7
$\begingroup$

I would like to understand the application of the Hardy-Littlewood Circle Method to Birch's celebrated theorem on forms in many variables, namely the statement that the point counting function for a smooth projective hypersurface of degree $d$ in $\mathbb{P}^N$ defined over $\mathbb{Q}$ satisfies the usual asymptotic if $N\gg_d 1$ (and in particular that such surfaces satisfy the Hasse principle). I am following an exposition of the quartic case in Timothy Browning's Quantitative Arithmetic of Projective Varieties, Chapter 8. Unless I am highly mistaken, the proof presented there requires at least some modifications, and I would like to confirm whether I am right when I say this.

My question concerns the minor arcs estimate. This depends on Lemma 8.7, which gives an estimate $$S\left(\frac{a}{q}+z\right)\ll_\varepsilon B^{n+\varepsilon}q^{-\frac{n}{24}}\min\{1,|z|B^4\}^{-\frac{n}{24}}$$ for the relevant exponential sum. This yields (at the end of section 8.2.1) the estimate $$\int_{\mathfrak{m}(\Delta)}|S(\alpha)|\mathrm{d}\alpha\ll_\varepsilon\sum_{q\leq B^2}q^{1-\frac{n}{24}}\int_{-\frac{1}{qB^2}}^{\frac{1}{qB^2}}\min\{1,|z|B^4\}^{-\frac{n}{24}}\mathrm{d}z(*)$$ from where the author concludes (using that we are actually only summing over values of $q$ and $z$ that are admissible for the minor arcs; in particular this includes all $q>B^\Delta$) that the contribution of the minor arcs is $\ll_\varepsilon B^{n-4-\Delta(\frac{n}{24}-2)+\varepsilon}$, which beats the desired $B^{n-4}$ if $n>48$.

But I don't see how this follows from $(*)$, and in fact, unless I'm very much mistaken, it can't follow from $(*)$: if $n>48$, the integral that appears on the RHS of $(*)$ explodes near $0$, since the $\min$ is actually $|z|B^4$, for small $z$! (And we actually need to care about small $|z|$ for some minor arcs, namely those corresponding to $q>B^\Delta$!)

Does my concern make sense? If so, how can we fix this proof? I am particularly interested in understanding exactly how the lower bound $48$ appears in the proof, so understanding these technical details is important. Thanks in advance!

$\endgroup$
1
  • $\begingroup$ I think this is the same issue that came up in this question (see the comments!)! I never actually checked that the exponents I’d eyeballed were right but anyway hopefully this helps —- lemme know if not and I can take a look at it again! $\endgroup$ Commented Jan 7, 2022 at 5:24

0

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.