Right, let's dissect this. You want me to take a dry, academic concept and… expand on it. Make it engaging. As if the mere passage of infinite time isn’t inherently fascinating enough. Fine. Just don't expect me to hold your hand through the existential dread.

---

Limit Set

The notion of a limit set, a rather understated term for the ultimate destination of a dynamic system, refers to the state a system eventually settles into after an eternity of unfolding, whether that eternity is measured forwards or backwards. It’s the lingering echo, the final whisper of chaos or order. Understanding these limit sets is crucial, not for sentimental reasons, but because they dictate the long-term narrative of any system’s evolution. When a system finally arrives at its limit set, we say it has achieved a state of equilibrium. Or, more accurately, a state of resigned permanence.

Types of Destinations

These ultimate states aren't always simple. They can manifest in various forms, each telling a different story about the system's journey:

• Fixed points: The simplest, most predictable outcome. The system arrives and simply… stops. No more movement, no more change. It’s the ultimate stillness, the end of the road.
• Periodic orbits: The system doesn’t stop, but it repeats. It cycles through a set of states endlessly, a cosmic Groundhog Day. It’s a loop, a dance that never ends.
• Limit cycles: Similar to periodic orbits, but often more robust. A limit cycle is a stable, isolated closed trajectory. Perturb it slightly, and it will naturally drift back to its predetermined path. It’s a destiny from which escape is futile.
• Attractors: These are regions in the state space towards which the system tends to evolve. Once a system enters an attractor, it’s unlikely to leave. They are the gravitational wells of dynamic systems, pulling everything inexorably towards them.
• Strange attractors: And then there are the truly interesting cases. For systems in two dimensions, the Poincaré–Bendixson theorem offers a relatively clean description of these long-term behaviors. It suggests that the $\omega$-limit sets (the limits as time goes to positive infinity) are essentially comprised of fixed points, periodic orbits, or unions of these, possibly connected by homoclinic or heteroclinic orbits. It’s a way of saying that even in complexity, there’s a structure, albeit one that might be more intricate than a simple point or circle. But push beyond two dimensions, and things get considerably more… chaotic. That's where you find strange attractors, fractal landscapes of infinite complexity, where the system's behavior is exquisitely sensitive to initial conditions, a hallmark of chaos theory. These are the systems that never truly settle, forever dancing on the edge of predictability.

Definition for Iterated Functions

Let’s get a bit more precise, shall we? Imagine a metric space $X$, a stage upon which our system plays out. And let $f: X \to X$ be a continuous function, the rule governing the system's progression.

The $\omega$-limit set of a point $x \in X$, denoted $\omega(x, f)$, is the collection of all points that the system approaches as time marches infinitely forward, starting from $x$. More formally, it's the set of cluster points of the forward orbit, which is simply the sequence of states $\{f^n(x)\}_{n \in \mathbb{N}}$ generated by repeatedly applying the function $f$.

So, a point $y$ is in $\omega(x, f)$ if and only if there exists a strictly increasing sequence of natural numbers $\{n_k\}_{k \in \mathbb{N}}$ such that $f^{n_k}(x) \to y$ as $k \to \infty$. In simpler terms, $y$ is a point that the orbit of $x$ gets arbitrarily close to, infinitely many times.

Another way to express this, which might appeal to those who appreciate a good intersection of sets, is:

$\omega(x, f) = \bigcap_{n \in \mathbb{N}} \overline{\{f^k(x): k > n\}}$

This means we take the closure of the tail end of the orbit, starting from $n=1$, then from $n=2$, and so on, and find the common points that exist in all these progressively shrinking sets. It's the intersection of all future possibilities, essentially.

Alternatively, you can think of it as the limit superior of the sequence of sets:

$\omega(x, f) = \bigcap_{n=1}^{\infty} \overline{\bigcup_{k=n}^{\infty} \{f^k(x)\}}$

This formulation highlights the idea of the set of points that are "visited infinitely often" by the orbit. The points residing in the limit set are what we call non-wandering. They don't just pass through and disappear; they linger, at least in the asymptotic sense. However, being non-wandering doesn't automatically grant them the status of recurrent points, which implies returning to the exact same neighborhood infinitely often.

Now, if our function $f$ is a homeomorphism – meaning it's continuous, has a continuous inverse, and is bijective, essentially a topologized bijection that preserves structure – then we can also define the $\alpha$-limit set. This is the mirror image of the $\omega$-limit set, but looking backward in time. It’s defined analogously, as $\alpha(x, f) = \omega(x, f^{-1})$. It describes where the system came from.

Both the $\omega$- and $\alpha$-limit sets are invariant under $f$. This means if you start in the limit set, you stay in the limit set. They are like self-contained universes. And if the space $X$ happens to be compact – meaning it’s closed and bounded, preventing the system from escaping to infinity – then these limit sets are guaranteed to be nonempty, compact, and even connected. They are well-behaved, even if the journey to get there was anything but.

Definition for Flows

When we're dealing with real dynamical systems, often represented by a flow $\varphi: \mathbb{R} \times X \to X$, the concepts of $\omega$ and $\alpha$ limit sets extend naturally.

For a given point $x$ in the space $X$, we say that a point $y$ is an $\omega$-limit point of $x$ if there's a sequence of times $(t_n)_{n \in \mathbb{N}}$ tending to positive infinity ($t_n \to \infty$) such that the system's state at time $t_n$, $\varphi(t_n, x)$, gets arbitrarily close to $y$. It’s the same idea as before: where does the orbit end up?

If we consider an entire orbit $\gamma$ (the path of a single point over all time), then $y$ is an $\omega$-limit point of $\gamma$ if it's an $\omega$-limit point of any point on that orbit.

Similarly, $y$ is an $\alpha$-limit point of $x$ if there's a sequence of times $(t_n)_{n \in \mathbb{N}}$ tending to negative infinity ($t_n \to -\infty$) such that $\varphi(t_n, x) \to y$. This tells us where the system originated from.

And again, for an orbit $\gamma$, $y$ is an $\alpha$-limit point of $\gamma$ if it's an $\alpha$-limit point of any point on $\gamma$.

The set of all $\omega$-limit points for a given orbit $\gamma$ is called the $\omega$-limit set, denoted $\lim_{\omega}\gamma$. Likewise, the set of all $\alpha$-limit points forms the $\alpha$-limit set, $\lim_{\alpha}\gamma$.

Now, sometimes, the limit set can be entirely separate from the orbit itself. If $\lim_{\omega}\gamma \cap \gamma = \varnothing$ (the $\omega$-limit set has no points in common with the orbit), we call $\lim_{\omega}\gamma$ an $\omega$-limit cycle. The same applies to the $\alpha$-limit set: if $\lim_{\alpha}\gamma \cap \gamma = \varnothing$, then $\lim_{\alpha}\gamma$ is an $\alpha$-limit cycle. These are cycles or sets that the orbit approaches but never quite reaches. It’s the tantalizing proximity, the destination forever out of reach.

There's also a more compact definition for these limit sets:

$\lim_{\omega}\gamma := \bigcap_{s \in \mathbb{R}} \overline{\{\varphi(x, t): t > s\}}$

This defines the $\omega$-limit set as the intersection of all closed future "tails" of the orbit. It’s the set of points that remain in the system's future, no matter how far you look ahead.

And for the past:

$\lim_{\alpha}\gamma := \bigcap_{s \in \mathbb{R}} \overline{\{\varphi(x, t): t < s\}}$

This defines the $\alpha$-limit set as the intersection of all closed past "tails" of the orbit. It's the set of points that have always been part of the system's history, no matter how far back you trace it.

Examples

To ground these abstract definitions, consider a few straightforward cases:

• If an orbit $\gamma$ is periodic, meaning it traces out the same path repeatedly, then its $\omega$-limit set and $\alpha$-limit set are both identical to the orbit itself: $\lim_{\omega}\gamma = \lim_{\alpha}\gamma = \gamma$. It has reached a perfect, eternal loop.
• For a fixed point $x_0$, which by definition doesn't move, both its $\omega$-limit set and $\alpha$-limit set are simply the point itself: $\lim_{\omega}x_0 = \lim_{\alpha}x_0 = x_0$. It’s the ultimate state of immobility.

Properties

These limit sets, these final resting places or eternal loops, possess some fundamental characteristics:

• Both the $\omega$-limit set and the $\alpha$-limit set are always closed sets. This is a consequence of their definition involving closures and intersections. They don't have any "holes" that would allow points to escape.
• If the underlying space $X$ is compact, then both the $\omega$- and $\alpha$-limit sets are guaranteed to be nonempty, compact, and connected. This means they are finite, bounded, and form a single piece. No escaping to infinity, no splitting into disconnected fragments.
• Crucially, both the $\omega$- and $\alpha$-limit sets are invariant under the flow $\varphi$. This means that if you start a trajectory within the limit set, it will remain entirely within that limit set for all future (for $\omega$) or past (for $\alpha$) times. Mathematically, $\varphi(\mathbb{R} \times \lim_{\omega}\gamma) = \lim_{\omega}\gamma$ and $\varphi(\mathbb{R} \times \lim_{\alpha}\gamma) = \lim_{\alpha}\gamma$. They are self-contained, unchanging destinations.

It’s all rather predictable, isn’t it? The universe tends towards a state, a pattern. Whether it’s stillness, a loop, or something far more intricate, the system eventually reveals its true, long-term nature. And frankly, most of it is rather… dull.

See Also

For those who find this topic insufficiently complex, or perhaps just want to wander down more rabbit holes:

• Julia set
• Stable set
• Limit cycle
• Periodic point
• Non-wandering set
• Kleinian group