RVCorrect.py

[Klaus130161]

Dear Thomas,

I found your Python code RVCorrect.py on the Internet and have a question about the accuracy of the result and another one about the sign of the geographical length of the observation site.

First of all, I searched the IRAF. net forum for an example that was calculated with IRAF: http://iraf.net/forum/viewtopic.php?showtopic=138738

The example was:

The object: (year, month, day, UT, RA, DEC) 2007 4 30 08:41:59 15:46:43 37:49:45:45

latitude = 19:49.7
longitude = 155:28.7 !!!! sign wrong ???
altitude = 4160.

The result of IRAF (Frank Valdes):
HJD VOBS VHELIO VLSR VDIURNAL VLUNAR VANNUAL VS. SOLAR
2454220.86577 0.00 0.54 17.96 0.239 -0.002 0.299 17.428

Then I calculated the same example with your code
and I received the following result:

VOBS: 0.000000 km. s-1
HJD: 2454220.865771
VHELIO: 0.566202 km. s-1
VLSR: 18.003718 km. s-1
VDIURNAL: 0.239563 km. s-1
VLUNAR: -0.002047 km. s-1
VANNUAL: 0.328686 km. s-1
VSOLAR: 17,437515 km. s-1

The difference between the two calculations is 0.566-0.54 = 0,026km/sec.

Do you know the reason for this discrepancy?

I have also done a comparison calculation with barycorr
http://astroutils.astronomy.ohio-state.edu/exofast/barycorr.html
Here is the result: 567.467035794 m/sec. The difference in this case is only 1.265 m/sec !!

The problem, however, is that the sign of longitude has to be accepted positively for this result.
Is that correct?

Thank you in advance for your effort
Klaus

[thomasorb]

Dear @Klaus130161,

I have counter checked the results returned by my module with the results returned by the module astropy (http://docs.astropy.org/en/stable/coordinates/velocities.html) with the following code

import astropy.units as u
from astropy.time import Time
from astropy.coordinates import SkyCoord, EarthLocation, FK5
keck = EarthLocation.from_geodetic(lat=19.828333*u.deg, lon=-155.47833333*u.deg, height=4160*u.m) # those are exactly the coordinates of the observatory returned by IRAF (observ.lat, observ.lon)
sc = SkyCoord('15:46:43 +37:49:45', FK5, unit=(u.hourangle, u.deg))
obstime = Time(2454220.865771, format='jd') # exactly the number of JD returned by IRAF (and my module)
barycorr = sc.radial_velocity_correction(obstime=obstime, location=keck)
print barycorr.to(u.km/u.s)
heliocorr = sc.radial_velocity_correction('heliocentric', obstime=obstime, location=keck)
print heliocorr.to(u.km/u.s)

which returns

0.565774937224 km / s # barycentric  (barrycorr.html: 0.567)
0.563097179679 km / s # heliocentric (rvcorrect.py: 0.566 / IRAF: 0.54)

It means that, in terms of barycentric velocity, astropy and the barycorr function (http://astroutils.astronomy.ohio-state.edu/exofast/barycorr.html) return comparable values within the 3 m/s precision error reported in both implementations. And, in terms of heliocentric velocity, both astropy and my implementation are giving similar results.
I thus believe that there might be a precision problem in IRAF's calculation or some other subtlety I am not aware of.

For the sign problem, in IRAF's function, the longitude is positive towards west which is the opposite for other conventions.

You pointed out a good problem which is the use of this module rvcorrect.py in ORCS which might have not been completely checked and compared for a large range of input. In fact I will remove it and consider using the astropy implementation which is certainly far more robust as it is checked by numerous people ;)

Thank you !

@Klaus130161 Please tell me if it answers your questions so that I can close this issue

[thomasorb]

I have changed the heliocentric radial velocity correction method. It has been replaced with the much more reliable astropy.coordinates.SkyCoord.radial_velocity_correction method (f89c9b9). rvcorrect.py module has been deleted (993a0be).

If you want to work with it feel free to make a branch from f89c9b9 (it is under GPL3 licence)

[Klaus130161]

Dear Thomas, thank you for your answer and your effort. You can close the issue. Best regards Klaus —————————————————————————— Dr. rer. nat. Klaus Vollmann Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR) Raumfahrtmanagement / Space Agency Technik für Raumfahrtsysteme und Robotik / General Technologies and Robotics Königswinterer Str. 522-524 53227 Bonn Telefon +49 (0)228 447-212 Telefax +49 (0)228 447-718 Internet www.dlr.de<http://www.dlr.de/> Von: Thomas Martin [mailto:notifications@github.com] Gesendet: Dienstag, 9. Januar 2018 23:31 An: thomasorb/orcs Cc: Vollmann, Klaus; Mention Betreff: Re: [thomasorb/orcs] RVCorrect.py (#18) Dear @Klaus130161<https://github.com/klaus130161>, I have counter checked the results returned by my module with the results returned by the module astropy (http://docs.astropy.org/en/stable/coordinates/velocities.html) with the following code import astropy.units as u from astropy.time import Time from astropy.coordinates import SkyCoord, EarthLocation, FK5 keck = EarthLocation.from_geodetic(lat=19.828333*u.deg, lon=-155.47833333*u.deg, height=4160*u.m) # those are exactly the coordinates of the observatory returned by IRAF (observ.lat, observ.lon) sc = SkyCoord('15:46:43 +37:49:45', FK5, unit=(u.hourangle, u.deg)) obstime = Time(2454220.865771, format='jd') # exactly the number of JD returned by IRAF (and my module) barycorr = sc.radial_velocity_correction(obstime=obstime, location=keck) print barycorr.to(u.km/u.s) heliocorr = sc.radial_velocity_correction('heliocentric', obstime=obstime, location=keck) print heliocorr.to(u.km/u.s) which returns 0.565774937224 km / s # barycentric (barrycorr.html: 0.567) 0.563097179679 km / s # heliocentric (rvcorrect.py: 0.566 / IRAF: 0.54) It means that, in terms of barycentric velocity, astropy and the barycorr function (http://astroutils.astronomy.ohio-state.edu/exofast/barycorr.html) return comparable values within the 3 m/s precision error reported in both implementations. And, in terms of heliocentric velocity, both astropy and my implementation are giving similar results. I thus believe that there might be a precision problem in IRAF's calculation or some other subtlety I am not aware of. For the sign problem, in IRAF's function, the longitude is positive towards west which is the opposite for other conventions. You pointed out a good problem which is the use of this module rvcorrect.py in ORCS which might have not been completely checked and compared for a large range of input. In fact I will remove it and consider using the astropy implementation which is certainly far more robust as it is checked by numerous people ;) Thank you ! @Klaus130161<https://github.com/klaus130161> Please tell me if it answers your questions so that I can close this issue — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<#18 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AhDf6XW-Do6eTZAZ-oM5ps6svX870QeLks5tI-iVgaJpZM4RF8WP>.