Who should I believe?

[Demonperformer]

The following is just an example of a more general problem; specifics are given in case this helps someone explain the discrepencies.

Starting with this website, I select Leda (Jupiter XIII) and in the Ephemeris Options select a start date of "2018/03/10" Number of outputs "1" Observatory code: "000" [this is Greenwich, which although not spot on to my location is about the closest.]
The result I get is: "2018 03 10 000000 15 22 53.6 -17 09 01   4.976   5.489  0.0699   116.3   9.3  20.5   575-  705+    0.052   066.7    301  +02   -43   0.45   032  -19"
The important bits to note are the date (2018 03 10) the UT (000000) and the topocentric RA & Dec (15 22 53.6 -17 09 01).

I now move to stellarium and enter the same date/time (2018-3-10 0:0:0). My location is Poole rather than Greenwich, but this should not make an enormous difference to the result (it's about 240km). I now do a search for Leda and the data on the screen gives me (amongst other things): "RA/Dec (on date): 15h22m37.06s/-/17°22'47.0"".

Now, this gives me a difference in RA of 16.5 seconds of time, which is 236" (16.5*15*cos(17.15°) [=0.95553]) and a difference in Dec of 826" (13'46"). Using Pythagoras, this makes a total difference of 859" or about 14.3' between these two sets of data allegedly reporting the same point in the sky.

When the FOV of my chip is 23', this is a significant difference. If I centre on the wrong one, Leda would be off the edge of the chip, and picking a point half way between the two figures in the hope that the satellite will appear on one edge of the chip or the other is (IMHO) just sloppy. With the faintness of the object, requiring a lot of imaging-time to capture it, I consider it important that I get the right figure, but I have no idea which set of data is correct or why such a massive discrepency between them.

So, is it me doing something stupid? is it MPC? is it Stellarium? - who should I believe???

Thanks.

[furrysocks2]

For completeness, what does Stellarium give you if the change location to Greenwich?

[Demonperformer]

Just changed to Greenwich and it gives me an identical line and location.

[michael8554]

What Epoch are these ephemeris in?

Michael

[daz]

I don't have Stellarium installed currently, but when I run your date/time settings I get:

2018 03 10 000000 15 22.89   -17 09.0    4.976   5.489  0.0699   116.3   9.3  20.5   575-  705+   +0.048   +0.021   301  +02   -43   0.45   032  -19

Different to yours, by a fraction. I note that the epoch is J2000 (as shown by the line) :

Date       UT      R.A. (J2000) Decl.    Delta     r      r_J     El.    Ph.   V      Offset      Sky Motion         Object    Sun   Moon

So, as Michael suggests, check the epoch in Stellarium is not JNOW

 

Update: If I select 'Full sexagesimal' as the Display, it matches your RA, but I can't see an option to change the epoch still.

[MikeODay]

For what it is worth, SkySafari ( for 2000.0 @ Poole ) gives:

So closer to your first set of figures: (15 22 53.6 -17 09 01) than the ones from your Stellarium output.

If I change to current Epoch, SkySafari gives:

Which does not agree with either of your sets of figures ( 15 22 53.6 -17 09 01 , 15h22m37.06s -17°22'47.0 )

I guess this would tend to indicate that ‘Epoch’ may not be the problem.  Although I guess it still does not help you determine which package is giving you the best estimate for the predicted position.   

........

Edit:

I suppose it also matters which corrections are being applied by the various packages.  For example, in SkySafari one can choose to apply the following corrections:

With all turned on, SkySafari gives the position of Leda as:

and all off:

Given the very significant differences in the predicted positions above, I would think that the selection of the corrections to be applied in the various packages ( and of course the different algorithms, input data, etc. ) will be playing a very large part in any differences between the predictions being reported. 

None of which really helps you, sorry.  All I can suggest is that you try both packages on an easily observable object and see which one is giving you the best prediction.

[Demonperformer]

OK, I had not considered the JNOW/J2000 issue. Even if that is not the whole story, it is certainly having an impact. 

Thinking this through a bit more, I wonder if the purpose of the two is a helpful thing to consider. MPC is a professional tool that is at least in part used by people who want a standardised set of data for theoretical research. Stellarium on the other hand is primarily used by people who want an answer to the question Where do I look to find the object I am interested in. As such, I am coming to the conclusion that the answers given to me by Stellarium are likely to be more useful to me even if not more accurate in an absolute sense.

As to where I go from here, I think Mike's suggestion is probably a good one. Find a test object that is easier and see how actual results compare with the figures given.

Thanks.