Explicit seam-region error constant
ProvedRSSeam.E1_const_learda-rh65k-transfernumber-theoryriemann-zeta
For every real ,
This supplies a numerical remainder constant for the height-range seam.
Preamble
import Definitions.Def_ArdaRH65K_RS_B2
import Definitions.Def_ArdaRH65K_RS_B3
import Definitions.Def_ArdaRH65K_RS_C0
import Definitions.Def_ArdaRH65K_RS_Entire
import Definitions.Def_ArdaRH65K_RS_Fold
import Definitions.Def_ArdaRH65K_RS_Kernel
import Definitions.Def_ArdaRH65K_RS_Mellin
import Definitions.Def_ArdaRH65K_RS_Strip
import Mathlib
namespace RSInt
end RSInt
/- RSSeam_B3Bound.lean -- lane SEAM: the C0-corrected Riemann-Siegel remainder bound with a LOWER
starting height, t >= 509 (saddle variable a = sqrt(t/2pi) >= 9) instead of t >= 10000 (a >= 39).
Imports `RS_B3Bound` read-only and reuses its lemmas; nothing in lane RS is edited.
## The headlines (no `sorry`, no `native_decide`, no new axioms)
`rs_remainder_C0_alpha_seam` : for t >= 509, 0 < p = rsFrac t, cos(2 pi p) ≠ 0, and every phase
phi with |phi - rsThetaMain t| <= 1/t,
|2 Re(e^{i phi} rsRem t) - (-1)^(N+1) a^(-1/2) Psi(p)| <= (13/20) a^(-3/2).
`rs_remainder_C0_seam` : the same in t,
|2 Re(e^{i phi} rsRem t) - (-1)^(N+1) (t/2pi)^(-1/4) Psi(p)| <= (13/5) t^(-3/4).
## What changed against `RS_B3Bound` (the error budget re-run at a >= 9)
* near/far split of the error integral at |v| = a/3 (was a/9); near the saddle the amplitude
bound |rsPhi - 1| <= e e^e still applies because |w| = sqrt 2 |v| <= a/2;
* near: e <= |w| (13/25 + 17/4 |w|^2)/a and e <= 27/5 v^2 + 1/600, so the integrand is
<= e^{1/600}/(pi a) (13/25 + 17/2 v^2) e^{-7 v^2};
* far: |integrand| <= 5 e^{-pi a^2/18} e^{-pi v^2/2}, and e^{-x} <= 8!/x^8 at x >= 14;
* |rsE1 a p| <= (49/200 + 1/200)/a = (1/4)/a for a >= 9 (was (9/50)/a for a >= 39);
* |rsJ p| <= 4 (unchanged, `RSInt.norm_rsJ_le`); phase term 8|delta| <= 4/(pi a^2) <= (3/20)/a;
* budget 2 (1/4) + 3/20 = 13/20 in the a-form; (2 pi)^(3/4) <= 4 gives 13/5 in t.
Untrusted numerics (scratchpad/seam/remainder.py): the true C0-corrected remainder is
<= 0.121 t^(-3/4) on samples in [200, 10100]; so 13/5 is honest with a ~20x margin.
conjecture1_proved = False. An explicit finite-height remainder inequality; nothing here bears on
the Riemann Hypothesis.
-/
open Complex MeasureTheory Filter Topology Set
open scoped Real ComplexConjugate
noncomputable section
section
namespace RSSeam
end RSSeam
open RSSeam
open RSInt
Formal statement
theorem RSSeam.E1_const_le {a : ℝ} (ha : 9 ≤ a) :
Real.exp (1 / 600) / (π * a) * (13 / 25) * Real.sqrt (π / 7) +
Real.exp (1 / 600) / (π * a) * (17 / 2) * (Real.sqrt (π / 7) / (2 * 7)) +
5 * Real.exp (-(π * a ^ 2 / 18)) * Real.sqrt (π / (π / 2)) ≤ 1 / 4 / a := by sorry
Source