---
title: 'Quasi-High-Energy Limit: Refined Asymptotics'
url: https://www.emergentmind.com/topics/quasi-high-energy-limit
type: topic
---

# Quasi-High-Energy Limit: Refined Asymptotics

Searching arXiv for the cited papers and closely related uses of “quasi-high-energy limit.”
“Quasi-high-energy limit” is not a single universally standardized notion. In the cited literature it denotes a refinement of a strict high-energy asymptotic regime in which one retains structures that the leading approximation discards: the small-\(x\) boost expansion of quasi-classical gluon fields in QCD, the regime \(|t|\ll s\) for infrared-singular gauge amplitudes, the first power-suppressed sector beyond Sudakov approximation, quasi-mode constructions for differential operators on vector bundles, the regime \(E\gg E_{rt}(N)\) in collision-induced tunneling, quasi-Regge kinematics for open spin chains, and the large-\(b_0\) high-energy regime of nonlinear trident production [2402.17568], [1201.2841], [1709.01092], [1109.2104], [1503.06339], [1801.04872], [2007.08492]. A plausible unifying description is that the term labels asymptotic limits that remain high-energy, but are enlarged so as to keep subleading kinematics, power corrections, dense backgrounds, or non-eigenstate data.

## 1. Terminological scope and common structure

In gauge-theory scattering, the phrase appears in two closely related but distinct senses. One is the high-energy limit \(|t|\ll s\) of massless amplitudes, where the infrared operator factorizes and Reggeization emerges from the dipole formula [1201.2841]. Another is the “beyond Sudakov” regime \(s\gg m^2\) in which one also keeps the first power-suppressed term in \(m^2/s\), i.e. the \(O(\rho)\) or \(\sim m/s\) contribution to the amplitude, and studies its own tower of double logarithms [1709.01092], [1412.0671], [1809.04950].

In small-\(x\) QCD, the same broad idea takes the form of a boost expansion. In the strict high-energy limit one works at eikonal order, but the quasi-high-energy description keeps the next-to-eikonal components \(A^+_{(0)}\) and \(A_i{}_{(0)}\), allowing Low’s soft theorem at small \(x\) to be recovered from matrix elements of quasi-classical fields and then extended toward saturation [2402.17568].

Outside perturbative scattering, the phrase is used more broadly. For geometric differential operators, the strict high-energy limit studies exact eigensections, whereas a quasi-high-energy limit is the same construction carried out on normalized quasi-modes whose energies go to infinity and which satisfy a dynamical localization condition [1109.2104]. In collision-induced false-vacuum decay, it denotes the regime \(E\gg E_{rt}(N)\), where the suppression exponent has already reached its minimum and becomes independent of further increases in \(E\) [1503.06339]. In nonlinear trident, it denotes the route to \(\chi\gg 1\) through large \(b_0\) at fixed or moderate \(a_0\), rather than through the locally-constant-field hierarchy \(a_0\gg 1\) [2007.08492].

## 2. Small-\(x\) QCD: quasi-classical fields and Low’s soft theorem

In the strict high-energy, or eikonal, limit of the small-\(x\) problem, soft gluons are approximately described by quasi-classical gluon fields. Under a large boost by a factor \(\xi\to 0\), the field scales as
\[
A^+\sim \xi^{-1},\qquad A^-\sim \xi,\qquad A_i\sim \xi^0.
\]
At leading order only \(A^+_{(-1)}\equiv A^+\) appears, and in Lorenz gauge \(\partial\!\cdot\!A=0\) the classical Yang–Mills equation reduces to
\[
-\nabla_\perp^2 A^+_{(-1)}(x^-,x_\perp)=J^+_{(-1)}(x^-,x_\perp),
\qquad
A^+_{(-1)}=-\frac{1}{\nabla_\perp^2}J^+_{(-1)}.
\]
Equivalently, in light-cone gauge \(A^-=A_i=0\), the transverse components can be written as the pure gauge
\[
A_i(x_\perp)=\frac{i}{g}\,U(x_\perp)\partial_iU^\dagger(x_\perp),
\]
with \(F_{ij}=0\), confirming the pure-gauge nature on the transverse plane [2402.17568].

The next-to-eikonal expansion organizes the field as
\[
A^+(x)=\xi^{-1}A^+_{(-1)}(x)+A^+_{(0)}(x)+O(\xi),
\]
\[
A^-(x)=\xi\,A^-_{(1)}(x)+\dots,\qquad
A_i(x)=A_{i(0)}(x)+\xi A_{i(1)}(x)+\dots .
\]
The first subeikonal components are therefore \(A^+_{(0)}(x)\) and \(A_i{}_{(0)}(x)\). In Lorenz gauge, and assuming \(A^+_{(-1)}\sim O(1/g)\), \(A_{(0)}\sim O(1)\), the order-\(\xi^0\) equations are
\[
-\nabla_\perp^2 A^+_{(0)}-2\partial_i D_- A_i{}_{(0)}=J^+_{(0)},
\qquad
-\nabla_\perp^2 A_i{}_{(0)}=J_i{}_{(0)},
\]
with \(D_-=\partial_-+ig[A^+_{(-1)},\,\cdot\,]\). The immediate solutions are
\[
A^+_{(0)}=-\frac{1}{\nabla_\perp^2}\Bigl[2\partial_iD_-A_i{}_{(0)}+J^+_{(0)}\Bigr],
\qquad
A_i{}_{(0)}=-\frac{1}{\nabla_\perp^2}J_i{}_{(0)}.
\]
The resulting next-to-eikonal field strengths include
\[
F^{i+}
=
\Bigl(\delta^{ij}-2\frac{\partial^i\partial^j}{\nabla_\perp^2}\Bigr)\,
D_- \Bigl(\frac{1}{\nabla_\perp^2}J^j\Bigr)
+
\frac{\partial^i}{\nabla_\perp^2}J^+,
\]
\[
F^{ij}=-\frac{\partial^iJ^j-\partial^jJ^i}{\nabla_\perp^2},
\qquad
F^{-+}=\frac{\partial_j}{\nabla_\perp^2}J^j,
\qquad
F^{-i}=0
\]
[2402.17568].

In the dilute regime \(A\sim O(g)\), the off-diagonal matrix element of \(F^{+i}\) between one-gluon “in” and “out” states reproduces the small-\(x\) expansion of Low’s soft factor. The standard soft factor expanded at \(k^+/p^+\equiv z\ll 1\), while allowing \(|k_\perp|\sim |p_\perp|\), is
\[
S(k)
=
2\,\delta_{\sigma\sigma'}
\Bigl[
-\frac{k^i}{k_\perp^2}
+
z\Bigl(\delta^{ij}-2\frac{k^ik^j}{k_\perp^2}\Bigr)\frac{p^j}{k_\perp^2}
+
z\,\sigma\,\frac{i\epsilon^{ij}k^j}{k_\perp^2}
+
z\,\frac{k^i}{k_\perp^2}(p^+\partial_{p^+})
\Bigr]\epsilon^*_{\lambda,i}
+O(z^2),
\]
and the field-theoretic matrix element yields exactly the same bracket structure [2402.17568].

The extension to the dense regime introduces saturation-scale dependence. In the McLerran–Venugopalan model,
\[
\langle \rho^a(x_\perp)\rho^b(y_\perp)\rangle
=
\delta^{ab}\mu_A^2\,\delta^2(x_\perp-y_\perp),
\qquad
\mu_A^2\sim Q_s^2L^-,
\]
with \(A^+_{(-1)}\sim -\rho/\nabla_\perp^2\). Since the subeikonal currents also carry \(\mu_A^2\to Q_s^2\) dependence, every instance of \(J_{(-1)}\), \(J_{(0)}\), and hence \(F^{+i}\), inherits a characteristic scale \(Q_s\). The proposed dense generalization is a “saturated” Low’s theorem,
\[
S(k)=\int d^2x_\perp\,e^{-ik_\perp\cdot x_\perp}\;
\Bigl\langle
\mathrm{Tr}\bigl[U(x_\perp)T^a\partial^iU^\dagger(x_\perp)\bigr]
\Bigr\rangle_{Q_s},
\]
which manifestly depends on the saturation scale \(Q_s\) [2402.17568].

## 3. Infrared singularities, Reggeization, and the \(|t|\ll s\) limit

For massless gauge amplitudes, the high-energy limit \(|t|\ll s\) can be derived from the universal properties of infrared singularities. In color-tensor notation an amplitude factorizes as
\[
{\cal M}=Z\,{\cal H},
\]
where \(Z\) collects all soft and collinear poles in dimensional regularization and satisfies
\[
\mu\frac{d}{d\mu}Z=-Z\,\Gamma .
\]
The dipole ansatz takes the anomalous dimension to be
\[
\Gamma^{\rm dip}
=
\frac14\,\widehat\gamma_K(\alpha_s)\sum_{i<j}
\ln\!\Bigl(\frac{-s_{ij}}{\mu^2}\Bigr)\,
\mathbf T_i\!\cdot\!\mathbf T_j
-
\sum_{i=1}^n \gamma_{J_i}(\alpha_s),
\]
with \(s_{ij}=2p_i\!\cdot\!p_j\), \(\mathbf T_i\) the color generator acting on parton \(i\), \(\widehat\gamma_K\) the universal cusp anomalous dimension stripped of its quadratic Casimir, and \(\gamma_{J_i}\) the collinear anomalous dimension of leg \(i\) [1201.2841].

For four-point scattering, one defines
\[
\mathbf T_s=\mathbf T_1+\mathbf T_2,
\qquad
\mathbf T_t=\mathbf T_1+\mathbf T_3,
\]
and the infrared operator factorizes, to leading power in \(|t|/s\), as
\[
Z(p_i,\mu,\epsilon)=Z_1(t,\mu,\epsilon)\,Z_{\rm rest}(s/t,\mu,\epsilon),
\]
where \(Z_1\) is color-diagonal and contains no large \(\ln(s/|t|)\). The nontrivial high-energy dependence is
\[
Z_{\rm rest}(s/t,\mu,\epsilon)
=
\exp\!\Biggl\{
K(\alpha_s(\mu),\epsilon)
\Bigl[
\ln\!\frac{s}{-t}\,\mathbf T_t^2+i\pi\,\mathbf T_s^2
\Bigr]
\Biggr\}.
\]
This is the all-orders eikonal soft factor, valid to leading power in \(t/s\) but to all logarithmic orders in \(\ln(s/|t|)\) [1201.2841].

At leading logarithmic accuracy, if the Born hard part is dominated by a definite \(t\)-channel color exchange so that \(\mathbf T_t^2\to C_t\,\mathbf 1\), the infrared-divergent part reggeizes:
\[
{\cal M}_{\rm div}^{\rm LL}
\propto
\exp\!\Bigl\{C_t\,K(\alpha_s,\epsilon)\,\ln\!\frac{s}{-t}\Bigr\}
=
\Bigl(\frac{s}{-t}\Bigr)^{C_tK(\alpha_s,\epsilon)}.
\]
The corresponding Regge trajectory is
\[
\alpha(t)=C_t\,K(\alpha_s,\epsilon)
=
\frac{\alpha_s}{4\pi}\frac{C_t\,\widehat\gamma_K^{(1)}}{\epsilon}
+
{\cal O}(\alpha_s^2,\epsilon^0),
\]
recovering the standard Regge-factorized form for the infrared-singular part of any four-point amplitude [1201.2841].

The same formalism also identifies the failure of simple Reggeization at higher logarithmic orders. At NLL the imaginary part acquires a non-diagonal color term proportional to \(K^2\ln(s/(-t))[\mathbf T_t^2,\mathbf T_s^2]\). At NNLL the real part also fails to reggeize, with leading Reggeization-breaking operator
\[
{\cal E}(s/t,\alpha_s,\epsilon)
=
-\frac{\pi^2}{3}K^3(\alpha_s,\epsilon)\,
\ln\!\frac{s}{-t}\,
\bigl[\mathbf T_s^2,[\mathbf T_t^2,\mathbf T_s^2]\bigr].
\]
In multi-Regge kinematics this structure generalizes to commuting \(t\)-channel color operators \(\mathbf T_{t_k}\), and
\[
Z_{\rm MRK}
=
\exp\!\Bigl\{
\sum_{k=1}^{n-1}
\bigl[
K\,\Delta y_k\,\mathbf T_{t_k}^2+i\pi K\,\mathbf T_s^2
\bigr]
\Bigr\},
\]
so each sub-channel reggeizes independently at LL, with the same pattern of mixing at subleading logs [1201.2841].

The same paper also uses Regge behavior to constrain possible corrections beyond the dipole formula. A prototypical quadrupole correction depending on conformal cross ratios would generate “super-leading” terms \(\alpha_s^3\ln^4(s/|t|)\) in conflict with known Regge behavior. This suggests that any admissible beyond-dipole correction must be a very special linear combination whose super-leading pieces cancel exactly [1201.2841].

## 4. Beyond Sudakov approximation: power-suppressed logarithms

In QED and QCD, one major use of the quasi-high-energy limit is the retention of the first power-suppressed term in the high-energy expansion. For the QED Dirac form factor \(F_1\) with \(\rho=m^2/s\ll 1\) and \(L=\ln(s/m^2)\gg 1\),
\[
F_1(s,m^2)=S_\lambda\sum_{n=0}^\infty \rho^n F_1^{(n)}(\alpha,L),
\qquad
F_1^{(0)}=e^{-x},
\qquad
x=\frac{\alpha}{4\pi}L^2.
\]
Thus the leading term reproduces the Sudakov factor
\[
S(s,m^2)=\exp\!\Bigl[-\frac{\alpha}{4\pi}\ln^2\frac{s}{m^2}\Bigr].
\]
The first genuinely new effect beyond Sudakov appears in \(F_1^{(1)}\) at two loops, and evaluating the four-fold double-logarithmic integral gives
\[
F_1^{(1)}\big|_{\rm 2-loop}
=
-\frac{x^2}{3}+{\cal O}(x^3),
\]
hence
\[
\Delta F_1
=
F_1^{(1)}\rho
=
-\frac13
\Bigl(\frac{\alpha}{4\pi}\Bigr)^2
L^4\,\frac{m^2}{s}
+\cdots .
\]
The mechanism is not soft-photon Sudakov exchange but exchange of a soft electron–positron pair between two eikonal electron lines [1412.0671].

The non-abelian generalization keeps the same logic but changes the organizing principle. The standard leading-power Sudakov limit keeps only \(O(1)\) terms as \(m^2/s\to 0\), with soft/collinear gauge-boson exchange exponentiating into the usual Sudakov factor. The quasi-high-energy, or beyond-Sudakov, limit keeps the first power-suppressed term and studies its double-logarithmic tower, which is induced by soft quark exchange and depends on eikonal color-charge nonconservation [1709.01092]. In the double-logarithmic approximation the factorization formula for a single soft-quark exchange is
\[
\mathcal M(s,m^2)
=
Z_{\rm in}\,Z_{\rm out}\;g(\pm z)\;\mathcal M^{(0)}(s,m^2),
\]
with
\[
z=(C_A-C_F)\,\frac{\alpha_s}{4\pi}L^2,
\qquad
g(z)=2\int_0^1 d\xi\int_0^{1-\xi} d\eta\;
\exp[2z\xi\eta].
\]
Equivalently,
\[
g(z)= {}_2F_2\!\Bigl(1,1;\tfrac32,2;\tfrac z2\Bigr)
=
2\sum_{n=0}^\infty \frac{n!}{(2n+2)!}(2z)^n .
\]
Its asymptotics differ sharply between \(z>0\) and \(z<0\), so the same topology leads to enhancement in QCD but suppression in QED [1709.01092].

For mass-suppressed quark form factors, the detailed resummation introduces a second universal function \(f(z)\). One has
\[
F_1 = Z_q^2\,[1+\rho\,F_1^{(1)}+O(\rho^2)],
\qquad
F_1^{(1)}=\frac{C_F(C_A-2C_F)}{6}\,x^2\,f(-z),
\]
with \(x=(\alpha_s/4\pi)L^2\). The same \(f(z)\) appears in the scalar, axial, and pseudoscalar form factors up to simple color prefactors:
\[
F_S^{(1)}=-\frac{C_FT_F}{3}\,x^2\,f(-z),
\qquad
F_A^{(1)}=F_1^{(1)}\big|_{f\to -f},
\qquad
F_P^{(1)}=F_1^{(1)}\big|_{f\to +f}.
\]
The source material describes this as “magical” universality [1709.01092].

A related detailed formulation writes the factorization of a leading fermion-mass-suppressed amplitude as
\[
A(s,t,m)=H(s,t)\,J(p_1,m)\,J(p_2,m)\,S_{\rm ns}(m,Q)+O(m^2/Q^2),
\]
or, for the quark-current example,
\[
A={\cal G}(Q^2,m)=Z_q^2\,g(-z)\,{\cal G}^{(0)}(Q^2),
\qquad
Z_q^2=\exp\!\Bigl[-C_F\,\frac{\alpha_s}{4\pi}\ln^2\frac{m^2}{Q^2}\Bigr].
\]
The same framework was applied to the bottom-quark contribution to Higgs production in gluon fusion, for which
\[
\delta\sigma_{gg\to H}
=
-3\,\rho\,\ln^2\rho\,g(z)\times \sigma^{(0)}_{gg\to H},
\]
with \(z=(C_F-C_A)(\alpha_s/4\pi)\ln^2\rho<0\) [1809.04950].

## 5. Geometric operators, quasi-modes, and frame-flow invariance

For geometric differential operators, the expression “quasi-high-energy limit” refers to a generalization of the usual spectral high-energy limit from exact eigenfunctions to quasi-modes. Let \(P\) be a positive, self-adjoint differential or pseudodifferential operator of order \(m>0\) on sections of a Hermitian vector bundle \(E\to M\) over a compact Riemannian manifold. The strict high-energy limit studies expectation values
\[
\langle \psi_\lambda, A\,\psi_\lambda\rangle,
\qquad
A\in \Psi^0(M;E),
\]
for exact eigensections \(P\psi_\lambda=\lambda^m\psi_\lambda\). A quasi-high-energy limit is the same construction carried out on normalized quasi-modes \(\psi_n\) whose energies satisfy
\[
\langle \psi_n,P\,\psi_n\rangle\to\infty
\]
and
\[
\|(P-\lambda_n^m)\psi_n\|_{L^2}=o(\lambda_n^m).
\]
The weak-\(*\) limit points of the states
\[
w_n(A)=\langle\psi_n,A\,\psi_n\rangle
\]
descend to states on \(C(S^*M;\pi^*\!\End(E))\) [1109.2104].

The principal symbol controls the limit. If \(P\) has full symbol \(P\sim \sum_{|\alpha|\le m} a_\alpha(x)D_x^\alpha\), then
\[
\sigma_P(x,\xi)=\sum_{|\alpha|=m} a_\alpha(x)\,\xi^\alpha,
\]
and for Laplace-type operators \(\sigma_P(x,\xi)=|\xi|^m\,\mathrm{Id}_{E_x}\). The microlocal lift
\[
W_\lambda(a)=\langle \psi_\lambda,\Op(a)\psi_\lambda\rangle
\]
converges, along subsequences, to a matrix-valued measure. In one formulation,
\[
w(\Op(a))
=
\int_{S^*M}\mathrm{tr}\bigl(a(x,\xi)\rho(x,\xi)\bigr)\,d\Vol_{S^*M}(x,\xi),
\]
where \(\rho(x,\xi)\ge 0\) and \(\mathrm{tr}\,\rho=1\) [1109.2104].

The classical invariance of the limit states follows from Egorov’s theorem. For the quantum propagator \(U(t)=e^{-itP^{1/m}}\),
\[
\Op(a)(t)=U(-t)\,\Op(a)\,U(t)\in \Psi^0(M;E),
\qquad
\sigma_{\Op(a)(t)}(x,\xi)=\beta_t(a(x,\xi)),
\]
so any limit measure is invariant under the classical flow \(\beta_t\). In the scalar-Laplace case \(\beta_t\) is the geodesic flow \(G^t\). For nontrivial bundles, and in particular for the Dirac and Maxwell cases, the relevant classical dynamics is the frame flow induced by parallel transport on the orthonormal frame bundle. Strohmaier and collaborators emphasize that it is the dynamical properties of this frame flow that determine the behavior of eigensections at large energies [1109.2104].

The same framework yields bundle-valued quantum-ergodicity statements. If there exist mutually orthogonal zeroth-order pseudodifferential projections \(\Pi_j\) commuting with \(P\) and giving an ergodic decomposition of the tracial state under \(\beta_t\), then almost all eigenfunctions in each spectral subspace \(\Pi_jL^2\) equidistribute toward the corresponding ergodic state. The cited examples are the scalar Laplacian, the Dirac operator with chirality projections \(\Pi_\pm=\tfrac12(1\pm \sign(D))\), the Hodge-Laplacian on \(1\)-forms, and the Dolbeault Laplacian on Kähler manifolds [1109.2104].

## 6. Nonperturbative tunneling and quasi-Regge integrability

In collision-induced false-vacuum decay, the quasi-high-energy limit is defined by
\[
E\gg E_{rt}(N),
\]
where the inclusive cross section for induced decay by \(N\) incoming quanta at total energy \(E\) is
\[
\sigma_N(E)\approx \exp[-F_N(E)/g^2].
\]
The suppression exponent decreases with energy, reaches its minimum
\[
F_{\min}(N)=F_N(E_{rt}(N))
\]
at the threshold \(E=E_{rt}(N)\) where \(T=0\), and for all \(E>E_{rt}(N)\) satisfies
\[
F_N(E)=F_{\min}(N)\quad\text{(constant).}
\]
The semiclassical solutions that compute \(E_{rt}(N)\) and \(F_{\min}(N)\) are the real-time instantons: complex solutions of the classical field equations evolving entirely on the real-time axis, with no Euclidean leg remaining. Around these solutions one develops a perturbative expansion whose terms stay bounded as \(E\to\infty\). Physically, above threshold the incoming particles emit many soft quanta of total energy \(E_{rt}\), while the excess energy \(E-E_{rt}\) remains in the colliding particles and does not reduce the suppression exponent further [1503.06339].

A different quasi-high-energy usage occurs in the quasi-Regge limit of multi-Regge amplitudes and effective reggeized-gluon dynamics. Here one refines the standard Regge limit by taking
\[
s\to\infty,\qquad |t_i|\sim O(1),\qquad \alpha_s\ln s\simeq O(1),
\qquad
t_i/s^{1-\delta_i}\sim O(1)\quad (0<\delta_i\ll 1),
\]
or equivalently by keeping rapidity gaps \(y_k\) of order \(O(1)\) or \(O(\ln s^\delta)\) rather than strictly of order \(\ln s\). In the planar eight-point amplitude of \(\mathcal N=4\) SYM, and similarly in adjoint BKP dynamics, the reggeon Green’s function obeys
\[
\partial_Y f = H_{\rm eff}\,f,
\qquad
H_{\rm eff}=\sum_{i=1}^{n-1}H_{i,i+1},
\]
with nearest-neighbor kernels built from the one-loop Regge trajectory \(\omega(\mathbf p)\) and Lipatov’s real-emission kernel \(\xi\). The resulting system maps to an integrable open spin chain, diagonalizable through an open-chain Baxter equation for the function \(Q(u)\) [1801.04872].

The same work introduces graph-theoretic complexity into the high-energy effective field theory. For a graph \(\Gamma\), the Kirchhoff complexity \(C(\Gamma)\) is the number of distinct spanning trees, given by any cofactor of the Laplacian \(L_\Gamma\). For a chain with \(n\) rungs, the maximal complexity grows as
\[
t(n)=\frac{(2+\sqrt3)^n-(2-\sqrt3)^n}{2\sqrt3}
\simeq \frac{1}{2\sqrt3}(2+\sqrt3)^n.
\]
Numerically, the paper reports “complexity democracy”: for fixed rung number \(n\), the average Monte Carlo weight of all graphs of a given complexity becomes essentially independent of \(C\) for large \(C\). This suggests that the quasi-Regge regime organizes exponentially many diagrams into a comparatively rigid integrable structure [1801.04872].

## 7. Strong-field QED: large-\(b_0\) trident and the failure of LCF

In nonlinear trident pair production in a plane-wave background, the relevant parameters are
\[
a_0=\frac{eE}{m\omega},
\qquad
b_0=\frac{kp}{m^2},
\qquad
\chi=a_0b_0.
\]
Two routes to \(\chi\gtrsim 1\) are distinguished. The usual locally-constant-field regime takes \(a_0\gg 1\) first and then \(\chi\to\infty\). The high-energy, or “quasi-high-energy,” regime instead takes
\[
b_0\gg 1
\]
at fixed or moderate \(a_0\), so that \(\chi=a_0b_0\gg 1\) while \(a_0\) need not be large. In this regime the one-step, or direct, part of trident eventually dominates, in contrast to the two-step cascade term that dominates when \(a_0\gg 1\) [2007.08492].

At leading order in large \(b_0\), but exact in \(a_0\), the total direct probability is
\[
\mathbb P
=
\mathbb P^{11}_{\rm dir}+\mathbb P^{22\to 1}_{\rm dir}
=
\frac{\alpha^2}{6\pi^2}
\int\!\frac{d\phi_3\,d\phi_4}{\theta\,\Theta}
\Bigl[
\frac83-\frac{19}{6}\Delta^2
+
(1-\Delta^2)\Bigl(\ln\!\bigl[-\frac{\theta\,\Theta}{(2b_0)^2}\bigr]+2\gamma_E\Bigr)
\Bigr],
\]
with \(\Delta=a(\phi_4)-a(\phi_3)\), \(\theta=\phi_4-\phi_3\), and \(\Theta\) the squared effective-mass integral. If one also takes \(a_0\gg 1\), this splits into the local large-\(a_0\) terms
\[
\mathbb P^{11}_{\rm dir}
=
\frac{\alpha^2}{3\sqrt3\,\pi}\int \frac{d\sigma}{b_0}\,\chi(\sigma),
\qquad
\mathbb P^{22\to1}_{\rm dir}
=
\frac{13\alpha^2}{6\sqrt3\,\pi}
\int \frac{d\sigma}{b_0}\,\chi(\sigma)\,
\ln\!\frac{\chi(\sigma)}{2\sqrt3\,e^{-\gamma_E-64/39}} .
\]
The same paper shows that replacing the incoming electron by a static Coulomb center yields Bethe–Heitler pair production in a plane wave, and that the large-\(b_0\) limit reproduces exactly the high-energy trident result. The Weizsäcker–Williams approximation reproduces the \(\ln b_0\) part of the exact high-energy approximation [2007.08492].

A central point is the failure of LCF in the true high-\(\chi\) regime reached through energy rather than intensity. From the large-\(\chi\) behavior of the one-step term in LCF, the dominant formation region scales as \(\tilde\theta\propto \chi^{2/3}\), with \(\tilde\theta=a_0\theta\). LCF remains valid only if
\[
\frac{\chi^{2/3}}{a_0}\ll 1
\qquad\Longleftrightarrow\qquad
a_0\gg \chi^{2/3}.
\]
Once \(\chi\gg a_0^{3/2}\), exactly in the regime \(b_0\gg a_0\), this condition fails; the LCF result ceases to be accurate and is replaced by the high-energy approximation governed by \(\ln b_0\) and nontrivial \(a_0\)-dependence [2007.08492].

The same analysis identifies a genuinely nonlocal next-to-leading correction in the \(1/a_0\) expansion. Writing
\[
F(\theta)=
\frac{[a(\phi_4)-a(\phi_3)]^2}
{\theta^2\bigl(\langle a^2\rangle-\langle a\rangle^2\bigr)},
\qquad
F_{\rm LCF}=\frac{\dot a^2\theta^2}{12},
\]
one obtains
\[
\Delta\mathbb P
=
-\frac{\alpha^2}{6\pi^2}\,\mathcal T
\int_0^\infty d\theta\,[F(\theta)-F_{\rm LCF}(\theta)].
\]
Its main support comes from \(\theta\sim O(1)\), not \(\theta\sim 1/a_0\), and it can be numerically important even at quite large \(a_0\). To bridge small-\(\chi\) and large-\(\chi\) regions, the paper employs Borel–Padé, Borel–conformal–Padé, and hypergeometric/Meijer-\(G\) resummations, obtaining high precision up to very large \(\chi\) [2007.08492].

Across these examples, the quasi-high-energy limit denotes not the abandonment of asymptotics, but their refinement. Whether the retained structure is next-to-eikonal field content, infrared color mixing, mass-suppressed logarithms, quasi-mode localization, threshold-saturating instantons, quasi-Regge substructure, or large-\(b_0\) nonlocality, the common theme is that the strict leading high-energy approximation is insufficient for the phenomenon of interest.

Source: https://www.emergentmind.com/topics/quasi-high-energy-limit