Zeros and Poles

In complex analysis, understanding the behavior of a complex function near specific points is crucial. Two of the most important types of points to consider are zeros and poles. These points tell us where a function evaluates to zero or becomes infinitely large, respectively.

Zeros of a Complex Function

A complex number z₀ is called a zero of a complex function f(z) if f(z₀) = 0. The order of a zero is an important characteristic. If f(z₀) = 0, f'(z₀) = 0, ..., f^(n-1)(z₀) = 0, but f⁽ⁿ⁾(z₀) ≠ 0, then z₀ is a zero of order n. This means that the function can be expressed in the form f(z) = (z - z₀)ⁿ g(z), where g(z) is an analytic function and g(z₀) ≠ 0.

For example, consider the function f(z) = z³ - z. We can factor this as f(z) = z(z² - 1) = z(z - 1)(z + 1). The zeros are at z = 0, z = 1, and z = -1. Each of these is a zero of order 1 because the function does not have repeated roots at these points.

Now consider f(z) = (z - 2)² sin(z). Here, z = 2 is a zero. Let's check the derivatives:

• f(2) = (2 - 2)² sin(2) = 0
• f'(z) = 2(z - 2)sin(z) + (z - 2)²cos(z), so f'(2) = 0
• f''(z) = 2sin(z) + 2(z - 2)cos(z) + 2(z - 2)cos(z) - (z - 2)²sin(z). At z = 2, f''(2) = 2sin(2) ≠ 0.

Therefore, z = 2 is a zero of order 2 for this function.

Poles of a Complex Function

A complex number z₀ is called a pole of a complex function f(z) if f(z) can be written in the form f(z) = g(z) / (z - z₀)^m, where g(z) is analytic and non-zero at z₀, and m is a positive integer. The integer m is called the order of the pole. In simpler terms, as z approaches z₀, the magnitude of f(z) tends to infinity.

A pole is a type of isolated singularity. If a function has a pole at z₀, then lim_z→z₀ |f(z)| = ∞.

Consider the function f(z) = 1 / (z - 3). Here, z = 3 is a pole of order 1. As z approaches 3, the denominator approaches 0, and the function value goes to infinity.

For the function f(z) = 1 / (z - 1)², z = 1 is a pole of order 2. The denominator becomes zero at a rate proportional to (z - 1)².

If a function f(z) has a pole at z₀, then the function 1/f(z) has a zero at z₀. For instance, if f(z) = 1 / (z - 5)³, then 1/f(z) = (z - 5)³, which has a zero of order 3 at z = 5. This implies that f(z) has a pole of order 3 at z = 5.

The order of a pole can be determined by checking the derivatives of the reciprocal of the function. If f(z₀) is infinite, consider h(z) = 1/f(z). If h(z₀) = 0, h'(z₀) = 0, ..., h^(m-1)(z₀) = 0, but h^(m)(z₀) ≠ 0, then f(z) has a pole of order m at z₀.

Local Mapping Properties

The concept of zeros and poles is intrinsically linked to how a complex function maps points in its domain to its range. The local mapping properties of an analytic function describe its behavior in the neighborhood of a point, particularly concerning its geometric effect on small regions.

Mapping near a Zero

If an analytic function f(z) has a zero of order n at z₀, then in a small neighborhood around z₀, f(z) behaves like c(z - z₀)ⁿ, where c = f⁽ⁿ⁾(z₀) / n! is a non-zero complex constant.

Consider a small disk centered at z₀. The mapping property near a zero of order n is that it magnifies small distances by a factor of |c| and rotates angles by arg(c). Crucially, it also n-to-1 maps small neighborhoods. This means that if we take a small disk around z₀, any value w in the corresponding mapped disk (excluding the image of z₀, which is 0) is attained n times by f(z) for z within the small disk around z₀.

For example, f(z) = z² has a zero of order 2 at z = 0. In the neighborhood of 0, f(z) maps a small disk around the origin such that any value in the resulting mapped region (a small disk around 0 in the w-plane) is attained twice. This is why the mapping of the unit disk by f(z) = z² covers the unit disk twice.

Mapping near a Pole

If an analytic function f(z) has a pole of order m at z₀, then in a small neighborhood around z₀, f(z) behaves like c / (z - z₀)^m, where c is a non-zero constant.

Near a pole, the function magnifies distances significantly and rotates angles. The mapping is m-to-1 in the sense that for a small punctured disk around z₀, the image under f(z) covers a large region (e.g., the entire complex plane excluding a small disk around the origin) m times.

Consider f(z) = 1/z, which has a pole of order 1 at z = 0. A small disk around the origin in the z-plane is mapped to a large region in the w-plane. For instance, the annulus ε < |z| < 1 is mapped to the annulus 1 < |w| < 1/ε. As ε → 0, the mapped region becomes unbounded.

Mapping near an Ordinary Point

If z₀ is an ordinary point of an analytic function f(z), meaning f'(z₀) ≠ 0, then the function acts as a local conformal mapping. Conformal mapping preserves angles between intersecting curves. In such a neighborhood, the mapping is one-to-one locally, meaning small disks are mapped to small disks with similar orientation (up to a rotation and scaling). The magnification factor is |f'(z₀)|.

Maximum Modulus Principle

The Maximum Modulus Principle is a fundamental theorem in complex analysis that deals with the maximum value of the modulus (absolute value) of an analytic function within a given region. It has significant implications for understanding the behavior of analytic functions.

Statement of the Principle

Let f(z) be an analytic function on a bounded, connected domain D. If f(z) is also continuous on the closure of D (denoted D̄), then the maximum value of |f(z)| on D̄ is attained on the boundary of D (denoted ∂D).

In other words, if M = max_z∈D̄ |f(z)|, then there exists some z₁ ∈ ∂D such that |f(z₁)| = M. For any z ∈ D, we have |f(z)| ≤ M.

Proof Sketch (Intuitive)

The principle can be understood by considering the behavior of analytic functions. If an analytic function had a maximum modulus at an interior point z₀ ∈ D, and |f(z₀)| = M, then the function would be locally bounded by M. However, analytic functions tend to grow away from points where they are "small" unless they are constant.

A more rigorous proof often uses the open mapping theorem or considers the behavior of log|f(z)|, which is a harmonic function. For a harmonic function, local maxima must occur on the boundary.

Consider a small disk around an interior point z₀. If f(z) is not constant, then by the open mapping theorem, f(z) maps this small disk to some region in the w-plane. If |f(z₀)| were the maximum, then all values in the mapped region would have modulus less than or equal to |f(z₀)|. However, the open mapping theorem states that analytic non-constant functions map open sets to open sets. This implies that there must be points in the mapped region with modulus strictly greater than |f(z₀)|, contradicting the assumption that |f(z₀)| is the maximum. Therefore, the maximum must occur on the boundary.

Consequences and Applications

The Maximum Modulus Principle has several important consequences:

• Non-Constant Analytic Functions: If f(z) is a non-constant analytic function on a bounded domain D, then |f(z)| does not attain its minimum value in the interior of D unless f(z) = 0 at that point. This is because if f(z₀) = 0 for some z₀ ∈ D, then |f(z₀)| = 0, which is the smallest possible modulus. If f(z)` is never zero in D, then 1/f(z)` is analytic on D. By the Maximum Modulus Principle applied to 1/f(z), the maximum modulus of 1/f(z) occurs on the boundary. This means the minimum modulus of f(z) also occurs on the boundary.
• Liouville's Theorem: A direct corollary is Liouville's Theorem. If an analytic function f(z) is defined on the entire complex plane (an unbounded domain) and is bounded (i.e., |f(z)| ≤ M for all z), then f(z)` must be a constant function.

  Proof: Consider a large disk of radius R. By the Maximum Modulus Principle, |f(z)| is bounded by M on this disk. Now consider the derivative f'(z). For any z₀ inside the disk, the Cauchy integral formula for derivatives gives f'(z₀) = 1/(2πi) ∫_∂DR f(z)/(z - z₀)² dz. The length of the contour is 2πR. The maximum value of |f(z)| on the contour is M. The minimum distance from z₀ to the contour is greater than 0. Thus, |f'(z₀)| ≤ (1/(2π)) * (M / (min distance)²) * 2πR. As R → ∞, this bound does not necessarily go to zero if we only know f is bounded. However, a more refined argument using the Mean Value Theorem for harmonic functions or direct estimation of the integral shows that |f'(z₀)| can be made arbitrarily small by choosing a large enough disk, implying f'(z₀) = 0 for all z₀. If the derivative is zero everywhere, the function is constant.

• Determining Function Behavior: The principle is useful for estimating the maximum values of functions, especially in regions where direct computation is difficult. If we know a function is analytic on a disk and can compute its values on the boundary, we know its maximum modulus within the disk.