Class number of real quadratic fields of explicit discriminant

summary

The gist

In this paper we are interested in the class numbers of a family of real quadratic fields for which the square roots of the discriminants have a known expansion in continued fraction…

In short

The paper investigates class numbers h(D) for a specific family of real quadratic fields defined by discriminants D(b, s, k). The main result proves that h(D) is greater than 1 for almost all cases in this family, with only a finite number of exceptions. This is established by analyzing the continued fraction expansions of the square root of the discriminant and showing contradictions when class numbers are assumed to be one.

Key concepts

Family Description D(b, s, k)
This describes a specific set of real quadratic fields where the discriminant D is calculated using parameters b, s, and k. A key feature is that if (1, ωD) is the standard integral basis for the ring of integers of Q(√D), then certain coefficients in this basis are divisible by b.
Continued Fraction Expansion
This mathematical tool represents a number as a sequence of integers and fractions. The paper uses these expansions for the square root of D to derive constraints on the discriminant. Contradictions found in these expansions are used to prove that class numbers must be greater than one.
Class Number h(D)
The class number is an important invariant for quadratic fields, measuring how far the ring of integers is from being a unique factorization domain. The paper aims to show that for this specific family of discriminants, this value h(D) is typically greater than 1.
Family n^2 + 1
This is a specific case analyzed in the paper where the discriminant D equals n^2 + 1, with n being an odd integer. The analysis shows that for these fields, the class number h(D) is greater than 1 unless D equals 2p (where p is an odd prime) or D takes on specific other forms.

Terminology used across episodes

This episode discusses

The paper

Class number of real quadratic fields of explicit discriminant · Read on arXiv

RICCARDO BERNARDINI

Transcript

Introduction to the show: ident: Mathematics. The day in mathematics: new theorems, new proofs, and what they open.

Ada: I'm Ada.

Emmy: And I'm Emmy, here with Kolya.

Ada: We're looking at the paper: "Class number of real quadratic fields of explicit discriminant".

Emmy: Here's the gist of the paper. In this paper we are interested in the class numbers of a family of real quadratic fields for which the square roots of the discriminants have a known expansion in continued fraction, and we prove that h(D) > 1, with possibly a finite number of exceptions Theorem 1.1 states that for the family of real quadratic fields with discriminants described by D(b, s, k), if b ≥ 0 and s, k > 0 are such that b + s is odd and D is squarefree, then h(D) > 1 unless b = 0 and s = 1 Family Description and Context The paper focuses on the family of real quadratic fields with discriminants described in

11, Theorem 6: by McLaughlin and Zimmer.

Ada: The discriminant is given by D(b, s, k) = (4bs + 1)k + (b(4bs + 1)k + s) squared.

Emmy: A common aspect between the families (3) and (4), which is DBiró(b, s, k), is that if (1, ωD) is the standard integral basis of the ring of integers of Q(√D), then the coefficients bj for j 6= 0, l are divisible by b.

Ada: A slight difference is that the continued fraction of the square root of (3) has a factor 2 which appears in the coefficients.

Paper summary: Ada: So we're looking at this paper called "Class number of real quadratic fields of explicit discriminant". It claims they can prove that for a specific family of real quadratic fields, the class number is always greater than one, except for maybe a very small, finite set of exceptions.

Emmy: That’s the main thrust. They are focusing on these families where you know exactly how to expand the square root of the discriminant using continued fractions. The real weight here is proving that h(D) is not always one, which is a big deal because class numbers are notoriously mysterious in number theory.

Kolya: So what does this mean for us? It suggests that if you pick a discriminant from this specific structure, you're unlikely to find one with a class number of one unless it falls into some very narrow circumstances they identify.

Ada: Exactly. The paper focuses on the family described by D(b, s, k), where the discriminant is defined by D being (4bs + one)k plus some other term involving b, s, and k <ref:2412.06351#pg2>. The crucial part of their main result is Theorem one point one which states that if you pick positive numbers for b, and s and k, such that when you add them, they're odd, and the discriminant is squarefree, then the class number h(D) must be greater than one unless you have the very specific case where b is zero and s is one <ref:2412.06351#pg1>.

Emmy: And they make a strong connection to other families. They mention that if you look at a slightly different family, DBiró(b, s, k), it shares a common feature where the coefficients of the ring of integers are divisible by b. But they point out that their family has a specific difference related to the continued fraction expansion of the square root of D.

Kolya: The paper mentions using certain primes like forty-one sixty-one one hundred seventy-five and one thousand eight hundred sixty-one in their proof <ref:2412.06351#pg2>. That’s interesting because those same primes show up when they look at a simpler family called n squared plus one.

Ada: Right. They actually analyze the n squared plus one case separately. They find that if the discriminant is D = n squared plus one, where n is odd and squarefree, then the class number will be greater than one unless you have a very specific condition on n.

Paper summary: Emmy: They classify the cases where the class number equals one for this family. It turns out that it happens when D is two times an odd prime, or in some other specific set of values for n. That helps them narrow down the possibilities significantly.

Kolya: So, from a probabilistic viewpoint, this suggests that for a randomly chosen discriminant within these families, the probability of having class number one is very low. The conditions they find restrict those exceptions quite tightly.

Ada: The methodology they use to get this result involves constructing some algebraic integers and then using explicit Dirichlet characters with conductors like forty-one sixty-one one hundred seventy-five and one thousand eight hundred sixty-one to systematically eliminate residue classes <ref:2412.06351#pg2>. This relies on a lemma showing that if you have an element beta in the ring of integers whose norm is less than 2n, it must be associated with an integer <ref:2412.06351#pg1>.

Emmy: And for the n squared plus one family, they use Shintani's zeta function and Bernoulli polynomials to evaluate something called zeta P(K)(zero chi) <ref:2412.06351#pg2>. It’s a deep way to connect analytic number theory tools to these specific field structures.

Kolya: What I wonder is how those bounds from the continued fraction coefficients actually lead to a contradiction in the final step for Theorem one point one <ref:2412.06351#pg1>? That seems like the core engine of their proof.

Ada: They use the continued fraction expansions of sqrt D or one plus root D squared for this contradiction. For cases where D is a square modulo some number q, they get inequalities, specifically inequality twenty-four and inequality twenty-five, which they show cannot both be true at the same time.

Emmy: The structure of the argument is that these bounds derived from the continued fraction coefficients simply cannot hold simultaneously for any discriminant satisfying those conditions. This forces the conclusion that h(D) must be greater than one unless we hit that single exception where b is zero and s is one.

Kolya: So, if you were looking at this for practical application, it means if you can find a discriminant matching this explicit form, you have a high confidence that the class number isn't just one. The caveat here is that they only prove it for the family where the square roots of the discriminants have known continued fraction expansions.

Paper summary: Ada: That’s right. They explicitly state their limitation: they are focused on fields where those continued fraction expansions are known, and they found that these conditions strongly favor class numbers greater than one, leaving only that one specific case as a potential exception.

Emmy: So the implication is that for this family of discriminants, finding fields with class number one is much harder than people might expect. It places a strong constraint on the structure of those fields when you have information about their square root expansions.

Kolya: If someone was just looking at the numbers, they'd see these primes forty-one sixty-one one hundred seventy-five and one thousand eight hundred sixty-one appearing in both their main theorem and their analysis of n squared plus one <ref:2412.06351#pg2>. That connection is a nice piece of structural information.

Ada: It is indeed. The fact that the prime factors involved in Theorem one point five are the same as those in Theorem one point one suggests this isn't just a coincidence, but rather a deep structural link between these different families of discriminants they are studying <ref:2412.06351#pg3,The fact that the prime factors>.

Emmy: So to wrap up on what they’ve done with "Class number of real quadratic fields of explicit discriminant", they’ve established that for their family D(b, s, k), the class number is greater than one unless b is zero and s is one. This result relies on connecting continued fraction properties to contradictions in analytic inequalities.

Kolya: And the caveat they give you is that this proof only holds when you have those known continued fraction expansions for the square roots of your discriminants. It’s a very specific setting for this kind of certainty.

Ada: That's the essence of it. The paper gives us a strong statement about class numbers in these structured families, pointing toward h(D) being greater than one most of the time, which is what we expected based on general theory but now with much more explicit constraints.

Emmy: It’s about showing that having explicit information about the square root expansion allows you to use tools like continued fractions to rule out class number one cases effectively. It's a nice demonstration of how specific structural knowledge can constrain abstract invariants.

Conclusion: Ada: So we've been looking at this paper, "Class number of real quadratic fields of explicit discriminant." Basically, they found that for these specific families of numbers, the class number is almost always greater than one.

Emmy: They’re focusing on these real quadratic fields where you know exactly how to expand the square root of their discriminant using continued fractions. It's about those specific structural forms D has.

Kolya: So, what does this mean for us? It suggests that if you pick a discriminant from this family, it's pretty unlikely to have a class number of one unless it falls into some very tight exceptions they identified.

Ada: Right. The main result is Theorem one point one, which says the class number h(D) is greater than one for the family described by D unless you have that specific case where b equals zero and s equals one.

Emmy: They use continued fraction expansions of sqrt D to build a contradiction. Specifically, they look at inequalities from the coefficients and show those bounds just can't happen at the same time. That forces the class number up.

Kolya: I'm interested in that contradiction part. What does that actually tell us about how these numbers behave? Is it like a general rule for all quadratic fields?

Ada: Not for all of them, no. It only applies to this specific family where you have that explicit information about the square root expansion. It’s very constrained research, but the result is strong for those fields.

Emmy: The authors also did some work on a simpler family, n squared plus one. They found that if n is odd and squarefree, the class number is greater than one unless n fits a specific small set of conditions.

Kolya: So they're connecting these two different families through shared prime factors like forty-one and sixty-one. It makes sense that those primes are appearing in both results.

Ada: Exactly. It shows there’s a deep structural link between the two families they studied, even though the discriminants look different on the surface.

Emmy: The implication is that having explicit information about how you calculate sqrt D really lets you use tools like continued fractions to rule out class number one cases quite effectively.

Kolya: So, for someone just listening, what does this change? It means if you can find a discriminant matching this form, you have high confidence the class number isn't one, unless it’s that very specific b and s combination.

Ada: That’s right. They’ve shown that with enough structural information about the discriminant, you can rule out most possibilities for the class number being one in these families.

Emmy: Now we need to see if this extends further, or if there are other structures where this kind of explicit analysis might work.

More episodes

← Home