Fix indexing issue in atkinson tables that resulted from adding 1.5 s

[pmpowers-usgs]

The addition of values for 1.5s to Gail Atkinson's ground motion tables caused indexing to be off by one, affecting the scaling of PGA values via a distance decay at BC sites.

See the attached documentation and figure from @YajieLee.

Steve Harmsen's tests do not indicate that this impacts the maps; not exactly sure why it wouldn't. His fix was included along with additional interpolated 2.5 sec implementations for several NGAW2 relations. For whatever reason, both these changes did not make their way into the 'current' hazFX* code.

JLee-amean11-bug-5-7-2015.docx