astro: Asteroid 2017 FW17 - an extinct comet?

Asteroid 2017 FW17 was first observed at Pan-STARRS 1, Haleakala on 2017-03-18.

Orbital Elements at Epoch 2457831.5 (2017-Mar-19.0) TDB
Reference: JPL 5 (heliocentric ecliptic J2000)

Element	Value	Uncertainty (1-sigma)	Units
e	.3756229227781228	0.020776
a	4.071577338499798	0.075338	au
q	2.542199558295334	0.13088	au
i	4.36476081532836	0.4957	deg
node	353.1159388040458	0.37693	deg
peri	206.4444762192174	1.1805	deg
M	350.7475904922186	0.39597	deg
t_p	2457908.624933945908 (2017-Jun-04.12493395)	5.191	JED
period	3000.837370760166 8.22	83.288 0.228	d yr
n	.1199665145161817	0.0033297	deg/d
Q	5.600955118704262	0.10364	au

Orbit Determination Parameters

# obs. used (total)	15
data-arc span	8 days
first obs. used	2017-03-18
last obs. used	2017-03-26
planetary ephem.	DE431
SB-pert. ephem.	SB431-N16
condition code	8
fit RMS	.96317
data source	ORB
producer	Otto Matic
solution date	2017-Apr-06 09:58:49

Additional Information

Earth MOID = 1.54121 au

Jupiter MOID = .0287021 au

T_jup = 2.913

JPL covariance

MPC orbit ref: MPEC 2017-FF0

The orbit condition code is 8 so it is extremely uncertain.

The object is interesting because it might have a cometary origin although this is difficult to say given the uncertainty.

I generated 100 clones with orbit parameters average and 1-sigma as shown above and simulated what happened in the last 10^8 days.

Note 1:
thanks to Tony Dunn for a comment that he sent me: if its current nominal orbit is correct, this asteroid is a Hilda with a 3:2 resonance with Jupiter.
But as Tony run it forwards and backwards, it leaves the 3:2 resonance within a few hundred years.
Tony has a online simulator that shows the short term Hilda behaviour .

Note 2:
I used Tony Dunn's simulator to check what happens with asteroid using its nominal parameters. The likely cometary origin seems confirmed

Simulation set-up

Mercury6 Integrator

reference:

J.E.Chambers (1999)

A Hybrid Symplectic Integrator that Permits Close Encounters between Massive Bodies''. Monthly Notices of the Royal Astronomical Society, vol 304, pp793-799.

           Integration parameters
           ----------------------

   Algorithm: Bulirsch-Stoer (conservative systems)

   Integration start epoch:         2457831.5000000 days
   Integration stop epoch:      -100000000.0000000
   Output interval:                     100.000
   Output precision:                 medium

   Initial timestep:                0.050 days
   Accuracy parameter:              1.0000E-12
   Central mass:                    1.0000E+00 solar masses
   J_2:                              0.0000E+00
   J_4:                              0.0000E+00
   J_6:                              0.0000E+00
   Ejection distance:               1.0000E+02 AU
   Radius of central body:          5.0000E-03 AU

Simulation results
The result of the simulation is that 47 out of 100 clones might have arrived in the solar system from a distance greater than 100 AU.
Their arrival time is very much uncertain.
Going back in the past, the first arrival was in year 2075 B.C. while the last arrival was in year 249000 B.C.
This is the distribution of arrival times showing a relative moderate peak about 40000 B.C:

Every clone has its own "orbital history" and different clones have completely different "histories".
While there is no reason to focus the attention on a specific clone, one might choose a clone just to have an idea of the general macroscopic behaviour.
One can also try to capture the effect of the close encounters with the major planets: the details are not necessarily true but the overall pattern might be reasonable.

Thus, let's see the history of a clone that arrived in the solar system about 45000 Years B.C.

The simulator calculated the following number close encounters (other clones had a greater number of encounters, I choose this because it is easier to be managed graphically...):

JUPITER SATURN URANUS
     72      20       1

These encounters occurred at a distance Dmin summarized here:

$JUPITER
   Min. 1st Qu. Median    Mean 3rd Qu.    Max.
0.08149 0.52310 0.73397 0.68313 0.90392 1.03749

$SATURN
   Min. 1st Qu. Median    Mean 3rd Qu.    Max.
0.2115 0.7622 0.8714 0.8798 1.1167 1.2990

$URANUS
   Min. 1st Qu. Median    Mean 3rd Qu.    Max.
0.9797 0.9797 0.9797 0.9797 0.9797 0.9797

This is a plot (made with package ggplot of the programming environment R that was also used to calculate the above tables) showing semi-major axis, perhelium and aphelium distance as a function of time.
The close encounter with Uranus is shown as a vertical dotted line.
It occurred at a distance 0f 0.97 AU and its effect is not so clear:

This second plot below shows the same thing but this time the vertical dotted lines are associated to the Saturn close encounters.

Finally, the same plot this time for Jupiter close encounters:

I made similar plots (only for Jupiter) showing the other orbital parameters:

Eccentricity

Inclination

Ascending Node and argument of perihelion

Close encounters: summary views

As said above, there is no reason to focus on a specific clone besides the possibility to look in a graphical form to one of the possible asteroid "histories".

More in general, we can consider that any given clone had multiple encounters with any given planet, so we can make an average of the these encounters (we can focus first of all on number_of_encounters and encounter distance). Even better: we can make a box-plot (showing min, max, median, 1st and 3rd percentile) to have an idea of the distribution.

In some special cases (like the case of encounters with planet Mercury that you see below), there is only one clone that contributes: thus, the boxplot of the numer of encounters is "flat" and equal to the number of encounters of this clone.

In the plots below, it is also possible to compare the clones suspected to be comets with the normal asteroid clones: look at "cometary-origin".

Neptune