Bug fix for GM-wrapping related to distbear
This is a bug fix. There were still issues with Greenwich Meridian (GM), specifically for grids that spanned across the GM. In this fix, all instances were related to the distbear subroutine, which finds a lat/lon point at a specified distance from an origin lat/lon. That distbear routine always returns the lon value in 0-360 longitude space, which can be inconsistent with the longitude space of the origin point (i.e., that origin point could be in 360+ space). Code was added after all calls to distbear in gettrk -- and there are a lot, probably 12 or so -- to properly adjust the target longitude. Finally, one additional bug fix is that the output_atcfunix subroutine was modified so that a couple of the basin IDs are fixed to match what EMC & NHC have for various basin IDs -- all other locations in the code already had the correct basin IDs to match EMC & NHC.