Draft: Extend coords benchmarks

The following benchmark files need to be covered:

Coordinate benchmarks

Includes calculations from azimuth/zenith to equatorial (J2000 and non-J2000) and back, and also galactic coordinate calculations (back and forth) from equatorial for each detector. The detector location are taken from reference detectors (transformation from UTM is handled by KM3Astro.jl).

• astro/antares_coordinate_systems_benchmark.csv (=> agrees within 0.007 deg)

Notes: rows 2, 9, 17 are off by 0.8 deg when calculating azimuth and zenith from RA and DEC.

• astro/ARCA_coordinate_systems_benchmark.csv (=> agrees within 0.007 deg)

Notes: rows 9, 17 are off by 0.8 deg when calculating azimuth and zenith from RA and DEC.

• astro/ORCA_coordinate_systems_benchmark.csv (=> agrees within 0.007 deg)

Notes: rows 9, 17 are off by 0.8 deg when calculating azimuth and zenith from RA and DEC.

Astro objects benchmarks

Includes calculations from azimuth/zenith to equatorial (only J2000) and back, for each detector. The detector location are taken from reference detectors (transformation from UTM is handled by KM3Astro.jl).

• astro/antares_astro_objects_benchmark.csv (=> agrees within 0.009 deg)
• astro/ARCA_astro_objects_benchmark.csv (=> agrees within 0.007 deg)
• astro/ORCA_astro_objects_benchmark.csv (=> agrees within 0.007 deg)

Moon and Sun benchmarks

Includes calculations from azimuth/zenith to equatorial (J2000 and non-J2000) and back, for each detector. The detector location are taken from reference detectors (transformation from UTM is handled by KM3Astro.jl).

• astro/antares_moon_sun_position_benchmark.csv
  • lon/lat to equatorial (=> agrees within 0.015 deg)
  • lon/lat to azimuth/zenith (=> agrees within 0.015 deg)
  • equatorial to azimuth/zenith
• astro/ARCA_moon_sun_position_benchmark.csv
  • lon/lat to equatorial (=> agrees within 0.015 deg)
  • lon/lat to azimuth/zenith (=> agrees within 0.017 deg)
  • equatorial to azimuth/zenith
• astro/ORCA_moon_sun_position_benchmark.csv
  • lon/lat to equatorial (=> agrees within 0.015 deg)
  • lon/lat to azimuth/zenith (=> agrees within 0.016 deg)
  • equatorial to azimuth/zenith

Yet another coordinate systems benchmark

• astro/local2eq_aanet_2.7.1.csv
  • dx/dy/dz,lon,lat,meridian_conv -> equatorial (ra/dec, ra-j2000, dec-j2000) (<= agrees within 0.00005 deg)
  • dx/dy/dz + KM3Astro-UTM-to-lonlat-conversion -> equatorial (ra/dec, ra-j2000, dec-j2000) (<= agrees within 0.005 deg)
  • equatorial -> dx/dy/dz
  • equatorial -> phi/theta, azimuth/zenith

Note on Moon position calculations

The Moon position is off by up to 0.017 deg compared to aanet, which is probably due to the fact that aanet is using RDPLAN from SLALIB, which ignores the height of the "observatory".

Here is a plot which shows the differences between KM3Astro.jl and AstroPy (24h dataset):