1: \begin{abstract}

2: Orbit determination is possible for a chaotic orbit of a dynamical

3: system, given a finite set of observations, provided the initial

4: conditions are at the central time. In a simple discrete model, the

5: standard map, we tackle the problem of chaotic orbit determination

6: when observations extend beyond the predictability horizon. If the

7: orbit is hyperbolic, a shadowing orbit is computed by the least

8: squares orbit determination. We test both the convergence of the orbit

9: determination iterative procedure and the behaviour of the

10: uncertainties as a function of the maximum number $n$ of map

11: iterations observed.  When the initial conditions belong to a chaotic

12: orbit, the orbit determination is made impossible by numerical

13: instability beyond a computability horizon, which can be approximately

14: predicted by a simple formula. Moreover, the uncertainty of the

15: results is sharply increased if a dynamical parameter is added to the

16: initial conditions as parameter to be estimated. The uncertainty of

17: the dynamical parameter decreases like $n^a$ with $a<0$ but not large

18: (of the order of unity). If only the initial conditions are estimated,

19: their uncertainty decreases exponentially with $n$.  If they belong to

20: a non-chaotic orbit the computational horizon is much larger, if it

21: exists at all, and the decrease of the uncertainty is polynomial in

22: all parameters, like $n^a$ with $a\simeq 1/2$. The Shadowing Lemma

23: does not dictate what the asymptotic behaviour of the uncertainties

24: should be. These phenomena have significant implications, which remain

25: to be studied, in practical problems of orbit determination involving

26: chaos, such as the chaotic rotation state of a celestial body and a

27: chaotic orbit of a planet-crossing asteroid undergoing many close

28: approaches.

29: 

30: 

31: 

32: \keywords{Chaotic Motions \and Numerical Methods}

33: \end{abstract}

34: