Non-UTC timezone handling

[gartavanis]

When the start and end dates cross the time change boundary (e.g. EST->EDT) the code can return one day more forcing data than requested. This should not be an issue, except in the case that the dataset ends and the additional day "falls off the edge".

To fix this, instead of calculating start and end dates separately, the time interval between the original UTC dates should be added to the new timezone start date in order to calculate the new end date. This should be done in all points in the code where end dates are converted from UTC to another timezone.