Pólya Groups in Number Fields

• Class Groups of Number Fields are structured by the Pólya group, defined as the subgroup generated by all Ostrowski ideals of uniform norm.
• Explicit classifications for imaginary and real quadratic, as well as multi-quadratic fields, are achieved using genus theory and rigorous analytic bounds.
• Cohomological methods and class number formulas link classical genus theory with modern computations, offering insights into open problems in field extensions.

The Pólya group $\mathrm{Po}(K)$ of a number field $K$ is the subgroup of the ideal class group $\mathrm{Cl}(K)$ generated by the classes of Ostrowski ideals, that is, products of all prime ideals of $K$ with the same absolute norm. The Pólya index $[\mathrm{Cl}(K) : \mathrm{Po}(K)]$ measures the extent to which the class group is generated by such “same-norm” products. This concept is motivated by the classical “one class per genus” problem and provides a structural lens for examining the generation of ideal class groups across families of number fields (Akbary et al., 15 Aug 2025).

1. Definitions and Structural Formulas

For a number field $K$ with ring of integers $\mathcal{O}_K$, the Ostrowski ideal associated to a prime power $q$ is

$$\Pi_{(q)}(K) = \prod_{N_{K/\mathbb{Q}}(\mathfrak{p}) = q} \mathfrak{p},$$

with the product taken over all prime ideals $\mathfrak{p}$ of norm $q$. The Pólya group is defined as

$$\mathrm{Po}(K) = \langle [\Pi_{(q)}(K)] : q = p^n,\, p \text{ prime},\, n \geq 1 \rangle \leq \mathrm{Cl}(K).$$

The Pólya index is then $[\mathrm{Cl}(K):\mathrm{Po}(K)]$. In particular, $\mathrm{Po}(K) = \mathrm{Cl}(K)$ if and only if the class group is generated by Ostrowski ideals, reflecting an analog of “one class in each genus.”

When $K/\mathbb{Q}$ is Galois with group $G$, there is an exact sequence [Zantema]: $$0 \to H^1(G, U_K) \to \bigoplus_{p | d_K} \mathbb{Z}/e_p(K/\mathbb{Q}) \mathbb{Z} \to \mathrm{Po}(K) \to 0,$$ giving

$$|\mathrm{Po}(K)| = \frac{\prod_{p | d_K} e_p(K/\mathbb{Q})}{|H^1(G, U_K)|},$$

where $e_p(K/\mathbb{Q})$ is the ramification index and $U_K$ is the unit group (Akbary et al., 15 Aug 2025).

2. Finiteness Theorems for Fixed Pólya Index

For infinite families of number fields, Pólya index controls are obtained by analytic class number bounds and genus theory. For infinite families of solvable CM-fields, if $h_K^-/h_{K^+}$ grows rapidly with discriminant, and taking into account ramification indices, one obtains that for any fixed $t$, only finitely many fields $K$ in the family have $[\mathrm{Cl}(K):\mathrm{Po}(K)] = t$. A similar philosophy applies to families of Galois fields with mild regulator growth and to real quadratic fields of extended Richaud–Degert (R-D) type, where explicit class number lower bounds and divisor count upper bounds for $\mathrm{Po}(K)$ yield similar finiteness (Akbary et al., 15 Aug 2025).

3. Explicit Classifications in Families

The paper achieves several unconditional classifications:

• Imaginary quadratic, bi-quadratic, and tri-quadratic fields with Pólya index one: Complete lists for these cases exist, following the fact that in genus-one abelian CM-fields, $\mathrm{Po}(K) = \mathrm{Cl}(K)$ iff the genus number $g_K$ equals $h_K$. For instance, there are precisely 57 imaginary bi-quadratic and 17 tri-quadratic fields with Pólya index one (Akbary et al., 15 Aug 2025).
• Real quadratic fields of extended R-D type with Pólya index one: All such fields with discriminant $D<6.3 \times 10^{16}$ (with possibly at most one exception outside this range) are classified explicitly, yielding 269 fields. This extends the prior classification for narrow R-D type (Akbary et al., 15 Aug 2025).
• Imaginary quadratic fields with Pólya index two (under GRH): There are exactly 161 such fields, verified via explicit discriminant bounds and computational class group evaluation (Akbary et al., 15 Aug 2025).

Family	Index One: Complete List Exists?	Method/Note
Imaginary quadratic	Yes	Genus theory
Imaginary bi-quadratic	Yes (57 fields)	Explicit parameter count
Imaginary tri-quadratic	Yes (17 fields)	Parameter restriction
Imaginary quadratic, index two	Yes (161 fields, under GRH)	Computational, via GRH
Real quadratic (extended R-D)	Yes (269 fields, except at most one)	Explicit bound/algorithm

4. Analytic and Cohomological Ingredients

• The Pólya group size $|\mathrm{Po}(K)|$ divides the genus number $g_K$, so $\mathrm{Po}(K) = \mathrm{Cl}(K)$ if and only if $g_K = h_K$ in abelian CM-cases—a principle used for classification.
• Cohomological exact sequences connect $\mathrm{Po}(K)$ to $H^1(G, U_K)$, particularly relevant in Galois and multi-quadratic fields, reducing the problem to effective computation and analysis of ramification.
• For real quadratics of extended R-D type, use of analytic class number formulas (including Tatuzawa’s $L$-function bounds) gives explicit lower bounds, and upper bounds for $|\mathrm{Po}(K)|$ are obtained via the divisor function $\tau(d_K)$.
• Conditional results, especially in the computational classification of imaginary quadratic fields of index two, rely on GRH and explicit implementation through PARI-GP (Akbary et al., 15 Aug 2025).

5. Broader Implications for Class Group Generation

The index $[\mathrm{Cl}(K):\mathrm{Po}(K)]$ quantifies the deviation of the class group from being generated by “same-norm” prime products. It emerges that only for very special fields is the class group generated this way (Pólya index one or two), and in almost all large-degree or large-discriminant families, the Pólya group is much smaller than the full class group. This observation fits well into heuristics à la Odlyzko for large class groups and complements statistical studies on class group generation (Akbary et al., 15 Aug 2025).

6. Extensions and Open Questions

• Extending explicit classification to non-abelian CM-fields and higher-degree fields (cubic, quartic) is open.
• Removing GRH from the index two imaginary quadratic classification remains an outstanding challenge.
• Understanding the distribution of $\mathrm{Po}(K)$ inside $\mathrm{Cl}(K)$—for instance, the statistical behaviour in large degree or random families—may provide new insights into class group randomness and realizability.
• Higher-dimensional analogues using Ostrowski ideals or genus-theoretic approaches are suggested as promising directions for future study (Akbary et al., 15 Aug 2025).

7. Connection to Classical Problems and Genus Theory

The study of Pólya groups is intimately linked to classical genus theory and the “one class in each genus” paradigm. The explicit connection between Ostrowski ideals and genus subgroups allows translation of deep analytic results (such as bounds for class numbers and $L$-values) into concrete group-theoretic conclusions about class group generation. Thus, the theory of Pólya groups bridges arithmetic statistics, explicit class group computation, and classical algebraic number theory (Akbary et al., 15 Aug 2025).