Observability plugin 0.11.4 RC2 Accuracy

[Khalid AlAjaji]

Comparing rise and set times for my location with http://aa.usno.navy.mil/data/ , There is a 4 minutes difference for the Sun and Moon rise\set times. For Mercury and Arcturus it is 2-3 minutes difference.

I run Stellarium with refraction settings active.

Another issue is that when the Sun is selected and time is changed to be near sunset the value reported by the plugin changes as time is changed. One hour before sunset in my case it shows 18:17, 15 minutes before sunset it shows 18:18, at sunset it shows 18:19 and 2 minutes after sunset it shows 18:20.

Stars show the same behavior.

[Ivan Marti-Vidal (i-martividal)]

Dear all,

The difference in rise/set times with respect to those of USNO has an explanation. As far as I know, they take into account refraction effects. Currently, the Observability plugin doesn't consider refraction. The same argument can be used for the Moon.

Regarding the changes in the rise/set time predictions as time goes by, this bug is also definitely related to the refraction effect. If you turn refraction off by removing the atmosphere, the predictions do not depend on time anymore. I thought that the "equinoxEquToAltAz" core function was returning the Alt/Az coordiantes free of refraction effects (which doesn't seem to be the case?). That of course affects the time predictions as a function of the Alt/Az coordinates, since the plugin was assuming that those coordinates were free of refraction effects. This bug can be easily solved by asking the core to give the geometric (i.e., not apparent) Az/Alt coordinates to the Observability plugin. I'll take a look to it asap.

On the other hand, I'm afraid that it will be more complicated to fully implement refraction, since some numerical approach should be used for all the objects (i.e., not only Solar-system objects, but also stars), e.g., by solving for the time of the (refraction-considered) null altitude. I would need to learn a bit about the refraction-related functions in Stellarium to solve this issue.

[Ivan Marti-Vidal (i-martividal)]

The bug of the time-dependent rise/set predictions is solved (I just realized that the equinoxEquToAltAz function can return the refraction-free alt/az just by setting "StelCore::RefractionOff" as the second argument). I uploaded the corrected code as revision 5545 of Observability-1.0.2.

Indeed, knowing this useful detail about this function, the full implementation of refraction effects in Observability should not be that difficult after all :D.

Shall I wait to fully implement the refraction before I propose a new merging?

[Alexander Wolf (alexwolf)]

Oh! I'm merge your fix.

[Alexander Wolf (alexwolf)]

Diff: http://bazaar.launchpad.net/~stellarium/stellarium/trunk/revision/5547