
Thomson measurement of *n*_{e}(*R*) is *performed* in the horizontal
midplane, so mapping to minor radius is not required to evaluate its
in this algorithm. The practical consequences of
using this more rudimentary algorithm are:

- You must choose the MIRI chord whose sightline most closely passes through the plasma center as the one to normalize the Thomson profile against (or, if available, use the waveform FM-YI-NE which represents the horizontal line-integral electron density inferred by Abel inverting the MIRI data).
- The Abel-inversion algorithm for FM-YI-NE makes an
approximate -dependent correction for the
plasma ellipticity, which is reasonably accurate for
most beam-heated plasmas. Its accuracy is not assured
for the very high regime plasmas produced
by our collaborators from Columbia University.
Using one of the individual MIRI chords for normalization (i.e. not FM-YI-NE) will effectively overestimate the plasma density by its elongation , which for typical plasmas is an error of a few percent.

Fri Jul 11 15:18:44 EDT 1997