The dust flags should be used with caution. Alessandra's studies
of veto efficiency and deadtime suggest that the high-threshold H1
flags should be seriously considered for the burst analysis, but
the H2 flags do not seem nearly as effective. 

There were no dust monitors running in the L1 dark port enclosure
during S4. The flags defined here depend only upon recorded human intrusions.

For reference, dust intervals flagged based on elog entries during
the run of recorded laser safety system intrusions are
flagged by DUST

Dust intervals based on measured dust levels (H1 & H2 only) are
flagged with DUST_ELEVATED and DUST_HIGH for the two different
thresholds Andri applied.

K. Riles
July 14, 2005