Lebesgue integration - Advanced Theorems and Applications

Fundamental Theorems of Lebesgue Integration Introduction: Why These Theorems Matter The fundamental theorems of Lebesgue integration are powerful tools that allow us to interchange limits and integrals—one of the most important operations in analysis. Unlike the Riemann integral, which struggles with this task, Lebesgue integration provides clean, general conditions under which limit and integral operations commute. This ability is essential for modern analysis, particularly in studying Fourier series, differential equations, and probability theory. These theorems work because the Lebesgue integral is built on measure theory, which handles arbitrary measurable sets rather than just intervals. This flexibility is the key to their power. Linearity: Integration is Linear The Statement: If $f$ and $g$ are integrable functions and $a, b$ are real numbers, then: $$\int (af + bg)\,d\mu = a\int f\,d\mu + b\int g\,d\mu$$ What This Means: The integral of a linear combination of functions equals the linear combination of their integrals. This is perhaps the most intuitive property—it says that integration respects the vector space structure of functions. Why It Matters: Linearity is essential because it allows us to break complicated integrals into simpler pieces. Combined with other theorems, it helps us understand the behavior of integrals under transformations. Example: If you know that $\int f\,d\mu = 3$ and $\int g\,d\mu = 5$, then you immediately know that $\int (2f - g)\,d\mu = 2(3) - 5 = 1$ without computing the integral directly. Monotonicity: Order is Preserved The Statement: If $f \le g$ pointwise almost everywhere, then: $$\int f\,d\mu \le \int g\,d\mu$$ What This Means: If one function is always smaller than another (except possibly on a set of measure zero), then its integral is smaller. This is intuitive but powerful—it allows us to compare integrals without computing them explicitly. Why It Matters: Monotonicity provides a way to establish bounds on integrals. If you can find simpler functions above and below your function of interest, you can bound its integral. Example: Consider functions on $[0,1]$ with Lebesgue measure. If $0 \le f(x) \le 1$ for all $x$, then $0 \le \int f\,d\mu \le 1$. Monotone Convergence Theorem: Exchanging Limits Under Order The Statement: Suppose $f1, f2, f3, \ldots$ is a sequence of non-negative measurable functions such that $fk \uparrow f$ (meaning the sequence is increasing: $f1(x) \le f2(x) \le f3(x) \le \cdots$ for all $x$, and converges to $f$). Then: $$\int f\,d\mu = \lim{k \to \infty} \int fk\,d\mu$$ What This Means: When functions increase monotonically, you can swap the order of taking the limit and the integral. This is one of the central results of Lebesgue integration. Why It Matters: The Riemann integral fails dramatically at this task. The Monotone Convergence Theorem is what makes Lebesgue integration so much more powerful for working with limits of sequences of functions. Intuition: Each $fk$ approximates $f$ from below. The integral of $fk$ steadily increases toward $\int f\,d\mu$. Since the functions are ordered and non-negative, there's no "bad behavior" that would prevent us from swapping the limit and integral. Example: Define $fk(x) = \min(x, k)$ on $[0, \infty)$ with Lebesgue measure. This sequence increases to $f(x) = x$. The Monotone Convergence Theorem guarantees that: $$\int0^\infty x\,dx = \lim{k \to \infty} \int0^\infty \min(x,k)\,dx$$ <extrainfo> Approximation by Simple Functions An important application of the Monotone Convergence Theorem involves simple functions. A simple function is one that takes only finitely many values. For any non-negative measurable function $f$, we can construct an increasing sequence of simple functions that converges to $f$. Specifically, for each positive integer $n$, define: $$sn(x) = \frac{k}{2^n} \text{ whenever } \frac{k}{2^n} \le f(x) < \frac{k+1}{2^n}$$ where $k$ ranges over non-negative integers. This construction creates a "staircase" approximation that gets finer as $n$ increases. The sequence $s1, s2, s3, \ldots$ increases monotonically to $f$. The Lebesgue integral is actually defined using such simple functions: the integral of a non-negative measurable function is the limit of integrals of the approximating simple functions. The Monotone Convergence Theorem guarantees this limit exists and is well-defined. </extrainfo> Fatou's Lemma: Lower Bounds for Limits The Statement: For any sequence of non-negative measurable functions $f1, f2, f3, \ldots$: $$\int \liminf{k \to \infty} fk\,d\mu \le \liminf{k \to \infty} \int fk\,d\mu$$ What This Means: The integral of the "limit inferior" (roughly speaking, the lower limit) of a sequence is at most the limit inferior of the integrals. Why It's Tricky: This is not an equality like the Monotone Convergence Theorem—it's an inequality. This is because Fatou's Lemma allows the functions to misbehave, as long as they don't grow too wildly. Unlike the Monotone Convergence Theorem, we don't require the sequence to be monotone, so we lose equality. Intuition: Consider what can go wrong. A sequence of functions could have "mass" that escapes to infinity or spreads out in complicated ways. Fatou's Lemma tells us that the integral of the limiting function is at least as small as the limit of integrals—the integral can only increase (or stay the same) due to these bad behaviors. When to Use It: Fatou's Lemma is useful when you know the integrals $\int fk\,d\mu$ but the functions $fk$ lack the monotonicity required for the Monotone Convergence Theorem. Dominated Convergence Theorem: The Ultimate Tool The Statement: Suppose: $f1, f2, f3, \ldots$ are measurable functions converging pointwise almost everywhere to a function $f$ There exists an integrable function $g$ such that $|fk(x)| \le g(x)$ for all $k$ and almost all $x$ Then: $$\int f\,d\mu = \lim{k \to \infty} \int fk\,d\mu$$ What This Means: If a sequence of functions converges pointwise and is "controlled" by an integrable function (the "dominating" function), then we can swap the limit and integral. Why It Matters: This is the most commonly used of the three main convergence theorems. The key insight is that if all functions in the sequence are bounded by a single integrable function, their limit behaves nicely enough that we can pass the limit inside the integral. Why the Dominating Function is Essential: Without the dominating function, swapping limits and integrals can fail. Consider functions that become more and more concentrated at a single point. The integrals might converge to one value, but the limit function might have zero integral. Intuition: The dominating function $g$ prevents the functions $fk$ from "running off to infinity" or spreading out uncontrollably. With this control, the integrals converge to the integral of the limit. Example: Consider on $\mathbb{R}$ with Lebesgue measure: $$fk(x) = \frac{1}{k} \sin(kx)$$ This sequence converges pointwise to $0$. Since $|fk(x)| \le \frac{1}{k}$ for all $x$, and $\frac{1}{k}$ is integrable (on any finite interval), the Dominated Convergence Theorem applies, guaranteeing: $$\int{\mathbb{R}} 0\,dx = \lim{k \to \infty} \int{\mathbb{R}} \frac{1}{k}\sin(kx)\,dx = 0$$ Why Riemann Integration Falls Short Understanding the limitations of the Riemann integral motivates why we need Lebesgue integration and these powerful theorems. The Dirichlet Function: A Classic Failure The Dirichlet function is defined as: $$D(x) = \begin{cases} 1 & \text{if } x \in \mathbb{Q} \\ 0 & \text{if } x \notin \mathbb{Q} \end{cases}$$ This function is not Riemann integrable. Here's why: in any subinterval, there are both rational and irrational numbers. When computing Riemann sums, different choices of sample points yield different values—some Riemann sums approach $0$, others approach $1$. The integral doesn't converge, so the Riemann integral doesn't exist. However, with Lebesgue integration, the story is simple: the rationals form a null set (measure zero). Since the function equals $1$ only on a null set, the Lebesgue integral is: $$\int{\mathbb{R}} D(x)\,d\mu = 0$$ This shows how Lebesgue integration naturally handles highly discontinuous functions. Failure of Monotone Convergence for Riemann Integral Consider the sequence: $$fk(x) = \begin{cases} 1 & \text{if } x \in [0, 1] \cap \mathbb{Q}k \\ 0 & \text{otherwise} \end{cases}$$ where $\mathbb{Q}k$ is an enumeration of the rationals. This sequence is monotone increasing and each $fk$ is Riemann integrable on any interval (with integral $0$). However, the pointwise limit is the Dirichlet function, which is not Riemann integrable. This dramatic failure illustrates a fundamental weakness of the Riemann integral: monotone convergence of integrable functions can produce non-integrable limits. Limited Domain Structure The Riemann integral requires integration over intervals (or finite unions of intervals). It does not extend naturally to: Highly irregular measurable sets (like Cantor-type sets) Unbounded domains (requires improper integrals, which are less robust) Abstract measure spaces The Lebesgue integral, built on measure theory, works seamlessly over any measurable set. Difficulty Interchanging Limits and Integrals Many important problems in analysis require computing: $$\lim{k \to \infty} \int fk\,d\mu \quad \text{vs.} \quad \int \lim{k \to \infty} fk\,d\mu$$ For Riemann integrals, very few general conditions guarantee these are equal. This severely limits applications in: Fourier series (convergence of partial sums) Fourier transforms Differential equations Probability theory The fundamental theorems of Lebesgue integration (Monotone Convergence, Fatou's Lemma, Dominated Convergence) provide exactly the tools needed for these applications. Summary: The Power of Lebesgue Integration The fundamental theorems of Lebesgue integration emerge from its foundation in measure theory: Linearity is the expected vector space property. Monotonicity provides basic comparison tools. The three main convergence theorems—Monotone Convergence, Fatou's Lemma, and Dominated Convergence—give us control over when we can exchange limits and integrals. Together, these theorems solve the central problem of the Riemann integral: they provide robust, general conditions for working with sequences of functions. The Dominated Convergence Theorem, in particular, is remarkably useful because the condition of domination is easy to verify in practice. The Lebesgue integral truly subsumes the Riemann integral. It handles discontinuous functions, irregular domains, and limits of sequences with grace. This is why it is the standard tool in modern analysis, probability theory, and partial differential equations.