Although not planned at the beginning of the campaign of 2021, the Galilean photometry method proved to provide accurate results most of the time. Here are the full results, including the two others methods I have experimented:
In green are the results found to be accurate given the state of the planet in 2021, where it is probable that the very dark system 1 lowered the value in U,B and V. In orange are the inaccurate or wrong results. The case of the z' value is different: the results are far from the reference given by Anthony Mallama, but have been consistent through three different observations - and the author would like to discuss further the reference of 0,35.
The results found from Alpy 600 spectro-photometry are in complete coherence, and provided the only acceptable value for the U band in 2021.
One session was dedicated to classical star photometry (6th September), but was clearly off.
The
results for 2019 are very interesting. From a good piece of luck, Europa
was found on an UBVRI series taken in June of that year. The results
look excellent, despite the fact that the planet was observed through an
airmass of 2,8 (elevation of 20° only), that failed completely the
attempt of using reference stars. There are good reason for this
unexpected success, see below.
Advantages
of the method
Clearly, the best advantage is that the use of a Galilean moon as a reference solves two tough problems that we face often when trying to use stars as a reference:
1) The moons can be frequently imaged at the same time as Jupiter, and in the same field. This largely eliminates the need to compensate for differences of airmass between the target and the reference. This is particularily striking for the series of 2019 for which the observation was made through nearly 3 airmasses;
2)
The moons, being observed as a resolved disk in contrary to stars that
are confined into the diffraction pattern, provide a much steadier light
flux. One big problem when using stars as reference is that the
twirkling behavior will change the amount of flux on a very short time
scale, which can noticeably reduce the accuracy of the measure.
Moreover, if the Galilean moon is found on the same field that the
globe, this means that we can use the exact same substract of frames, in
the same quality order, making almost sure that the seeing will
influence equally both the planet and the moon. It will also save much
time!
3) Another possible advantage is that the Galilean Moons can be said to have a comparable colour than that of Jupiter. Even if the colours of the four Galileans can differ substantially, as being illuminated by the same source (the Sun), they will have a relatively warm/yellow colour. This can also increase the accuracy since when doing star photometry, the difference of colour index will introduce complexity (like the need of a secondary coefficient of correction mixed from colour and airmass).
Uncertainties
of the method
In
contrary to non variable stars, the Galilean moons of Jupiter are not
completely stable light flux sources. Their light flux will vary first
following the observed orbital phase angle (the longitude of their
central meridian), as well as if we observe either the leading side
(when they are "behind" Jupiter) or their trailing side (when they are
"before" Jupiter). All moons present noticeably brighter magnitudes from
their leading side.
However, this variations can be successfully taken into account since for at least some of the moons, they can be predicted from ephemerides and from the article of Millis and Thompson. Europa and Ganymede look as reliable references (see below).
From that first experience, it looks like the highest uncertainty comes from the values of the V magnitude. It can sometimes vary quite a lot following the source. The values provided by Millis and Thompson can sometimes differ from that given from WinJupos, and both can also differ from the value provided from other ephemeride sources. I have tried two - that can disagree as well between them!!:
The
NASA Horizon system application: this website is a good
reference, especially because the provided values of the V band are
reputed to be corrected from the solar phase angle;
The
IMCCE web service for ephemerides: this one is from the
renowned french institute of ephemerides. It may have provided the most
accurate values, except for one date.
The
fact that the V value can differ substantially for the same observing
time is puzzling and could lead to inaccuracies. However, so far, each
time that I faced different values, only one provided a coherent result
when looking at the measured albedo of Jupiter. All potential different
values lead to obviously wrong albedos. At this time, this how I handled
the situation.
Which
Galilean moon is best ?
Not all Galilean moons are equally interesting for the method. Best is to use a moon that presents a low evolution of its albedo and colour index relatively to its observed longitude. If the surface varies substantially with the longitude, it will introduce errors. Best also is to use a moon that will be frequently enough found near Jupiter, to make it possible to observe both at the same time.
Considering these elements, the Galilean satellites can be classified from the best to the worst as:
1)
The best one may be Europa (JII). It presents a relatively smooth
surface that may not evolve significantly with the orbital phase angle,
especially on its trailing side (the leading side looks more
complicated). The colour index (from Millis and Thompson's article)
looks flat for every longitude for B-V; however it looks to evolve more
for U-B. It will also be found near Jupiter quite often. Results
obtained from Europa were good.
2)
The second best one is Ganymede (JIII). It will not be observed quite
often near the globe, but following Millis and Thompson, its B-V and U-B
colour index look to be very regular through out the observed longitude.
Ganymede also provided accurate results during the campaign of 2021.
3) The possible third best would be Callisto (JIV). Callisto looks to present the most regular, flat colour index through all its longitudes. However, the main problem is that it is often found very far from the planet, and it is very rare to see it in the same camera's field as Jupiter. It is the only one that I did not test so far, for that reason.
4)
Io (JI) appears clearly as the worst reference. It is observed very
often near Jupiter, but it presents the most complicated surface, and
the least smooth spectrum, leading to profund variations of albedo and
colour index following the orbital phase angle. The experience I had
with it lead cleary to a failure.