Curves in ${\mathbb P}^n$ of analytic spread at most $n$

This paper investigates closed subschemes of dimension one in projective space $\mathbb{P}^n$ defined by ideals with analytic spread at most $n$, proving that under mild conditions their powers have positive depth, their Rees rings have regularity at most one, and their fiber cones are Cohen-Macaulay, results which notably apply to monomial curves in $\mathbb{P}^3$.

The Gist

Imagine you are an architect designing a city. In this city, the buildings are mathematical shapes called curves, and they exist in a vast, multi-dimensional space (like a 4D or 5D universe).

The paper you're asking about is like a blueprint inspection report. The authors, Marc Chardin and Clare D'Cruz, are checking these curves to see how "stable" and "predictable" they are when you start stacking them or multiplying them.

Here is the breakdown of their findings using simple analogies:

1. The Setting: The "Curve" and the "Blueprint"

• The Curve ($X$): Think of a curve as a winding road or a wire sculpture. In math, it's a 1-dimensional object floating in a higher-dimensional space ($P^n$).
• The Defining Ideal ($I$): This is the set of instructions (equations) used to build the curve. If you want to build a specific wire sculpture, you need a list of rules (e.g., "must pass through point A," "must be flat").
• The "Analytic Spread": This is the paper's main character. Imagine you have a huge pile of blueprints (equations) to build the curve. The "analytic spread" asks: "What is the absolute minimum number of unique instructions I actually need to describe the shape's core behavior?"
  • If the spread is low, the shape is simple and tightly controlled.
  • If the spread is high, the shape is chaotic and needs many independent rules to define it.

2. The Main Discovery: The "Sweet Spot"

The authors focus on curves where the number of essential instructions is at most equal to the dimension of the space (e.g., in a 4D space, the curve needs no more than 4 essential rules).

They found that when a curve fits this "sweet spot," it behaves beautifully. Here is what happens:

A. The "Depth" Never Dries Up

In math, "depth" is like the structural integrity of a building. If depth is zero, the building is about to collapse (mathematically speaking, it has "holes" or singularities that break the rules).

• The Finding: For these well-behaved curves, no matter how many times you multiply the instructions (building higher and higher towers of the same shape), the structure never collapses. It always has at least a little bit of "depth" (stability).
• The Metaphor: Imagine stacking blocks. Usually, if you stack too many, the tower wobbles and falls. But for these special curves, the tower is made of a magical material that stays stable no matter how high you go.

B. The "Regularity" is Low (The "Speed Limit")

"Regularity" measures how complicated the equations get as you build higher powers of the curve.

• The Finding: The complexity of these curves stays very low. The "Rees algebra" (a fancy way of organizing all the powers of the curve) is defined by very simple, short equations (linear and quadratic).
• The Metaphor: Usually, as you build a complex skyscraper, the blueprints get thicker and more confusing. For these curves, the blueprints stay simple. It's like building a skyscraper using only basic Lego bricks rather than custom-molded, intricate parts.

C. The "Fiber Cone" is Perfectly Smooth

The "fiber cone" is a mathematical object that represents the "shape" of the curve without the messy details of the surrounding space.

• The Finding: This object is "Cohen-Macaulay." In plain English, this means it is perfectly uniform and has no hidden cracks.
• The Metaphor: Think of a smooth marble statue. A "Cohen-Macaulay" statue has no internal fractures; if you tap it, it rings true. The authors proved that for these specific curves, the underlying shape is always a perfect, unbroken marble statue.

3. The Real-World Test: Monomial Curves

The authors didn't just do theory; they tested this on Monomial Curves (curves defined by simple powers, like $x=t^2, y=t^3$).

• In 3D Space ($P^3$): They found that every monomial curve fits the "sweet spot." They are all stable, simple, and perfect.
• In 4D Space ($P^4$): This is where it gets interesting.
  • Case A (The Good): Some curves in 4D still fit the rules. They are stable and simple.
  • Case B (The Bad): They found an infinite family of curves in 4D that look similar but break the rules. Their "analytic spread" is too high (4), and they become unstable. Their blueprints get messy, and their structures develop cracks (depth issues).

Summary: Why Does This Matter?

This paper is like a quality control manual for mathematicians building shapes in high-dimensional spaces.

It tells us:

1. If your curve is "tightly defined" (low analytic spread), you can rest easy. Its powers will always be stable, its equations will remain simple, and its shape will be perfect.
2. If you step outside these rules (like in certain 4D examples), chaos ensues. The equations get messy, and the structure becomes unpredictable.

The authors essentially drew a line in the sand: "Stay on this side of the line, and your mathematical buildings will stand forever. Cross it, and you might find a crack in the foundation."

They also used computer software (Macaulay2) to simulate these buildings, proving that their theoretical predictions hold up in the real (digital) world.

Technical Summary

Here is a detailed technical summary of the paper "Curves in $\mathbb{P}^n$ of Analytic Spread at Most $n$" by Marc Chardin and Clare D'Cruz.

1. Problem Statement and Context

The paper investigates the algebraic properties of blowup algebras associated with closed subschemes $X$ of dimension one (curves) in projective space $\mathbb{P}^n$. Specifically, the authors focus on curves whose defining ideal $I \subseteq S = k[x_0, \dots, x_n]$ satisfies the condition that the analytic spread of the ideal localized at the maximal homogeneous ideal, denoted $a(I\mathfrak{m})$, is at most $n$.

The core problem addresses the behavior of:

• Depth functions: The depth of powers of the ideal, $d_I(t) = \text{depth}(S/I^t)$.
• Castelnuovo-Mumford regularity: The regularity of the Rees algebra $\mathcal{R}(I)$ and the associated graded ring $\mathcal{G}(I)$.
• Cohen-Macaulayness: Whether the fiber cone $\mathcal{F}(I)$ and the Rees algebra are Cohen-Macaulay.

The study is motivated by the concept of analytic deviation, defined as $ad(I) = a(I) - \text{ht}(I)$. The authors focus on the case where the analytic deviation is at most 1 (since $\text{ht}(I)=2$ for curves in $\mathbb{P}^n$ and $a(I) \leq n$, but the specific condition $a(I) \leq n$ in the context of the paper implies a tight bound on the deviation relative to the embedding dimension).

2. Methodology

The authors employ a combination of commutative algebra techniques, specifically:

• Approximation Complexes: They utilize the Z-complex and M-complex defined by Herzog, Simis, and Vasconcelos. These complexes relate the Koszul homology of the generators of an ideal to the structure of the Rees algebra and the associated graded ring.
• Local Cohomology and Spectral Sequences: The proof of the main lemma involves comparing spectral sequences derived from double complexes to compute local cohomology modules of the symmetric algebra, which helps determine the regularity and depth of the Rees algebra.
• Reduction Theory: The concept of minimal reductions and reduction numbers ($r(I)$) is central. The authors construct specific reductions $J$ of the ideal $I$ to analyze the structure of the fiber cone and the powers of $I$.
• Gröbner Bases and Initial Ideals: In the examples section, the authors use Gröbner basis computations (graded reverse lexicographic order) to determine initial ideals, Hilbert series, and verify properties like the d-sequence property and the non-membership of specific elements in ideals.

3. Key Contributions and Main Results

A. General Theoretical Framework (Lemma 2.2)

The authors establish a lemma for a Noetherian unmixed local ring $(R, \mathfrak{m})$ and an ideal $J$ generated by $r+1$ elements with $\text{depth}(J) \geq r$. Under conditions where $J$ is of linear type locally at primes of height equal to $\text{ht}(J)$, they prove:

• $J$ is of linear type and $\text{reg}(\mathcal{R}(J)) = 0$.
• If $I = J^{\text{sat}}$ (the saturation of $J$) and $\text{depth}(R) \geq r+2$, then the depth of powers of $I$ is at least 2, and $\text{reg}(\mathcal{R}(I)) \leq 1$.

B. Main Theorem (Theorem 2.3)

For a closed subscheme $X \subset \mathbb{P}^n$ of dimension 1, locally defined by at most $n$ equations, with defining ideal $I$ such that $I_{\mathfrak{p}}$ is a complete intersection for all minimal primes $\mathfrak{p}$, if the analytic spread $a(I\mathfrak{m}) \leq n$, then:

1. Positive Depth: $\text{depth}(S/I^t) > 0$ for all $t \geq 1$. Consequently, the limit depth is 1 (unless $I$ is a complete intersection, in which case it is higher).
2. Regularity Bound: $\text{reg}(\mathcal{R}(I)) \leq 1$. This implies the reduction number $r(I) \leq 1$, and the Rees algebra is defined by linear and quadratic equations.
3. Cohen-Macaulay Fiber Cone: The fiber cone $\mathcal{F}(I)$ is Cohen-Macaulay.
4. Generator Count: An explicit formula is provided for the minimal number of generators of $I^t$:
   $$\mu(I^t) = \binom{t+a-1}{a-1} + (\mu(I) - a)\binom{t+a-2}{a-1}$$
   where $a = a(I\mathfrak{m})$.

C. Application to Monomial Curves in $\mathbb{P}^3$ (Example 2.4)

The authors apply the main theorem to monomial curves in $\mathbb{P}^3$. Citing a result by Gimenez, Morales, and Simis, they note that the analytic spread of such curves is at most 3. Therefore, all monomial curves in $\mathbb{P}^3$ satisfy the conditions of Theorem 2.3, ensuring their Rees algebras have regularity at most 1 and their fiber cones are Cohen-Macaulay.

4. Counter-Examples and Limitations (Section 3)

The paper provides a nuanced analysis by constructing examples in $\mathbb{P}^4$ to demonstrate that the assumptions in Theorem 2.3 are necessary.

• Example 3.1 (Success Case): Monomial curves $C(1, 2, 3, 3a)$.

  • The defining ideal is generated by a d-sequence.
  • Analytic spread is 4 (maximal for $\mathbb{P}^4$), but the reduction number is 0.
  • $\text{reg}(\mathcal{R}(I)) = 0$ and the fiber cone is a polynomial ring (Cohen-Macaulay).
  • This confirms the theorem holds when the "linear type" condition is met via d-sequences.

• Example 3.2 (Failure Case): Monomial curves $C(1, 2, 3, 3a+1)$.

  • Analytic spread is 4.
  • The reduction number is 2 (violating the $r(I) \leq 1$ conclusion of the main theorem).
  • $\text{reg}(\mathcal{R}(I)) \geq 2$.
  • The fiber cone is Cohen-Macaulay, but the Rees algebra regularity is higher. This example shows that having $a(I) \leq n$ is not sufficient without the specific local complete intersection conditions or d-sequence properties.

• Example 3.3 (Failure Case): Monomial curves $C(1, 2, 3, 3a+2)$.

  • Analytic spread is 4.
  • Reduction number is 2.
  • $\text{reg}(\mathcal{F}(I)) = 2$ and $\text{reg}(\mathcal{R}(I)) \geq 2$.
  • This further illustrates that for certain monomial curves in $\mathbb{P}^4$, the "nice" behavior (low regularity) predicted by the main theorem fails.

5. Significance

• Unification of Depth and Regularity: The paper provides a precise characterization of the limit depth and regularity for a broad class of curves, linking the analytic spread directly to the homological properties of the blowup algebras.
• Resolution of Monomial Curves in $\mathbb{P}^3$: It settles the question for monomial curves in $\mathbb{P}^3$, proving they universally possess Cohen-Macaulay fiber cones and low regularity Rees algebras.
• Sharpness of Bounds: Through the $\mathbb{P}^4$ examples, the authors demonstrate that the condition $ad(I) \leq 1$ (or the specific structural assumptions) is sharp. Without these conditions, the regularity can jump, and the reduction number can exceed 1.
• Computational Insight: The use of Macaulay2 and explicit Gröbner basis calculations in the examples provides concrete evidence for the theoretical bounds, bridging the gap between abstract commutative algebra and computational algebraic geometry.

In summary, the paper establishes that for curves in $\mathbb{P}^n$ with analytic spread at most $n$ (and satisfying mild local conditions), the blowup algebras exhibit optimal homological behavior (regularity $\leq 1$, Cohen-Macaulay fiber cones), while explicitly delineating the boundaries where this behavior breaks down in higher-dimensional projective spaces.