Reconstruction of a dark energy model for the Dirac-Born-Infeld scalar field with the Hubble and DESI data via Gaussian process

Technical Summary: Reconstruction of a Dark Energy Model for the Dirac-Born-Infeld Scalar Field with Hubble and DESI Data via Gaussian Process

Problem Statement
The standard $\Lambda$CDM cosmological model, while phenomenologically successful, faces theoretical challenges regarding the origin of the cosmological constant ($\Lambda$) and the "Hubble tension" between early-universe (Planck) and late-universe (SH0ES) measurements. Furthermore, identifying the specific nature of Dark Energy (DE) often relies on assuming a specific parametric form for the DE equation of state or scalar field potential, which introduces model-dependent biases. The Dirac-Born-Infeld (DBI) scalar field, arising from open string theory and D-brane dynamics, offers a non-canonical alternative to standard quintessence that can explain late-time acceleration. However, determining the viability of specific DBI potentials (e.g., exponential, power-law) without assuming their functional forms a priori remains a challenge. This study addresses the need for a model-independent reconstruction of the DBI scalar field potential and its parameters using the latest observational data, including the recently released DESI dataset.

Methodology
The authors employ a two-stage, non-parametric approach using Gaussian Processes (GP) to reconstruct cosmological quantities directly from data, followed by statistical fitting to specific theoretical models.

1. Data Utilization: The reconstruction utilizes a combined dataset of 58 Hubble data points (32 Cosmic Chronometers [CC] and 26 Radial Baryon Acoustic Oscillations [BAO]) and 5 newly released data points from the Dark Energy Spectroscopic Instrument (DESI). The CC data provides model-independent $H(z)$ measurements, while BAO and DESI data offer high-precision constraints on the expansion history.
2. Gaussian Process Reconstruction: A GP algorithm with a squared-exponential covariance kernel is used to reconstruct the Hubble parameter $H(z)$ and its derivatives ($H'(z)$, $H''(z)$) without assuming a specific cosmological model. This allows for the derivation of the DE equation of state ($w_\phi$), the DE density parameter ($\Omega_\phi$), and the scalar field potential $V(\phi)$ and kinetic term ($\dot{\phi}^2$) directly from the reconstructed expansion history.
3. Hubble Constant Estimation: Monte Carlo realizations of the GP-reconstructed $H(z)$ are generated to propagate observational uncertainties. These samples are flattened to produce a posterior distribution for the Hubble constant ($H_0$), which is then fitted with a Gaussian profile.
4. Potential Reconstruction and Model Selection: Using the reconstructed $H(z)$ and the DBI field equations, the authors numerically solve for the scalar field $\phi(z)$ and the potential $V(\phi)$. This reconstructed dataset ($V$ vs. $\phi$) is then subjected to a chi-square ($\chi^2$) curve fitting procedure against four standard theoretical potentials: Exponential, Power-law, Free Field (Quadratic), and Higgs-like.
5. Parameter Constraints: Markov Chain Monte Carlo (MCMC) analysis is performed to constrain the free parameters of the best-fitting potentials, providing confidence intervals and examining parameter degeneracies.

Key Contributions

• First Application of DESI to DBI Fields: The paper claims to be the first to utilize the latest DESI dataset to reconstruct and constrain the coefficients of a DBI scalar field potential.
• Model-Independent Framework: By using GP, the study avoids assuming a specific functional form for the potential or the equation of state during the initial reconstruction phase, relying solely on data and GP priors.
• Comprehensive Potential Analysis: The work systematically evaluates four distinct scalar field potentials against the reconstructed data, providing a comparative assessment of their viability within the DBI framework.
• Integration of String Theory and Observations: The study bridges high-energy string theory concepts (DBI fields) with low-energy cosmological observations, offering a pathway to test string-theoretic predictions against data.

Results

• Hubble Constant: The model-independent reconstruction yields $H_0 = 69.53 \pm 2.68 \text{ km s}^{-1} \text{ Mpc}^{-1}$. This value lies between the Planck ($67.4 \pm 0.5$) and SH0ES ($73.04 \pm 1.04$) measurements, offering an unbiased intermediate estimate that is statistically consistent with both at the $\sim 0.8\sigma$ and $\sim 1.3\sigma$ levels, respectively.
• Equation of State: The reconstructed DE equation of state is $w_{\phi,0} = -0.9521 \pm 0.0109$, and the density parameter is $\Omega_{\phi,0} = 0.7220 \pm 0.3923$. These are consistent with the standard $\Lambda$CDM model. The analysis notes that $w_\phi$ deviates from $-1$ at high redshifts ($z > 2.1$), entering a region where the kinetic term becomes negative (unphysical for the proposed model), limiting the viable reconstruction range to $0 < z \le 2.1$.
• Potential Fitting:
  • Power-law Potential ($V_0 + V_1\phi^\lambda$): Exhibits the best statistical agreement with the reconstructed data, with a reduced chi-square ($\chi^2_{red}$) of $0.9888$, described as an "excellent fit."
  • Exponential Potential ($V_0 + V_1e^{-\lambda\phi}$): Shows a "good fit" with $\chi^2_{red} = 0.7694$.
  • Higgs-like Potential ($V_0 + \frac{m^2}{2}\phi^2 + \frac{\lambda}{4}\phi^4$): Shows a "reasonably good fit" with $\chi^2_{red} = 0.8120$.
  • Free Field Potential ($V_0 + \frac{m^2}{2}\phi^2$): Yields a "good fit" with $\chi^2_{red} = 1.2150$.
• Parameter Constraints: MCMC analysis reveals correlations between parameters (e.g., negative correlations between $V_0$ and other parameters in most models), which the authors attribute to the fixed nature of the current cosmological constant $\Lambda$. All models require a non-zero $V_0$ term (mimicking a zero-point energy or $\Lambda$) to match observational data.

Significance and Claims
The authors assert that this work provides a complete and systematic, model-independent analysis of the DBI scalar field. The primary significance lies in demonstrating that the DBI framework is consistent with current observational data (CC, BAO, DESI) without requiring ad hoc assumptions about the potential's shape. The derived $H_0$ value offers a balanced perspective on the Hubble tension, neither exacerbating nor dismissing it.

The study highlights that while the DBI field was originally motivated by high-energy string theory (D-brane inflation), it serves as a viable phenomenological candidate for late-time acceleration, potentially unifying dark energy and dark matter descriptions. The authors suggest that their non-parametric reconstruction of the potential provides a robust method to test "Swampland conjectures" (such as the de Sitter conjecture) which place bounds on scalar field potentials in quantum gravity. They conclude that the Power-law potential currently offers the closest match to the reconstructed data, though all four tested models are compatible within statistical uncertainties. The paper positions this methodology as a tool for future extensions into generalized K-essence fields and modified gravity theories (e.g., $f(T)$ and $f(Q)$ gravity).