Add LTO emissions to grid post-processing

[ian-ross]

This PR adds "surface" emissions to the gridded NetCDF output generated by the trajectory-to-grid post-processing. For the moment, this includes only LTO emissions. For each flight, LTO emissions are assigned to the locations of the origin and destination airports as follows:

• Takeoff and climb emissions are assigned to the origin airport;
• Approach emissions are assigned to the destination airport;
• Idle emissions are split 50/50 between the origin and destination airports.

This approach is based on the behavior of the legacy trajectory builder, which starts it trajectory simulation at the origin airport's location at the altitude of the beginning of the "non-LTO" part of the trajectory. There is thus not a part of the trajectory ground track between the origin airport and the beginning of real trajectory points over which we could distribute takeoff and climb emissions, so the only thing we can really do is to assign those emissions to the origin airport location. The same sort of reasoning goes for approach emissions at the destination airport. And I split idle emissions 50/50 because I couldn't think of anything better to do.

@aditeyashukla Does that seem like a reasonable approach? This will have to be reviewed if the takeoff and landing phases are ever modeled more directly, but that's something for the future.

You can see what the NetCDF data containing the LTO outputs looks like in /home/iross/data/AEIC/oag-2019/test-oag-2019.nc on Hex.