For what concern the Antonio draft, I have a main comment.
The last figure (minimum energy) gives essentially no information;
(I don't know has it is after the Doug suggestion to plot L/E).
Instead, a plot of the ratio between (IU/throughgoing) vs. cos(zenith)
should be included. It is a plot which was presented by SK, and of
course people can ask us: "way do not use the ratio to reduce
the systematic error?". Moreover, it is very easy for people
outside the collaboration to get this plot using the two
papers of throughgoing and of IU.
This plot was presented by Antonio in the last collaboration meeting,
and I have produced a new version, using the throughgoing data set
for the ICRC99. The figure is in:
WSGS03$DKA0:[MACROSCRA.SPURIO.LOWNU]ratio_iu_th.ps
The theoretical error in the ratio is 18% (in my opinion, too
conservative)
and correspond to the white error box in the figure. The data are the
point
with error. The expectation for max mix and 2.5 10^-3 eV2 is the bold
line.
The chi^2/d.o.f for no-oscillation is 9.5/5;
The chi^2/d.o.f for the oscillation curve is 4.9/5.
Maurizio