Deeply inelastic scattering as an inverse problem

summary

The gist

Deeply inelastic scattering as an inverse problem proposes a novel formalism for high-energy inclusive QCD processes based on analyticity of a scattering amplitude…

In short

This work proposes a novel method for analyzing proton structure functions using deep inelastic scattering as an inverse problem based purely on analyticity, bypassing conventional factorization theorems. By applying dispersion relations to double deeply virtual Compton scattering amplitudes, it solves for nonperturbative structure functions like F2 and FL directly from perturbative inputs, successfully reproducing experimental data in the intermediate Bjorken region.

Key concepts

Deeply Inelastic Scattering as an Inverse Problem
This approach treats QCD sum rules as an inverse problem. It uses known perturbative results (from quark side) to solve for unknown nonperturbative quantities (like proton structure functions F2 and FL) on the hadron side by relying only on the analyticity of physical observables, rather than standard factorization.
Dispersion Relations for DDVCS
The analysis is built upon constructing dispersion relations for the double deeply virtual Compton scattering amplitude defined on a complex baryon mass plane. This mathematical tool links the perturbative quark-level descriptions to the nonperturbative hadron structure through dispersive integrals, allowing for the extraction of structure functions.
Operator Product Expansion (OPE) Inputs
The OPE provides inputs from heavy quark DIS processes, which follow short-distance factorization. These perturbative results serve as constraints within the dispersion relation framework to help solve for the nonperturbative structure functions in the low-mass region.

Terminology used across episodes

This episode discusses

The paper

Deeply inelastic scattering as an inverse problem · Read on arXiv

Hsiang-nan Li

Institute of Physics, Academia Sinica

Transcript

Introduction to the show: ident: High Energy Physics: Phenomenology. The day in particle phenomenology: models, signatures and constraints.

Kai: We're looking at the paper: "Deeply inelastic scattering as an inverse problem".

Mira: Here's the gist of the paper. Deeply inelastic scattering as an inverse problem proposes a novel formalism for high-energy inclusive QCD processes based on analyticity of a scattering amplitude, which is conceptually distinct from the conventional factorization theorem The paper illustrates how to analyze proton deeply inelastic scattering by means of the dispersion relations for double deeply virtual Compton scattering defined on a complex baryon mass plane Inputting the perturbative results for the heavy baryon scattering amplitudes, it solves for the proton structure functions F2 and FL in the nonperturbative low-mass region directly from the dispersion relations, i.e., as an inverse problem Conceptual Framework The core idea is to manage QCD sum rules as an inverse problem by treating a spectral density on the hadron side of a dispersion relation, including resonance and continuum contributions, as an unknown solved with standard operator-product expansion (OPE) inputs on the quark side This approach relies solely on analyticity of physical observables to derive nonperturbative quantities like proton structure functions F2 and FL The formalism is applied to high-energy inclusive QCD processes, using the double deeply virtual Compton scattering (DDVCS) amplitude on a complex baryon mass plane as the basis for the dispersion relation Input and Constraints The analysis incorporates prior information, including the scaling of structure functions in photon virtuality Q2 and quark counting rules for the threshold region with Bjorken variable xB → 1 The only nonperturbative parameter at leading-power precision in 1/Q is the proton mass, with which agreement is demonstrated for F2 and FL in intermediate xB = 0.18-0.65 across a wide range of Q2 = 102-104 GeV2 Parton distribution functions can be extracted as a byproduct by treating the associated factorization formulas as another inverse problem Methodological Steps The derivation involves constructing the dispersion relation for a DDVCS amplitude based on its analyticity, linking hadron- and quark-level descriptions via dispersive integrals The imaginary part of the forward DDVCS amplitude is related to the DIS cross section, while OPE inputs for heavy quark DIS are used to solve the dispersion relation as an inverse problem The analytical scattering amplitudes are expressed as contour integrals, leading to a Fredholm integral equation (14) which is solved using the inverse matrix method Phenomenological Outcomes The analysis yields predictions for the proton structure function F2 and FL that agree well with experimental data in the considered xB-Q2 domain The results for the pion structure function Fπ2 also demonstrate straightforward applications to DIS, explaining its data with a pion mass as the nonperturbative parameter The framework allows for a theoretical access to the large xB region which is hard to explore experimentally The work proposes a novel approach that differs dramatically from the conventional factorization theorem by relying only on analyticity of physical observables The framework is not constructed on the light cone, so it does not demand the introduction of PDFs It has been corroborated that our results for the proton structure functions F2 and FL agree with the data in the range xB = 0.18-0.65 and Q2 = 102-104 GeV2 The extension to the pion structure function also accommodates the data for xπ = 0.12 and Q2 ≈ 103 GeV2 This work offers not only a completely distinct viewpoint but a deeper insight into inclusive QCD processes Page 4 --- The imaginary part of the forward DDVCS amplitude, which appears in the dispersive integral, is related to DIS cross section The OPE for the heavy quark DIS follows the short-distance factorization, so our approach differs conceptually from the factorization theorem established on the light cone The inputs remain reliable even in the threshold region with the Bjorken variable xB → 1, because the effective Bjorken variable xN = mpxB/mN for the heavy baryon DIS (see the detail in Sec.

Kai: II), mp being the proton mass, is always below unity There is thus sufficient phase space for arXiv:2611.15665v1

hep-ph: 11 Nov 2026 Page 8 --- The radius R in Eq.

Mira: (8) can be pushed toward the infinity, when the dispersion relation is formulated in terms of the subtracted unknown The variable changes mN − m0 ≈ mQ − m0 = xΛ and m − m0 = yΛ recast Eq.

Kai: (12) into Z ∞ 0 dy ∆ρ(y) x − y = Z ∞ 0 dy ImΠpert(yΛ + m0) exp(−yγ) x − y The above equation aims at solving for a structure function, which depends on the dimensionless variable y = (m − m0)/Λ Page 9 --- The maximal integer N will be prescribed later, and the value of the index α depends on the behavior of ∆ρ(y) near the threshold y → 0 We have checked that subtractions with powers higher than 3, like γ = 4, alter our predictions by only few percent (less than 5%) Therefore, we designate α = 3/2 according to the aforementioned quark counting rules ρ(y) ∝ y cubed near the threshold Page 9 --- The series ai would grow eventually with i owing to the amplification by the huge elements of M−1, though the series bi decreases The last minimal magnitude of ai close to the high end of N occurs at i = 16, which sets the integer N = 16 for the polynomial expansion It is known that priors help determine solutions to an inverse problem The series ai corresponding to Λ = Λm is selected to construct the physical solution of ρ(m), which gives rise to the central value of our predictions for F2 A ±20% variation around mb is postulated to allow room for theoretical uncertainties The solutions corresponding to Λ near Λm are all similar, describing the structure function for an arbitrary baryon mass from the nonperturbative region with m ≈ mp to a heavy-mass region Page 9 --- We read off the proton structure function F2 defined at m = mp = 0.938 GeV

43: , F2(xB = 0.25, Q2 = 102 GeV 2) = 0.27+0.05 −-0.07 where the central (upper bound, lower bound) corresponds to Λ = 200 GeV and N = 16 Page 9 --- F2(xB = 0.40, Q2 = 103 GeV 2) = 0.14+0.03 −-0.04 where the central value (upper bound, lower bound) corresponds to Λ = 320 GeV and N = 15 Page 9 --- F2(xB = 0.40, Q2 = 103 GeV 2) = 0.14+0.03 −-0.

Title and authors: Kai: Let's talk about the title and authors of "Deeply inelastic scattering as an inverse problem." The title itself suggests that they're treating this whole field not just as a forward calculation but as a way to work backward from measurements.

Mira: I think the core idea is that instead of trying to build up the structure functions from first principles using factorization theorems, they are setting up a system where the known perturbative parts give you constraints on the unknown nonperturbative stuff >

Lev: And it's interesting because they are solving for proton structure functions F2 and FL directly from these dispersion relations rather than needing explicit parton distribution functions to start with >

Kai: So, what does this actually mean in practice for how we think about proton structure? Is it a shortcut around the traditional methods?

Mira: It's a conceptual shift. The paper says it's not constructed on the light cone, which means it doesn't demand the introduction of PDFs in the same way conventional approaches do >

Lev: That’s important because when you try to run this on real hardware, you have to figure out how those constraints translate into something measurable without needing a full PDF input from scratch >

The paper's summary: Kai: So, summarizing the paper's summary, they are using prior information like the scaling of structure functions with photon virtuality Q2 and quark counting rules for the threshold region with Bjorken variable xB to guide their solution.

Mira: They incorporate these priors to solve for F2 and FL in a specific intermediate range of xB, between zero point one eight and zero point six five, across a wide range of photon virtuality Q2 from one hundred two gigaparsecs squared up to one hundred four gigaparsecs squared >

Lev: That's the data region where they are demonstrating agreement for F2 and FL, which is a good benchmark for seeing if this works >

Kai: So, it’s not just theory; they have these specific numerical results showing how well their model matches experimental data in that xB Q2 domain >

Mira: Yes. And they even extend this to the pion structure function F pi two showing straightforward applications by using the pion mass as the nonperturbative parameter to explain its data at x pi equal zero point one two and Q2 around one hundred three gigaparsecs squared > <ref:2611.15665#pg1>

Lev: If we think about running this on a real computer, extracting those parton distribution functions becomes another inverse problem that you'd have to solve using similar techniques, like with lattice QCD correlation functions >

The paper's improvements: Kai: The paper points out several ways they improve the approach. One major thing is treating QCD sum rules as an inverse problem by solving for a spectral density on the hadron side, using standard operator product expansion inputs on the quark side >

Mira: That’s a big improvement because it gives you better control over theoretical uncertainties compared to just doing traditional QCD sum rules where you have to guess some things >

Lev: And they use the inverse matrix method for solving those ill-posed integral equations, which is a mature mathematical tool for tackling that kind of problem >

Kai: They also address the challenge of controlling theoretical errors sourced from ill-posedness by using these established mathematical tools to attack an inverse problem directly >

Mira: They even use prior information, like the scaling of structure functions in photon virtuality Q2 and quark counting rules for the threshold region with Bjorken variable xB to one, to settle the behavior of their solution at low and high masses >

Lev: And they found that subtractions with powers higher than three, like gamma equal four, only alter their predictions by less than five percent >

Conclusion: Kai: So wrapping up on the paper "Deeply inelastic scattering as an inverse problem," the main implication is that this formalism offers a completely distinct viewpoint to inclusive QCD processes by relying only on analyticity of physical observables.

Mira: They showed that their results for the proton structure functions F2 and FL agree well with experimental data in the range xB equals zero point one eight to zero point six five and Q2 equals one hundred two gigaparsecs squared to one hundred four gigaparsecs squared >

Lev: From an engineering side, it means this approach is viable because it doesn't fundamentally demand PDFs, which simplifies the input requirements for running simulations on real hardware, though you still have to solve those byproduct inverse problems >

Kai: So, to summarize what we've heard about "Deeply inelastic scattering as an inverse problem," it’s a novel way to access inclusive QCD processes by turning them into an inverse problem using analyticity and solving for structure functions directly from the known perturbative inputs.

Mira: It gives us a deeper insight into how these high-energy processes work, even though the paper itself notes that controlling theoretical errors sourced from ill-posedness is still a challenge >

Lev: And to finish up on this piece, they found that when you read off the proton structure function F2 defined at mass equal point four zero GeV and Q2 equal one hundred three gigaparsecs squared, it gives us zero point one four plus zero point zero three minus negative zero point zero four > <ref:2611.15665#pg1>

Kai: That’s a concrete result to bring back to the listener. We'll take a quick break and come back with something else from arXiv.

More episodes

← Home