betaFeedback

WorksPhysics and Astronomy

Lorentz Boosts and Governing Equations of the Quantum Geometric Tensor in 4D Energy-Momentum Space

Abstract

In solids, the quantum geometric tensor (QGT), Qμν=gμν−i2ΩμνQ_{μν}=g_{μν}-\frac{i}{2} Ω_{μν}, combines the quantum metric and Berry curvature. Its imaginary part governs anomalous velocity and quantized Hall transport, while its real part determines the quantum weight and geometric contributions to flat-band superfluid stiffness and dielectric response. Here, we extend the QGT to four-dimensional (4D) energy-momentum space (ω/c,kω/c,k), treating frequency and momentum as independent coordinates. Invariance of the real-spacetime plane-wave phase fixes their Lorentz transformation in energy-momentum space. Lorentz boosts of static Dirac and Weyl systems then generate frequency metrics, quadrupolar frequency-momentum shear metrics, and electric-like Berry-Maxwell fields. We identify five Lorentz-invariant contractions and derive the governing field equations and flow conservation laws for both the quantum metric and Berry curvature from static Weyl geometry and Lorentz covariance. We further connect Gaussian spectral amplitudes of the QGT to localized Gaussian events in 4D spacetime. Finally, we discuss experimental realizations in spacetime crystals or time-varying media and clarify how this unified quantum geometry of a gravitational-like metric and electromagnetic-like gauge fields connects geometric self-consistency with physical event dynamics.

Einstein and Schrödinger sought to unify gravity and electromagnetism through Hermitian and nonsymmetric extensions of spacetime geometry [1, 2]. Their construction combines the symmetric spacetime metric gμνg_{\mu\nu} and the antisymmetric electromagnetic field tensor FμνF_{\mu\nu} in a complex Hermitian tensor, g^μν=gμν+iαFμν\hat{g}_{\mu\nu}=g_{\mu\nu}+i\alpha F_{\mu\nu}, and motivates this work. In condensed-matter and photonic systems, quantum mechanics provides an intrinsic Hermitian tensor in momentum space. We obtain the gauge-invariant quantum geometric tensor (QGT) by projecting the derivatives of a quantum state ∣u⟩|u\rangle onto its orthogonal complement P^⊥=1−∣u⟩⟨u∣\hat{P}_{\perp}=1-|u\rangle\langle u| [3, 4]:

Qμν=⟨∂μu∣(1−∣u⟩⟨u∣)∣∂νu⟩=gμν−i2Ωμν,(1)Q_{\mu\nu}=\langle\partial_{\mu}u|(1-|u\rangle\langle u|)|\partial_{\nu}u\rangle=g_{\mu\nu}-\frac{i}{2}\Omega_{\mu\nu}, \tag*{(1)}

where the symmetric real part gμν=Re⁡Qμνg_{\mu\nu}=\operatorname{Re}Q_{\mu\nu} defines the quantum metric and the antisymmetric imaginary part Ωμν=−2Im⁡Qμν\Omega_{\mu\nu}=-2\operatorname{Im}Q_{\mu\nu} gives the Berry curvature. The same state ∣u⟩|u\rangle determines both the real and imaginary parts of QμνQ_{\mu\nu}.

In static bands, the two QGT components govern distinct physical observables. For a gapped two-dimensional (2D) Dirac cone, integrating the Berry curvature Ωxy\Omega_{xy} over momentum space yields a half-integer Berry flux C=±1/2C=\pm1/2. In a lattice system, the global occupied bands and their ultraviolet (UV) completion determine the measurable Hall conductance [5, 6]. Meanwhile, the real part (quantum metric gijg_{ij}) measures the quantum distance between neighboring Bloch states and sets the quantum geometric weight [7, 8, 9, 10, 11]. This geometric weight contributes to the superfluid stiffness in flat-band superconductors [7] and to dielectric and optical responses in insulators [9, 10, 11]. Around a gapped Dirac point, integrating the metric trace Tr⁡(g)\operatorname{Tr}(g) gives a logarithmic UV divergence, ∫∣k∣<ΛUVd2k2πTr⁡g≃18ln⁡(ΛUV2/m2)+18\int_{|\mathbf{k}|<\Lambda_{\mathrm{UV}}}\frac{d^{2}k}{2\pi}\operatorname{Tr}g\simeq\frac{1}{8}\ln(\Lambda_{\mathrm{UV}}^{2}/m^{2})+\frac{1}{8}, while the metric and curvature obey the local bound Tr⁡(g)≥∣Ωxy∣\operatorname{Tr}(g)\geq|\Omega_{xy}| [8].

Spacetime crystals, time-varying media, and spatiotemporally modulated metamaterials provide settings for periodic spatial and temporal modulation [12]. Many dynamical descriptions formulate quantum geometry in a mixed parameter space of three-dimensional (3D) momentum k\mathbf{k} and one-dimensional (1D) external time tt. Such parameter–time descriptions connect the QGT to postquench wavepacket spreading [13]. They also organize the Berry connections of time-dependent states into Berry–Maxwell equations [14]. Energy-momentum Berry curvature provides a complementary description [15, 16]. Building on the Berry–Maxwell formulation in Ref. [16], we extend this description to the full QGT, including its symmetric metric sector. Here, we consider normalized spectral states ∣u(k,ω)⟩|u(\mathbf{k},\omega)\rangle with frequency and momentum as independent coordinates in four-dimensional (4D) energy-momentum space kμ=(ω/c,k)k^{\mu}=(\omega/c,\mathbf{k}) (Figure 1). Because the plane-wave phase ϕ=k⋅r−ωt=ημνkμxν\phi=\mathbf{k}\cdot\mathbf{r}-\omega t=\eta_{\mu\nu}k^{\mu}x^{\nu} is a Lorentz scalar under the Minkowski metric ημν=diag⁡(−1,1,1,1)\eta_{\mu\nu}=\operatorname{diag}(-1,1,1,1), the 4D energy-momentum coordinates (ω/c,k)(\omega/c,\mathbf{k}) inherit the same Lorentz transformation as real spacetime (ct,r)(ct,\mathbf{r}) [16]. For our Dirac and Weyl models, we evaluate the rest-frame eigenstates at Lorentz-boosted momenta to make their frequency dependence explicit (see End Matter).

Figure 1. From static Weyl geometry to the 4D QGT field equations. (a) A matter-wave wavepacket in a spacetime crystal defines a Floquet-Bloch state and its QGT. (b) Static Weyl geometry in the rest frame S′. (c) A phase-preserving boost compresses fixed-frequency momentum contours by a factor of 1/γ1/\gamma. (d) We derive the metric and Berry–Maxwell field equations together with their flow conservation laws. Blue and orange denote the metric and curvature sectors, respectively. Here s=Tr⁡g−gωωs=\operatorname{Tr}g-g_{\omega\omega}, ϕ=s/2\phi=\sqrt{s/2}, and c=1c=1. The second-order metric equation applies away from the monopole core.

In this Letter, we establish the Lorentz transformation laws, relativistic invariants, and governing field equations of the 4D QGT defined by spectral states ∣u(k,ω)⟩|u(\mathbf{k},\omega)\rangle in 4D energy-momentum space (Figure 1). First, we identify five Lorentz-invariant contractions (I1I_1–I5I_5) of the QGT and apply Lorentz boosts to 1D, 2D, and 3D Dirac and Weyl systems, showing how static spatial metrics and magnetic-like Berry curvatures generate pure frequency metrics gωωg_{\omega\omega}, quadrupolar frequency-momentum shear metrics gωk\mathbf{g}_{\omega\mathbf{k}}, and electric-like Berry-Maxwell fields γ\boldsymbol{\gamma}. Second, we use static Weyl monopole geometry and Lorentz covariance to derive the governing field equations and flow conservation laws for both the quantum metric and Berry curvature. Third, we Fourier-transform an independent Gaussian scalar spectral function in (ω/c,k)(\omega/c,\mathbf{k}) space to obtain a localized Gaussian event in real 4D spacetime (ct,r)(ct,\mathbf{r}). We distinguish this localization example from physical event wavepackets. Related work demonstrates topological spacetime localization and studies nonlinear event formation and vortices [17, 18, 19, 20].

4D QGT and Lorentz invariants. As shown in Figure 1(a), we describe a spatiotemporally modulated matter wave by normalized Floquet-Bloch states ∣u(k,ω)⟩|u(\mathbf{k},\omega)\rangle in 4D energy-momentum space kμ=(ω,kx,ky,kz)k^\mu=(\omega,k_x,k_y,k_z) (setting the light speed c=1c=1). For the models below, we evaluate the rest-frame Dirac or Weyl eigenstates at Lorentz-boosted momenta to obtain these states (see End Matter). Using the derivatives ∂μ≡∂/∂kμ=(∂ω,∇k)\partial_\mu\equiv\partial/\partial k^\mu=(\partial_\omega,\nabla_{\mathbf{k}}) in Eq. (1), we obtain the 4×44\times4 Hermitian QGT,

Qμν(k,ω)=gμν(k,ω)−i2Ωμν(k,ω),Q_{\mu\nu}(\mathbf{k},\omega)=g_{\mu\nu}(\mathbf{k},\omega)-\frac{i}{2}\Omega_{\mu\nu}(\mathbf{k},\omega),

with two coupled geometric sectors [Figure 1(a)]. Its antisymmetric imaginary part Ωμν=∂μAν−∂νAμ\Omega_{\mu\nu}=\partial_\mu\mathcal{A}_\nu-\partial_\nu\mathcal{A}_\mu [with 4D Berry connection Aμ=i⟨u∣∂μu⟩=(−χ,A)\mathcal{A}_\mu=i\langle u|\partial_\mu u\rangle=(-\chi,\mathbf{A})] forms the 6-component Berry-Maxwell field tensor. We decompose it into the magnetic-like Berry curvature Ωi=12ϵijkΩjk=(∇k×A)i\Omega_i=\frac{1}{2}\epsilon_{ijk}\Omega_{jk}=(\nabla_{\mathbf{k}}\times\mathbf{A})_i and the electric-like Berry curvature γi=Ωiω=−Ωωi=−∂ωAi−∂iχ\gamma_i=\Omega_{i\omega}=-\Omega_{\omega i}=-\partial_\omega A_i-\partial_i\chi [16]. The symmetric real part gμν=Re⁡⟨∂μu∣(1−∣u⟩⟨u∣)∣∂νu⟩g_{\mu\nu}=\operatorname{Re}\langle\partial_\mu u|(1-|u\rangle\langle u|)|\partial_\nu u\rangle defines the 10-component 4D quantum metric, comprising the 3D momentum metric gijg_{ij}, the pure frequency metric gωωg_{\omega\omega}, and the frequency-momentum cross metric gω=(gωx,gωy,gωz)\mathbf{g}_\omega=(g_{\omega x},g_{\omega y},g_{\omega z}). Physically, the frequency metric measures how rapidly the normalized state u(k,ω)u(\mathbf{k},\omega) changes with frequency at fixed momentum. The cross metric gives the mixed contribution to quantum-state distance when frequency and momentum vary together. We use Tr⁡g≡gxx+gyy+gzz\operatorname{Tr}g\equiv g_{xx}+g_{yy}+g_{zz} for the trace of the momentum block and Tr⁡(g2)\operatorname{Tr}(g^2) for the trace of its square. We raise four-dimensional indices with ημν\eta^{\mu\nu} and choose ϵ0123=+1\epsilon^{0123}=+1.

Because the plane-wave phase ϕ=k⋅r−ωt=ημνkμxν\phi=\mathbf{k}\cdot\mathbf{r}-\omega t=\eta_{\mu\nu}k^\mu x^\nu is a Lorentz scalar, a Lorentz boost in real spacetime with velocity β=v/c\boldsymbol{\beta}=\mathbf{v}/c induces the same coordinate transformation Λ(β)\Lambda(\boldsymbol{\beta}) in 4D energy-momentum space [Figure 1(c)]. We obtain the boosted field components from the invariance of the quantum distance ds2=1−∣⟨u(k)∣u(k+dk)⟩∣2=gμνdkμdkνds^2=1-|\langle u(\mathbf{k})|u(\mathbf{k}+d\mathbf{k})\rangle|^2=g_{\mu\nu}dk^\mu dk^\nu and the Berry curvature two-form Ω=12Ωμνdkμ∧dkν\Omega=\frac{1}{2}\Omega_{\mu\nu}dk^\mu\wedge dk^\nu. Consider a system that is static in its rest frame S′S' (∂ω′∣u⟩=0\partial_{\omega'}|u\rangle=0, so that gω′ω′′=0g'_{\omega'\omega'}=0, gω′′=0\mathbf{g}'_{\omega'}=0, and γ′=0\boldsymbol{\gamma}'=0). Under a Lorentz boost along the xx-axis to the laboratory frame SS, the energy-momentum coordinates and differentials transform as kx′=γ(kx−βω)k'_x=\gamma(k_x-\beta\omega), dkx′=γ(dkx−βdω)dk'_x=\gamma(dk_x-\beta d\omega), dky′=dkydk'_y=dk_y, and dkz′=dkzdk'_z=dk_z, where γ=1/1−β2\gamma=1/\sqrt{1-\beta^2} is the Lorentz factor. We substitute dkx′dk'_x into the rest-frame line element dsFS2=gij′dki′ dkj′ds_{\mathrm{FS}}^2=g'_{ij}dk'_i\,dk'_j and two-form Ω=12ϵijkΩk′ dki′∧dkj′\Omega=\frac{1}{2}\epsilon_{ijk}\Omega'_k\,dk'_i\wedge dk'_j, collect terms in dωd\omega and dkd\mathbf{k}, and obtain the boosted 4D quantum metric and Berry-Maxwell fields in frame SS:

gωω=β2γ2gxx′,gωx=−βγ2gxx′,gωa=−βγgxa′(2)g_{\omega\omega}=\beta^2\gamma^2g'_{xx},\qquad g_{\omega x}=-\beta\gamma^2g'_{xx},\qquad g_{\omega a}=-\beta\gamma g'_{xa} \tag*{(2)}
gxx=γ2gxx′,gxa=γgxa′,gab=gab′(a,b∈y,z),(3)g_{xx}=\gamma^2g'_{xx},\qquad g_{xa}=\gamma g'_{xa},\qquad g_{ab}=g'_{ab}\quad(a,b\in y,z), \tag*{(3)}
Ωx=Ωx′,Ω⊥=γΩ⊥′,Υ=−β×Ω=−βγ(0,−Ωz′,Ωy′).(4)\Omega_x=\Omega'_x,\qquad\boldsymbol{\Omega}_{\perp}=\gamma\boldsymbol{\Omega}'_{\perp},\qquad\boldsymbol{\Upsilon}=-\boldsymbol{\beta}\times\boldsymbol{\Omega}=-\beta\gamma(0,-\Omega'_z,\Omega'_y). \tag*{(4)}

Equations (2)–(4) show different Lorentz scaling at corresponding spectral points: transverse Berry curvatures scale linearly as γ\gamma, while the longitudinal metric gxxg_{xx}, pure frequency metric gωωg_{\omega\omega}, and cross metric gωxg_{\omega x} acquire quadratic γ2\gamma^2 factors.

Despite the frame dependence and γ2\gamma^2 enhancement of individual components, we contract QμνQ_{\mu\nu} with the Minkowski metric ημν=diag⁡(−1,1,1,1)\eta^{\mu\nu}=\operatorname{diag}(-1,1,1,1) to obtain five Lorentz-invariant quantities:

I1≡ημνgμν=Tr⁡(g)−gωω=Tr⁡(g′),I2≡12ΩμνΩμν=Ω2−Υ2=Ω′2,I3≡−18εμνρσΩμνΩρσ=Ω⋅Υ,(5)\begin{aligned} I_1&\equiv\eta^{\mu\nu}g_{\mu\nu}=\operatorname{Tr}(\mathbf{g})-g_{\omega\omega}=\operatorname{Tr}(\mathbf{g}'),\\ I_2&\equiv\frac{1}{2}\Omega_{\mu\nu}\Omega^{\mu\nu}=\boldsymbol{\Omega}^2-\boldsymbol{\Upsilon}^2=\boldsymbol{\Omega}'^2,\\ I_3&\equiv-\frac{1}{8}\varepsilon^{\mu\nu\rho\sigma}\Omega_{\mu\nu}\Omega_{\rho\sigma}=\boldsymbol{\Omega}\cdot\boldsymbol{\Upsilon}, \tag*{(5)} \end{aligned}
I4≡gμνgμν=gωω2−2gω2+Tr⁡(g2),I5≡QμνQμν=I4−12I2.(6)\begin{aligned} I_4&\equiv g_{\mu\nu}g^{\mu\nu}=g_{\omega\omega}^2-2g_{\omega}^2+\operatorname{Tr}(\mathbf{g}^2),\\ I_5&\equiv Q_{\mu\nu}Q^{\mu\nu}=I_4-\frac{1}{2}I_2. \tag*{(6)} \end{aligned}

These invariants have direct physical interpretations: (i) I1I_1 is the Minkowski metric trace, showing how the γ2\gamma^2 factor in gxxg_{xx} combines with the −β2γ2-\beta^2\gamma^2 contribution of the emergent frequency metric gωωg_{\omega\omega} to preserve an invariant 4D quantum geometric spread; (ii) I2I_2 is the Berry-Maxwell field Lagrangian and remains nonnegative for a subluminal boost of a static band structure (∣β∣<1|\boldsymbol{\beta}|<1); (iii) I3=Ω⋅ΥI_3=\boldsymbol{\Omega}\cdot\boldsymbol{\Upsilon} is a pseudoscalar density dual to the chiral anomaly (E⋅B\mathbf{E}\cdot\mathbf{B}). For a boost from a static frame (Υ′=0\boldsymbol{\Upsilon}'=0), Υ=−β×Ω\boldsymbol{\Upsilon}=-\boldsymbol{\beta}\times\boldsymbol{\Omega} is perpendicular to Ω\boldsymbol{\Omega}, so I3=0I_3=0 identically; a nonzero value therefore rules out a purely static rest-frame origin and signals intrinsic 4D spatiotemporal coupling. Finally, Eq. (6) relates the quadratic contractions I4I_4 and I5I_5 of the metric and 4D QGT. The five quantities are therefore not algebraically independent. For a static band structure and its Lorentz boosts, QGT positivity relates the first two invariants through the quantum geometric bound:

I1=Tr⁡(g)−gωω≥I2=Ω2−Υ2,(7)I_1=\operatorname{Tr}(\mathbf{g})-g_{\omega\omega}\ge\sqrt{I_2}=\sqrt{\boldsymbol{\Omega}^2-\boldsymbol{\Upsilon}^2}, \tag*{(7)}

where equality holds for the isotropic Weyl model (along with I4=12I12=12I2I_4=\frac{1}{2}I_1^2=\frac{1}{2}I_2 and I5=0I_5=0) but need not hold for other two-band systems. Saturation requires equal nonzero transverse metric eigenvalues; the massive 1D and 2D Dirac examples below generally give a strict inequality.

Lorentz boost of Dirac and Weyl systems. We illustrate Eqs. (2)–(7) with two-band Dirac and Weyl Hamiltonians H(k′)=d(k′)⋅σH(\mathbf{k}')=\mathbf{d}(\mathbf{k}')\cdot\boldsymbol{\sigma} in 1D, 2D, and 3D (Figure 2). For the upper band, we define the Bloch unit vector n^(k′)=d/∣d∣\hat{\mathbf{n}}(\mathbf{k}')=\mathbf{d}/\lvert\mathbf{d}\rvert and write the rest-frame metric and curvature as gij′=14(∂i′n^)⋅(∂j′n^)g'_{ij}=\frac{1}{4}(\partial'_i\hat{\mathbf{n}})\cdot(\partial'_j\hat{\mathbf{n}}) and Ωij′=−12n^⋅(∂i′n^×∂j′n^)\Omega'_{ij}=-\frac{1}{2}\hat{\mathbf{n}}\cdot(\partial'_i\hat{\mathbf{n}}\times\partial'_j\hat{\mathbf{n}}). In 1D, we consider the massive Dirac model H1D(kx′)=kx′σx+mσzH_{\mathrm{1D}}(k'_x)=k'_x\sigma_x+m\sigma_z with energy E′=kx′2+m2E'=\sqrt{k_x'^2+m^2} and unit vector n^=(kx′,0,m)/E′\hat{\mathbf{n}}=(k'_x,0,m)/E'. In the rest frame S′S', the Berry curvature vanishes (Ω′=0\Omega'=0), leaving only a bell-shaped spatial quantum metric gxx′(kx′)=m2/(4E′4)g'_{xx}(k'_x)=m^2/(4E'^4) [dashed gray curve in Figure 2(a)]. We apply the Lorentz boost kx′=γ(kx−βω)k'_x=\gamma(k_x-\beta\omega) through Eq. (2) and obtain a 2×22\times2 frequency-momentum metric in (ω,kx)(\omega,k_x) space:

Lorentz-boosted Dirac and Weyl geometry

Figure 2. Lorentz-boosted Dirac and Weyl geometry for β=0.75\beta= 0.75, ω=0\omega= 0, c=1c = 1, and m=0.8m = 0.8 in the massive models. (a) Rest-frame and boosted 1D metric components. (b) Negative Γy\Gamma_y in the 2D Dirac model; the dashed circle and solid ellipse show Lorentz compression. (c) Quadrupolar mixed metric gωyg_{\omega y}. (d) Weyl frequency metric gωωg_{\omega\omega} (green) and electric-like vortex Γ\boldsymbol{\Gamma} (orange) on the offset slice qx=a=0.35q_x = a = 0.35. The dashed ring r⊥=∣a∣r_\perp= |a| marks the metric maximum; the offset slice excludes the Weyl node.

(gωωgωxgxωgxx)1D=γ2m24[γ2(kx−βω)2+m2]2(β2−β−β1)(8)\left(\begin{array}{cc} g_{\omega\omega} & g_{\omega x} \\ g_{x\omega} & g_{xx} \end{array}\right)_{\mathrm{1D}}=\frac{\gamma^2m^2}{4[\gamma^2(k_x-\beta\omega)^2+m^2]^2}\left(\begin{array}{cc}\beta^2&-\beta\\-\beta&1\end{array}\right) \tag*{(8)}

Figure 2(a) shows that, for β=0.75\beta=0.75 (γ2≈2.29\gamma^2\approx2.29), a Lorentz boost sharpens the longitudinal metric gxxg_{xx} and raises its peak by a factor of γ2\gamma^2 (solid blue curve), even though this 1D model has no Berry curvature. The boost also generates a positive pure frequency metric gωω=β2γ2gxx′g_{\omega\omega}=\beta^2\gamma^2g'_{xx} (solid green curve) and a negative cross metric gωx=−βγ2gxx′g_{\omega x}=-\beta\gamma^2g'_{xx} (purple curve), while preserving the Lorentz-invariant difference I1=gxx−gωω=m2/(4E′4)I_1=g_{xx}-g_{\omega\omega}=m^2/(4E'^4). These components quantify the state's sensitivity to frequency and joint frequency-momentum variations in the geometric description of pulse deformation.

In 2D, the massive Dirac model H2D(kx′,ky′)=kx′σx+ky′σy+mσzH_{\mathrm{2D}}(k'_x,k'_y)=k'_x\sigma_x+k'_y\sigma_y+m\sigma_z with E′=kx′2+ky′2+m2E'=\sqrt{k_x'^2+k_y'^2+m^2} hosts both geometric sectors. In the static frame, the out-of-plane Berry curvature Ωz′=−m/(2E′3)\Omega'_z=-m/(2E'^3) integrates to the half-integer cone flux C=−12sgn⁡(m)C=-\frac{1}{2}\operatorname{sgn}(m). The static metric gij′=E′2δij−ki′kj′4E′4g'_{ij}=\frac{E'^2\delta_{ij}-k'_ik'_j}{4E'^4} has trace Tr⁡(g′)=(k′2+2m2)/(4E′4)\operatorname{Tr}(g')=(k'^2+2m^2)/(4E'^4). Integrating Tr⁡(g′)\operatorname{Tr}(g') up to a cutoff ΛUV\Lambda_{\mathrm{UV}} gives the quantum geometric weight WΛ=∫∣k′∣<Λd2k′2πTr⁡g′≃18ln⁡(Λ2/m2)+18W_\Lambda=\int_{\lvert\mathbf{k}'\rvert<\Lambda}\frac{d^2k'}{2\pi}\operatorname{Tr}g'\simeq\frac{1}{8}\ln(\Lambda^2/m^2)+\frac{1}{8}. This weight exhibits a logarithmic UV divergence and obeys WΛ≥∣CΛ∣W_\Lambda\geq\lvert C_\Lambda\rvert, where CΛC_\Lambda is the Berry flux through the same cutoff disk (see End Matter). A Lorentz boost along xx, with E′=γ2(kx−βω)2+ky2+m2E'=\sqrt{\gamma^2(k_x-\beta\omega)^2+k_y^2+m^2}, compresses the isotropic rest-frame contours (dashed gray circle) by a factor of 1/γ1/\gamma along kxk_x [solid ellipse in Figure 2(b)]. According to Eqs. (2)–(4), the boost induces a unidirectional in-plane electric-like Berry curvature [vertical streamlines and color density in Figure 2(b)] and a 3×33\times3 metric in (ω,kx,ky)(\omega,k_x,k_y) space:

γy=−βγm2E′3,gωω=β2γ2(ky2+m2)4E′4,gωy=βγ2(kx−βω)ky4E′4.(9)\gamma_y=-\frac{\beta\gamma m}{2E'^3},\qquad g_{\omega\omega}=\frac{\beta^2\gamma^2(k_y^2+m^2)}{4E'^4},\qquad g_{\omega y}=\frac{\beta\gamma^2(k_x-\beta\omega)k_y}{4E'^4}. \tag*{(9)}

The frequency-momentum shear metric gωy(kx,ky)g_{\omega y}(k_x,k_y) has a dxyd_{xy}-type quadrupolar pattern with alternating positive (+) and negative (−) lobes across the four quadrants of the (kx,ky)(k_x,k_y) plane [Figure 2(c)].

In 3D, the massless Weyl Hamiltonian H3D(k′)=k′⋅σH_{\mathrm{3D}}(\mathbf{k}')=\mathbf{k}'\cdot\boldsymbol{\sigma} hosts a topological monopole at k′=0\mathbf{k}'=0.

In the rest frame, its radial magnetic-like Berry curvature is Ω′=−k′/(2k′3)\boldsymbol{\Omega}'=-\mathbf{k}'/(2k'^{3}) and its transverse spherical metric is gij′=(δijk′2−ki′kj′)/(4k′4)g'_{ij}=(\delta_{ij}k'^{2}-k'_{i}k'_{j})/(4k'^{4}). For a Lorentz boost along xx, with k′=γ2(kx−βω)2+ky2+kz2k'=\sqrt{\gamma^{2}(k_{x}-\beta\omega)^{2}+k_{y}^{2}+k_{z}^{2}}, Eqs. (2)–(4) give the 4D Weyl geometric fields:

Ω=−γ(kx−βω,ky,kz)2k′3,Υ=−βγ(0,kz,−ky)2k′3,(10)\boldsymbol{\Omega}=-\frac{\gamma(k_{x}-\beta\omega,k_{y},k_{z})}{2k'^{3}},\qquad\boldsymbol{\Upsilon}=-\frac{\beta\gamma(0,k_{z},-k_{y})}{2k'^{3}}, \tag*{(10)}
gωω=β2γ2(ky2+kz2)4k′4,gω=βγ24k′4(−(ky2+kz2),(kx−βω)ky,(kx−βω)kz).(11)g_{\omega\omega}=\frac{\beta^{2}\gamma^{2}(k_{y}^{2}+k_{z}^{2})}{4k'^{4}},\qquad\mathbf{g}_{\omega}=\frac{\beta\gamma^{2}}{4k'^{4}}\left(-(k_{y}^{2}+k_{z}^{2}),(k_{x}-\beta\omega)k_{y},(k_{x}-\beta\omega)k_{z}\right). \tag*{(11)}

On the transverse (ky,kz)(k_{y},k_{z}) slice offset from the moving Weyl node (kx′=γ(kx−βω)=a=0.35)(k'_{x}=\gamma(k_{x}-\beta\omega)=a=0.35), the boosted monopole generates a circulating vortex of electric-like Berry curvature Υ\boldsymbol{\Upsilon} (orange streamlines) over a ring-shaped frequency metric gωωg_{\omega\omega} (green color map) [Figure 2(d)]. The metric vanishes at the center of this offset slice (ky=kz=0)(k_{y}=k_{z}=0) and peaks at ky2+kz2=∣a∣\sqrt{k_{y}^{2}+k_{z}^{2}}=|a|. A slice through the node intersects the monopole singularity.

Governing equations and spectral flow conservation. We derive the governing differential equations of both Ωμν\Omega_{\mu\nu} and gμνg_{\mu\nu} [Figure 1(d)] from the static geometry of the 3D Weyl monopole and Lorentz covariance. In the rest frame S′S', the Weyl monopole field Ω′=−k′/(2k′3)\boldsymbol{\Omega}'=-\mathbf{k}'/(2k'^{3}) is curl-free (∇k′×Ω′=0)(\nabla_{\mathbf{k}'}\times\boldsymbol{\Omega}'=0) and satisfies Gauss’s law ∇k′⋅Ω′=ρW′=−2πδ3(k′)\nabla_{\mathbf{k}'}\cdot\boldsymbol{\Omega}'=\rho'_{W}=-2\pi\delta^{3}(\mathbf{k}'). For the rest-frame quantum metric gij′=(δijk′2−ki′kj′)/(4k′4)g'_{ij}=(\delta_{ij}k'^{2}-k'_{i}k'_{j})/(4k'^{4}), the trace g′=1/(2k′2)=∣Ω′∣g'=1/(2k'^{2})=|\boldsymbol{\Omega}'| defines a root-trace scalar potential φ′=g′/2=1/(2k′)\varphi'=\sqrt{g'/2}=1/(2k') that obeys the Poisson equation ∇k′2φ′=ρW′\nabla_{\mathbf{k}'}^{2}\varphi'=\rho'_{W}. We define the trace-reversed quantum metric gˉμν′≡gμν′−12ημνg′\bar{g}'_{\mu\nu}\equiv g'_{\mu\nu}-\frac{1}{2}\eta_{\mu\nu}g' and differentiate gˉij′=−ki′kj′/(4k′4)\bar{g}'_{ij}=-k'_{i}k'_{j}/(4k'^{4}) to show that its momentum divergence vanishes away from the monopole core: ∂i′gˉij′=−14∂j′(ki′kj′/k′4)=0\partial'_{i}\bar{g}'_{ij}=-\frac{1}{4}\partial'_{j}(k'_{i}k'_{j}/k'^{4})=0.

We transform these static identities to the laboratory frame SS using ∂kx′=γ(∂kx+β∂ω)\partial_{k'_{x}}=\gamma(\partial_{k_{x}}+\beta\partial_{\omega}), while leaving the transverse derivatives unchanged, to obtain the governing equations. For the imaginary sector, we transform ∇k′×Ω′=0\nabla_{\mathbf{k}'}\times\boldsymbol{\Omega}'=0 and ∇k′⋅Ω′=ρW′\nabla_{\mathbf{k}'}\cdot\boldsymbol{\Omega}'=\rho'_{W} to obtain the Berry-Maxwell equations [16]:

∇k⋅Ω=ρW,∇k⋅Υ=0,∇k×Υ+∂ωΩ=−jW,∇k×Ω=∂ωΥ,(12)\nabla_{\mathbf{k}}\cdot\boldsymbol{\Omega}=\rho_{W},\qquad\nabla_{\mathbf{k}}\cdot\boldsymbol{\Upsilon}=0,\qquad\nabla_{\mathbf{k}}\times\boldsymbol{\Upsilon}+\partial_{\omega}\boldsymbol{\Omega}=-\mathbf{j}_{W},\qquad\nabla_{\mathbf{k}}\times\boldsymbol{\Omega}=\partial_{\omega}\boldsymbol{\Upsilon}, \tag*{(12)}

where ρW=γρW′\rho_{W}=\gamma\rho'_{W} and jW=ρWβ\mathbf{j}_{W}=\rho_{W}\boldsymbol{\beta} satisfy the monopole current continuity equation ∂ωρW+∇k⋅jW=0\partial_{\omega}\rho_{W}+\nabla_{\mathbf{k}}\cdot\mathbf{j}_{W}=0. For the real metric sector, we use the covariance of the trace-reversed divergence ∂μ(gμν−12ημνg)=0\partial^{\mu}\left(g_{\mu\nu}-\frac{1}{2}\eta_{\mu\nu}g\right)=0 to obtain the 4D quantum metric flow conservation laws. Here, g≡I1=Tr g−gωωg\equiv I_{1}=Tr\,\mathbf{g}-g_{\omega\omega}. In components:

∇k⋅gω=12∂∂ω(Tr(g)+gωω),∂igij−12∂j(Tr(g)−gωω)=∂gωj∂ω.(13)\nabla_{\mathbf{k}}\cdot\mathbf{g}_{\omega}=\frac{1}{2}\frac{\partial}{\partial\omega}\left(Tr(\mathbf{g})+g_{\omega\omega}\right),\qquad\partial_{i}g_{ij}-\frac{1}{2}\partial_{j}\left(Tr(\mathbf{g})-g_{\omega\omega}\right)=\frac{\partial g_{\omega j}}{\partial\omega}. \tag*{(13)}

For this isotropic Weyl model, we compare Eqs. (10) and (11) and relate the real metric components to the Berry-Maxwell fields through the exact constitutive relations gωω=Υ2/(2g)g_{\omega\omega}=\boldsymbol{\Upsilon}^{2}/(2g), $\mathbf{g}_{\omega}= −(γ×Ω)/(2g)-\left(\boldsymbol{\gamma} \times\boldsymbol{\Omega}\right)/(2g), and gij=(δijΩ2−ΩiΩj−γiγj)/(2g)g_{ij} = \left(\delta_{ij}\Omega^2-\Omega_i\Omega_j-\gamma_i\gamma_j\right)/(2g), where g=Ω2−γ2g = \sqrt{\Omega^2-\boldsymbol{\gamma}^2} (see End Matter). We substitute gω=−(γ×Ω)/(2g)\mathbf{g}_{\omega} = -(\boldsymbol{\gamma} \times\boldsymbol{\Omega})/(2g) into the first equation of Eq. (13) and obtain a reciprocal Poynting-type conservation law in energy-momentum space. The frequency derivative of the quantum geometric spread Ug=(Tr⁡g+gωω)/2U_g = (\operatorname{Tr}g + g_{\omega\omega})/2 balances the momentum divergence of the spread flux Sg=γ×Ω/(2g)\mathbf{S}_g = \boldsymbol{\gamma} \times\boldsymbol{\Omega}/(2g), so that ∂ωUg+∇k⋅Sg=0\partial_{\omega}U_g + \nabla_{\mathbf{k}} \cdot\mathbf{S}_g = 0. Note that this flow describes quantum-state spread rather than a physical energy current.

4D spacetime Gaussian events. In addition to first-order flow conservation, Lorentz covariance gives second-order field equations in energy-momentum space. Away from the monopole singularity (ρW=0,jW=0)(\rho_W = 0,\mathbf{j}_W = 0), we take the curl of Eq. (12) to obtain a homogeneous wave equation for the Berry-Maxwell field Ωμν=(γ,Ω)\Omega_{\mu\nu} = (\boldsymbol{\gamma},\boldsymbol{\Omega}). Differentiating the Weyl metric gives a nonzero source determined by its spatial variation. With φ=g/2=1/(2k′)\varphi= \sqrt{g/2} = 1/(2k') and □k=∇k2−∂ω2\Box_{\mathbf{k}} = \nabla_{\mathbf{k}}^2 - \partial_{\omega}^2 in units c=1c = 1, the curvature obeys a homogeneous equation, while gradients of the scalar potential source the metric equation:

□kΩμν=0,□kgμν=4∂μφ∂νφ.(14)\Box_{\mathbf{k}}\Omega_{\mu\nu} = 0,\qquad\Box_{\mathbf{k}}g_{\mu\nu} = 4\partial_{\mu}\varphi\partial_{\nu}\varphi. \tag*{(14)}

To illustrate QGT localization in real spacetime, we consider an independent Gaussian scalar spectral function in (ω,k)(\omega,\mathbf{k}) space. Figure 3(a,Figure 3b) show a Gaussian spectral amplitude and its localized Fourier envelope. The cone in Figure 3(a) serves as a dispersion reference; frequency and momentum remain independent spectral coordinates. We write a 4D Gaussian spectral amplitude, centered at (k0,ω0)(\mathbf{k}_0,\omega_0) with momentum width σk\sigma_k, frequency width σω\sigma_\omega, and phase center (r0,t0)(\mathbf{r}_0,t_0), as

Fourier duality of a Gaussian-shaped QGT and its localized spacetime profile

Figure 3. Fourier duality of a Gaussian-shaped QGT and its localized spacetime profile. (a,b) An independent scalar spectral amplitude a(k,ω)a(\mathbf{k},\omega) and its spacetime envelope a~(r,t)\tilde{a}(\mathbf{r},t), with conjugate widths σr=1/σk\sigma_r = 1/\sigma_k and σt=1/σω\sigma_t = 1/\sigma_\omega. The cone in (a) is a dispersion reference. (c) Normalized spectral envelope a=f(δkx−δω)W(δω)a = f(\delta k_x - \delta\omega)W(\delta\omega), where δkx=kx−k0\delta k_x = k_x - k_0 and δω=ω−ω0\delta\omega= \omega- \omega_0; the Gaussian profile and frequency window have amplitude widths σk=0.55\sigma_k = 0.55 and σω=1\sigma_\omega= 1, respectively. (d) Normalized Fourier amplitude ∣a~(x,τ)∣|\tilde{a}(x,\tau)|, with τ=ct\tau= ct and c=1c = 1. The dashed line τ=x\tau= x marks one light-cone branch.

a(k,ω)=a0exp⁡[−∣k−k0∣22σk2−(ω−ω0)22σω2]exp⁡[−ir0⋅(k−k0)+it0(ω−ω0)],(15)a(\mathbf{k},\omega) = a_0\exp\left[-\frac{\lvert\mathbf{k}-\mathbf{k}_0\rvert^2}{2\sigma_k^2}-\frac{(\omega-\omega_0)^2}{2\sigma_\omega^2}\right]\exp\left[-i\mathbf{r}_0\cdot(\mathbf{k}-\mathbf{k}_0)+it_0(\omega-\omega_0)\right], \tag*{(15)}

where a0a_0 is the spectral amplitude prefactor. The QGT characterizes the normalized internal-state family across energy-momentum space. Independently, the scalar spectral function sets the spatial and temporal localization scales of its Fourier profile. The 4D Fourier transform (F4)(\mathcal{F}_4) from energy-momentum space (k,ω)(\mathbf{k},\omega) to real 4D spacetime (r,ct)(\mathbf{r},ct), a~(r,t)=∫d3kdω(2π)4a(k,ω)ei(k⋅r−ωt)\tilde{a}(\mathbf{r},t) = \int\frac{\mathrm{d}^3k\mathrm{d}\omega}{(2\pi)^4}a(\mathbf{k},\omega)e^{i(\mathbf{k}\cdot\mathbf{r}-\omega t)}, gives an explicit Gaussian event in 4D spacetime:

a~(r,t)=a0σk3σω(2π)2exp⁡[−∣r−r0∣22σr2−(t−t0)22σt2]ei(k0⋅r−ω0t),(16)\tilde{a}(\mathbf{r},t) = \frac{a_0\sigma_k^3\sigma_\omega}{(2\pi)^2}\exp\left[-\frac{\lvert\mathbf{r}-\mathbf{r}_0\rvert^2}{2\sigma_r^2}-\frac{(t-t_0)^2}{2\sigma_t^2}\right]e^{i(\mathbf{k}_0\cdot\mathbf{r}-\omega_0t)}, \tag*{(16)}

where the conjugate amplitude widths σr=1/σk\sigma_r = 1/\sigma_k and σt=1/σω\sigma_t = 1/\sigma_\omega set the spatial and temporal localization scales of the envelope around the spacetime coordinate (r0,ct0)(\mathbf{r}_0,ct_0). Equations (15) and (16) give an exact Fourier duality (k↔r,ω↔t)(\mathbf{k} \leftrightarrow\mathbf{r},\omega\leftrightarrow t): a Gaussian spectral amplitude maps one-to-one to a localized Gaussian envelope in real 4D spacetime [Figure 3(a,Figure 3b)]. The windowed 1+1-dimensional example in Figure 3(c,Figure 3d) shows how frequency-momentum correlations stretch and tilt the joint spacetime profile. The dashed line τ=x\tau= x is a light-cone reference. The finite frequency window gives a nonzero differential residual (see End Matter). We use this independent scalar example to illustrate Fourier localization without constructing a source-dependent QGT field solution or independently radiating metric and curvature pulses.

Experimental considerations. Our framework and the duality of Gaussian events suggest three implications for experiments and the physical interpretation of the 4D QGT:

First, the Lorentz boost of the quantum metric gμνg_{\mu\nu} and Berry curvature Ωμν\Omega_{\mu\nu} [Eqs. (2)–(11)] and Figure 2 suggests two experimental routes: physical motion and synthetic spacetime crystals. In the first route, a physical crystal, photonic lattice, or Dirac medium moving at velocity v\mathbf{v} relative to the laboratory frame undergoes a relativistic Doppler shift and Lorentz transformation in (ω/c,k)(\omega/c,\mathbf{k}) space. In the second route, we can design a synthetic spacetime crystal—such as an acoustic or optical metamaterial with a traveling-wave modulation δn(r−vmt)\delta n(\mathbf{r}-\mathbf{v}_{m}t)—to emulate the boosted Dirac or Weyl states, with the modulation phase velocity vm\mathbf{v}_{m} setting the tunable parameter β=vm/c\boldsymbol{\beta}=\mathbf{v}_{m}/c. We must verify the state dependence assumed in the boost construction; a traveling modulation or cone tilt alone is insufficient. The γ2\gamma^{2}-amplified frequency metric gωωg_{\omega\omega} [Fig. 2(a,d)], quadrupolar shear metric gωyg_{\omega y} [Fig. 2(c)], and electric-like curvature γ\boldsymbol{\gamma} [Fig. 2(b,d)] provide targets for ultrafast pump-probe spectroscopy and wavepacket interferometry. Numerically tested protocols in real-spectrum pseudo-Hermitian systems extract the QGT from energy fluctuations or generalized forces, suggesting complementary readout strategies [21].

Second, the governing equations of the 4D QGT [Eqs. (12)–(14)] connect its metric and curvature components through field conservation and a common microscopic state. We therefore treat the 4D quantum metric gμν(k)g_{\mu\nu}(k) and Berry-Maxwell field Ωμν(k)\Omega_{\mu\nu}(k) as related aspects of band geometry. This complements the quantum-metric contribution to geodesic-like Bloch-electron motion at second order in an electric field [22]. The analogy with gravitational and electromagnetic fields is useful here, but the Weyl metric retains a nonlinear source even away from the monopole. Its geometric conservation law alone does not establish an autonomous radiation field.

Third, the independent Gaussian scalar example illustrates Fourier localization in real spacetime [Fig. 3(a,b)]. Related work demonstrates topological events or studies nonlinear event solitons in specific media [17, 19]. We ask how the coupled metric (gμν)(g_{\mu\nu}) and curvature (Ωμν)(\Omega_{\mu\nu}) influence a physical event's trajectory and deformation. The governing equations constrain how the QGT varies across energy-momentum space. To predict a matter wave event's trajectory and deformation, we must also specify the medium's dispersion and equations of motion [12, 22]. We consider Hermitian bands; non-Hermitian wavepacket dynamics can involve distinct right–right and left–right QGTs [23]. Experiments can test this connection by comparing reconstructed spectral geometry with measured pulse trajectories and shapes. Thus, the historical motivation of Einstein and Schrödinger has a quantum-geometric counterpart: the real and imaginary sectors share a single Hermitian tensor. This common origin concerns quantum-state geometry; it does not unify fundamental gravitational and electromagnetic interactions.

Conclusion. We formulated the 4D quantum geometric tensor Qμν=gμν−i2ΩμνQ_{\mu\nu}=g_{\mu\nu}-\frac{i}{2}\Omega_{\mu\nu} in relativistic energy-momentum space (ω/c,k)(\omega/c,\mathbf{k}), with frequency and momentum as independent coordinates. We identified five Lorentz-invariant contractions of the 4D QGT and showed in 1D, 2D, and 3D Dirac and Weyl systems that Lorentz boosts generate pure frequency metrics gωωg_{\omega\omega}, quadrupolar shear metrics gωig_{\omega i}, and electric-like Berry curvatures Υ\mathbf{\Upsilon}. We used static Weyl geometry and Lorentz covariance to derive the governing equations and flow conservation laws. An independent Gaussian scalar spectral function in (ω/c,k)(\omega/c,\mathbf{k}) space maps to a localized Gaussian event in 4D spacetime. This framework provides a common quantum-geometric description of metric and gauge fields in energy-momentum space and opens an experimental route to connect spectral geometry with event wavepacket motion and deformation.

Acknowledgments

We thank the discussions with Liang Zhang, Jinze He, and Zirui Zhao. We also acknowledge the use of Google Gemini Argon-4 and OpenAI GPT-6.1 Sol for fruitful discussions and cross-checks. Y. P. acknowledges support from the NSFC (No. 2023X0201-417-03) and start-up funding from ShanghaiTech University.

Please address correspondence and requests for materials to Y.P.: [email protected].

Data and code availability

We will make the code repository publicly available upon acceptance.

Conflict of interest

The author declares no competing interests.

End Matter

Floquet-Bloch states, Lorentz Jacobian, and geometric bounds.

We use c=1c=1 throughout the main text and End Matter, where cc denotes the speed of light in vacuum. The 4D energy-momentum coordinates kμ=(ω/c,k)k^\mu=(\omega/c,\mathbf{k}) are conjugate to real spacetime xμ=(ct,r)x^\mu=(ct,\mathbf{r}), with Minkowski metric ημν=diag⁡(−1,1,1,1)\eta_{\mu\nu}=\operatorname{diag}(-1,1,1,1). We write τ=ct=t\tau=ct=t for the real-space Fourier plot and □k=∇k2−∂ω2\Box_{\mathbf{k}}=\nabla_{\mathbf{k}}^2-\partial_\omega^2 for the wave operator in these units. We treat ω\omega and k\mathbf{k} as independent spectral coordinates and construct the laboratory-frame state ∣u(k,ω)⟩|u(\mathbf{k},\omega)\rangle for the Dirac and Weyl models by evaluating the static eigenstate ∣u0(k′)⟩|u_0(\mathbf{k}')\rangle at the boosted momentum:

∣u(k,ω)⟩=∣u0(k′)⟩,k′=(γ(kx−βω),ky,kz),(17)|u(\mathbf{k},\omega)\rangle=|u_0(\mathbf{k}')\rangle,\qquad\mathbf{k}'=(\gamma(k_x-\beta\omega),k_y,k_z), \tag*{(17)}

where q≡k′=(qx,ky,kz)\mathbf{q}\equiv\mathbf{k}'=(q_x,k_y,k_z) is the rest-frame momentum label used in Figure 1 and Figure 2(d). Equation (17) defines the spectral family in a fixed internal basis by evaluating the rest-frame state at the boosted momentum. The chain rule ∂μ=Λμα∂α′\partial_\mu=\Lambda^\alpha_\mu\partial'_\alpha, with Jacobian Λμα=∂k′α/∂kμ\Lambda^\alpha_\mu=\partial k'^\alpha/\partial k^\mu, gives the QGT transformation Qμν=ΛμαΛνβQαβ′Q_{\mu\nu}=\Lambda^\alpha_\mu\Lambda^\beta_\nu Q'_{\alpha\beta}. Since ∂ω∣u0⟩=0\partial_\omega|u_0\rangle=0 in S′S', only spatial indices α,β∈{x,y,z}\alpha,\beta\in\{x,y,z\} enter with Λωx=−βγ\Lambda^x_\omega=-\beta\gamma, Λxx=γ\Lambda^x_x=\gamma, and Λyy=Λzz=1\Lambda^y_y=\Lambda^z_z=1, yielding Eqs. (2)–(4). In Eq. (6), index order matters: I5=QμνQμν=I4−12I2I_5=Q^{\mu\nu}Q_{\mu\nu}=I_4-\frac{1}{2}I_2, whereas QμνQνμ=I4+12I2Q^{\mu\nu}Q_{\nu\mu}=I_4+\frac{1}{2}I_2. To prove Eq. (7), we align Ω′\boldsymbol{\Omega}' along axis 3 in S′S'. Positive semi-definiteness of Qij′Q'_{ij} requires g11′g22′−(g12′)2≥(Ω3′)2/4g'_{11}g'_{22}-(g'_{12})^2\ge(\Omega'_3)^2/4, so Tr⁡g′≥2g11′g22′≥∣Ω′∣\operatorname{Tr}g'\ge2\sqrt{g'_{11}g'_{22}}\ge|\boldsymbol{\Omega}'|. Lorentz invariance of I1=Tr⁡g′I_1=\operatorname{Tr}g' and I2=∣Ω′∣2I_2=|\boldsymbol{\Omega}'|^2 gives I1≥I2I_1\ge\sqrt{I_2} in every frame. Equality holds if and only if g11′=g22′=∣Ω′∣/2g'_{11}=g'_{22}=|\boldsymbol{\Omega}'|/2 and g12′=g33′=0g'_{12}=g'_{33}=0 (Weyl monopole).

Explicit Lorentz-boosted forms for 1D, 2D, and 3D models.

We substitute dkx′=γ(dkx−βdω)dk'_x=\gamma(dk_x-\beta d\omega) and k⊥′=k⊥\mathbf{k}'_\perp=\mathbf{k}_\perp into the rest-frame quantum geometric tensor of the massive Dirac and massless Weyl models to obtain the laboratory-frame field components. For the 1D massive Dirac model in the (1+1)(1+1)-dimensional (ω,kx)(\omega,k_x) space, the Lorentz-boosted 2×22\times2 frequency-momentum quantum metric is:

gμν1D=γ2m24[γ2(kx−βω)2+m2]2(β2−β−β1),Ωμν1D=0.(18)g^{1\mathrm{D}}_{\mu\nu}=\frac{\gamma^2m^2}{4[\gamma^2(k_x-\beta\omega)^2+m^2]^2} \begin{pmatrix} \beta^2 & -\beta\\ -\beta& 1 \end{pmatrix},\qquad\Omega^{1\mathrm{D}}_{\mu\nu}=0. \tag*{(18)}

For the 2D massive Dirac model in the (2+1)(2+1)-dimensional (ω,kx,ky)(\omega,k_x,k_y) space, with kx′=γ(kx−βω)k'_x=\gamma(k_x-\beta\omega) and E′=kx′2+ky2+m2E'=\sqrt{k_x'^2+k_y^2+m^2}, the Lorentz-boosted 3×33\times3 quantum metric is:

gμν2D=14E′4(β2γ2(ky2+m2)−βγ2(ky2+m2)βγkx′ky−βγ2(ky2+m2)γ2(ky2+m2)−γkx′kyβγkx′ky−γkx′kykx′2+m2).(19)g^{2\mathrm{D}}_{\mu\nu}=\frac{1}{4E'^4} \begin{pmatrix} \beta^2\gamma^2(k_y^2+m^2) & -\beta\gamma^2(k_y^2+m^2) & \beta\gamma k'_xk_y\\ -\beta\gamma^2(k_y^2+m^2) & \gamma^2(k_y^2+m^2) & -\gamma k'_xk_y\\ \beta\gamma k'_xk_y & -\gamma k'_xk_y & k_x'^2+m^2 \end{pmatrix}. \tag*{(19)}

The corresponding Lorentz-boosted magnetic-like curvature Ωz=Ωxy\Omega_z=\Omega_{xy} and emergent electric-like curvature γy=−Ωωy\gamma_y=-\Omega_{\omega y} are:

Ωz=−γm2E′3,γy=−βγm2E′3,γx=0.(20)\Omega_z=-\frac{\gamma m}{2E'^3},\qquad\gamma_y=-\frac{\beta\gamma m}{2E'^3},\qquad\gamma_x=0. \tag*{(20)}

For the 3D massless Weyl monopole in the (3+1)(3+1)-dimensional (ω,k)(\omega,\mathbf{k}) space, with k′=kx′2+ky2+kz2\mathbf{k}'=\sqrt{k_x'^2+k_y^2+k_z^2} and k⊥2=ky2+kz2k_\perp^2=k_y^2+k_z^2, the full 4×44\times4 quantum metric gμν3Dg_{\mu\nu}^{3\mathrm{D}} is:

gμν3D=14k′4(β2γ2k⊥2−βγ2k⊥2βγkx′kyβγkx′kz−βγ2k⊥2γ2k⊥2−γkx′ky−γkx′kzβγkx′ky−γkx′kykx′2+kz2−kykzβγkx′kz−γkx′kz−kykzkx′2+ky2).(21)g_{\mu\nu}^{3\mathrm{D}}=\frac{1}{4k'^4} \begin{pmatrix} \beta^2\gamma^2 k_\perp^2 & -\beta\gamma^2 k_\perp^2 & \beta\gamma k'_x k_y & \beta\gamma k'_x k_z \\ -\beta\gamma^2 k_\perp^2 & \gamma^2 k_\perp^2 & -\gamma k'_x k_y & -\gamma k'_x k_z \\ \beta\gamma k'_x k_y & -\gamma k'_x k_y & k_x'^2+k_z^2 & -k_y k_z \\ \beta\gamma k'_x k_z & -\gamma k'_x k_z & -k_y k_z & k_x'^2+k_y^2 \end{pmatrix}. \tag*{(21)}

The boosted Berry-Maxwell fields Ω\boldsymbol{\Omega} and Γ=−β×Ω\boldsymbol{\Gamma}=-\boldsymbol{\beta}\times\boldsymbol{\Omega} are:

Ω=−γ2k′3(kx−βωkykz),Γ=−βγ2k′3(0kz−ky).(22)\boldsymbol{\Omega}=-\frac{\gamma}{2k'^3} \begin{pmatrix} k_x-\beta\omega\\ k_y\\ k_z \end{pmatrix}, \qquad \boldsymbol{\Gamma}=-\frac{\beta\gamma}{2k'^3} \begin{pmatrix} 0\\ k_z\\ -k_y \end{pmatrix}. \tag*{(22)}

Ultraviolet scaling and topological bound of the 2D Dirac quantum weight.

For the 2D massive Dirac model H2D(k′)=kx′σx+ky′σy+mσzH_{\mathrm{2D}}(\mathbf{k}')=k'_x\sigma_x+k'_y\sigma_y+m\sigma_z with E′=k′2+m2E'=\sqrt{k'^2+m^2}, we integrate the rest-frame metric trace Tr⁡g′=(k′2+2m2)/(4E′4)\operatorname{Tr}g'=(k'^2+2m^2)/(4E'^4) and Berry curvature Ωz′=−m/(2E′3)\Omega'_z=-m/(2E'^3) over the UV cutoff disk k′<ΛUVk'<\Lambda_{\mathrm{UV}} with measure d2k′/(2π)d^2k'/(2\pi) to obtain:

WΛ=18[ln⁡(1+ΛUV2m2)+ΛUV2ΛUV2+m2].(23)W_\Lambda=\frac{1}{8}\left[\ln\left(1+\frac{\Lambda_{\mathrm{UV}}^2}{m^2}\right)+\frac{\Lambda_{\mathrm{UV}}^2}{\Lambda_{\mathrm{UV}}^2+m^2}\right]. \tag*{(23)}
CΛ=−sgn⁡(m)2(1−∣m∣ΛUV2+m2),WΛ≥∣CΛ∣.(24)C_\Lambda=-\frac{\operatorname{sgn}(m)}{2}\left(1-\frac{|m|}{\sqrt{\Lambda_{\mathrm{UV}}^2+m^2}}\right),\qquad W_\Lambda\geq|C_\Lambda|. \tag*{(24)}

The pointwise inequality Tr⁡g′−∣Ωz′∣=(E′−∣m∣)2/(4E′4)≥0\operatorname{Tr}g'-|\Omega'_z|=(E'-|m|)^2/(4E'^4)\geq0 implies WΛ≥∣CΛ∣W_\Lambda\geq|C_\Lambda| on any cutoff disk. For ΛUV≫∣m∣\Lambda_{\mathrm{UV}}\gg|m|, CΛ→−sgn⁡(m)/2C_\Lambda\to-\operatorname{sgn}(m)/2 gives the single-cone half-integer flux. The full lattice Hall conductance depends on the global occupied bands and their UV completion [5, 6]. In this limit, the quantum geometric weight WΛ≃[ln⁡(ΛUV2/m2)+1]/8W_\Lambda\simeq[\ln(\Lambda_{\mathrm{UV}}^2/m^2)+1]/8 diverges logarithmically. More generally, quantum geometric weight enters the geometric contributions to flat-band superfluid stiffness and dielectric response [7, 10, 11].

Weyl monopole constitutive reconstruction and governing field equations.

For the 3D Weyl monopole H3D(k′)=k′⋅σH_{\mathrm{3D}}(\mathbf{k}')=\mathbf{k}'\cdot\boldsymbol{\sigma}, let k′=∣k′∣=γ2(kx−βω)2+ky2+kz2k'=|\mathbf{k}'|=\sqrt{\gamma^2(k_x-\beta\omega)^2+k_y^2+k_z^2}. The invariant trace g≡I1=Tr⁡g−gωωg\equiv I_1=\operatorname{Tr}g-g_{\omega\omega} (written as s≡gs\equiv g in Figure 1) saturates Eq. (7): g=Ω2−Γ2=1/(2k′2)g=\sqrt{\Omega^2-\Gamma^2}=1/(2k'^2). We eliminate k′\mathbf{k}' and β\boldsymbol{\beta} between Eqs. (10) and (11) to reconstruct all ten metric components from (Ω,Γ)(\boldsymbol{\Omega},\boldsymbol{\Gamma}):

gωω=γ22g,gω=−Γ×Ω2g,(25)g_{\omega\omega}=\frac{\gamma^2}{2g},\qquad\mathbf{g}_{\omega}=-\frac{\boldsymbol{\Gamma}\times\boldsymbol{\Omega}}{2g}, \tag*{(25)}
gij=δijΩ2−ΩiΩj−γiγj2g.(26)g_{ij} = \frac{\delta_{ij}\Omega^2-\Omega_i\Omega_j-\gamma_i\gamma_j}{2g}. \tag*{(26)}

In S′S', the monopole flux is −2π-2\pi, so ∇k′⋅Ω′=ρw′=−2πδ3(k′)\nabla_{\mathbf{k}'} \cdot\mathbf{\Omega}' = \rho'_w = -2\pi\delta^3(\mathbf{k}'). We account for the longitudinal delta-function Jacobian δ(kx′)=γ−1δ(kx−βω)\delta(k'_x) = \gamma^{-1}\delta(k_x-\beta\omega), so the boosted monopole density ρw=γρw′\rho_w = \gamma\rho'_w and current jw=ρwβ\mathbf{j}_w=\rho_w\boldsymbol{\beta} obey Eq. (12). The root-trace potential φ=g/2=1/(2k′)\varphi=\sqrt{g/2}=1/(2k') satisfies:

□kφ=∇k′2φ=−2πδ3(k′)=ρwγ.(27)\Box_{k}\varphi= \nabla_{\mathbf{k}'}^2\varphi= -2\pi\delta^3(\mathbf{k}') = \frac{\rho_w}{\gamma}. \tag*{(27)}

For the trace-reversed metric gˉμν≡gμν−12ημνg\bar{g}_{\mu\nu}\equiv g_{\mu\nu}-\frac{1}{2}\eta_{\mu\nu}g, we differentiate gˉij′=−ki′kj′/(4k′4)\bar{g}'_{ij}=-k'_ik'_j/(4k'^{4}) and obtain ∂i′gˉij′=0\partial'_i\bar{g}'_{ij}=0 on k′>0k'>0. No singular δ3(k′)\delta^3(\mathbf{k}') source arises in this divergence at k′=0\mathbf{k}'=0 because the small-sphere surface integral is proportional to the vanishing odd angular average ∮n^jdΩ=0\oint\hat{n}_j d\Omega=0. We boost ∂μgˉμν=0\partial^\mu\bar{g}_{\mu\nu}=0 to obtain Eq. (13) and the reciprocal Poynting law ∂tUg+∇k⋅Sg=0\partial_t U_g+\nabla_{\mathbf{k}}\cdot\mathbf{S}_g=0. At second order, we evaluate the Laplacian ∇k′2gij′=ki′kj′/k′6=4(∂i′φ)(∂j′φ)\nabla_{\mathbf{k}'}^2g'_{ij}=k'_ik'_j/k'^{6}=4(\partial'_i\varphi)(\partial'_j\varphi) on k′>0k'>0 and boost the result to obtain Eq. (14) and its trace:

□kgμν=4(∂μφ)(∂νφ),□kg=4g2.(28)\Box_k g_{\mu\nu}=4(\partial_\mu\varphi)(\partial_\nu\varphi),\qquad\Box_k g=4g^2. \tag*{(28)}

The k′−4k'^{-4} singularity at k′=0\mathbf{k}'=0 in the second-order source requires a core mass regularization or finite-part extension. On the regular offset slice qx≡kx′=a=0.35q_x\equiv k'_x=a=0.35 [Figure 2(d)], gωω=β2γ2r⊥2/[4(a2+r⊥2)2]g_{\omega\omega}=\beta^2\gamma^2r_\perp^2/[4(a^2+r_\perp^2)^2] peaks on the ring r⊥=ky2+kz2=∣a∣r_\perp=\sqrt{k_y^2+k_z^2}=|a|.

Windowed spectral envelopes and QGT event wavepacket.

For Figure 3(c,Figure 3d), we define δkx=kx−k0\delta k_x=k_x-k_0 and δω=ω−ω0\delta\omega=\omega-\omega_0, with c=1c=1. The characteristic Gaussian f(q)=exp⁡[−q2/(2σk2)]f(q)=\exp[-q^2/(2\sigma_k^2)], where q=δkx−δωq=\delta k_x-\delta\omega, satisfies (∂kx+∂ω)f=0(\partial_{k_x}+\partial_\omega)f=0. We apply the frequency window W(δω)=exp⁡[−(δω)2/(2σω2)]W(\delta\omega)=\exp[-(\delta\omega)^2/(2\sigma_\omega^2)] to obtain a=fWa=fW and (∂kx+∂ω)a=−δωσω2a(\partial_{k_x}+\partial_\omega)a=-\frac{\delta\omega}{\sigma_\omega^2}a. We Fourier-transform the windowed amplitude in these detunings, with measure dkxdω/(2π)2dk_xd\omega/(2\pi)^2 and τ=ct=t\tau=ct=t, to obtain:

a~(x,τ)=σkσω2πexp⁡[−σk2x22−σω2(x−τ)22].(29)\tilde{a}(x,\tau)=\frac{\sigma_k\sigma_\omega}{2\pi}\exp\left[-\frac{\sigma_k^2x^2}{2}-\frac{\sigma_\omega^2(x-\tau)^2}{2}\right]. \tag*{(29)}

Spectral windowing produces the nonzero residual above. Here, a(k,ω)a(\mathbf{k},\omega) is an independent scalar spectral function of the QGT, and its 4D Fourier transform a~(r,t)\tilde{a}(\mathbf{r},t) illustrates spacetime localization. Neither function is identified here with a physical matter-wave or optical-pulse envelope multiplying the normalized internal state ∣u(k,ω)⟩|u(\mathbf{k},\omega)\rangle. Equations (12)–(14) constrain how the components of Qμν(k,ω)Q_{\mu\nu}(\mathbf{k},\omega) vary across (k,ω)(\mathbf{k},\omega) space. A physical matter-wave field ∣Ψ(r,t)⟩|\Psi(\mathbf{r},t)\rangle instead obeys the medium's dynamical equations. In semiclassical wavepacket descriptions, the curvature Ωμν\Omega_{\mu\nu} drives anomalous deflection, while the metric gμνg_{\mu\nu} contributes to finite-width spreading and geodesic-like gradient forces [12, 22]. Figure 2 uses the upper band with β=0.75\beta=0.75 and m=0.8m=0.8 for the massive models; Figure 3(c,Figure 3d) use σk=0.55\sigma_k=0.55 and σω=1\sigma_\omega=1.

References

  1. [1]A. Einstein, A generalization of the relativistic theory of gravitation, Ann. Math. 46, 578 (1945).DOI
  2. [2]E. Schrödinger, Space-Time Structure (Cambridge University Press, Cambridge, 1950).
  3. [3]J. P. Provost and G. Vallee, Riemannian structure on manifolds of quantum states, Commun. Math. Phys. 76, 289 (1980).DOI
  4. [4]M. V. Berry, Quantal phase factors accompanying adiabatic changes, Proc. R. Soc. Lond. A 392, 45 (1984).
  5. [5]D. J. Thouless, M. Kohmoto, M. P. Nightingale, and M. den Nijs, Quantized Hall conductance in a two-dimensional periodic potential, Phys. Rev. Lett. 49, 405 (1982).DOI
  6. [6]F. D. M. Haldane, Model for a quantum Hall effect without Landau levels: Condensed-matter realization of the parity anomaly, Phys. Rev. Lett. 61, 2015 (1988).pubmed.ncbi.nlm.nih.gov/10038961
  7. [7]S. Peotta and P. Törmä, Superfluidity in topologically nontrivial flat bands, Nat. Commun. 6, 8944 (2015).DOI
  8. [8]T. Ozawa and B. Mera, Relations between topology and the quantum metric for Chern insulators, Phys. Rev. B 104, 045103 (2021).DOI
  9. [9]I. Souza, T. Wilkens, and R. M. Martin, Polarization and localization in insulators: Generating function approach, Phys. Rev. B 62, 1666 (2000).DOI
  10. [10]Y. Onishi and L. Fu, Fundamental bound on topological gap, Phys. Rev. X 14, 011052 (2024).DOI
  11. [11]Y. Onishi and L. Fu, Quantum weight: A fundamental property of quantum many-body systems, Phys. Rev. Research 7, 023158 (2025).DOI
  12. [12]A. Li, L. Dong, and Q. Niu, Geodynamics and artificial gravity in spacetime crystals under slow perturbation and deformation, Phys. Rev. Research 8, 013028 (2026).arxiv.org/abs/2310.16918
  13. [13]J.-C. Tang, X.-Y. Hou, Y.-H. Huang, H. Guo, and C.-C. Chien, Evolution of quantum geometric tensor of one-dimensional periodic systems after a quench, Phys. Rev. B 113, 174301 (2026).DOI
  14. [14]G. Konstantinou, Time-Independent Parameters in Quantum Systems: Revisiting Berry Phase, Curvature and Gauge Connections, arXiv:2507.23347v2 (2025).arxiv.org/abs/2507.23347
  15. [15]R. Shindou and L. Balents, Artificial electric field in Fermi liquids, Phys. Rev. Lett. 97, 216601 (2006).DOI
  16. [16]Y. Pan and R. Yin, Constructing Berry-Maxwell equations with Lorentz invariance and Gauss's law of Weyl monopoles in four-dimensional energy-momentum space, Phys. Rev. B 110, 075139 (2024).DOI
  17. [17]J. Feis, S. Weidemann, T. Sheppard, H. M. Price, and A. Szameit, Space-time-topological events in photonic quantum walks, Nat. Photonics 19, 518 (2025).DOI
  18. [18]L. Zhang, Z. Zhao, Q. Pan, C. Pan, Q. Cheng, and Y. Pan, Topological event wavepackets in energy-momentum-gapped photonic spacetime crystals, Laser Photonics Rev. 20, e02603 (2026).DOI
  19. [19]L. Zhang, Z. Fan, and Y. Pan, Event soliton formation in mixed energy-momentum gaps of nonlinear spacetime crystals, Phys. Rev. Research 8, 023107 (2026).arxiv.org/abs/2503.16113
  20. [20]Y. Pan and G. Chen, Generation and manipulation of multipole and vortex events in (1+1)-dimensional spacetime, arXiv:2608.00618 (2026).arxiv.org/abs/2608.00618
  21. [21]Z.-H. Huang, H.-T. Ding, and L.-J. Lang, Direct measurement of the quantum geometric tensor in pseudo-Hermitian systems, Phys. Rev. A 112, 062225 (2025).DOI
  22. [22]T. B. Smith, L. Pullasseri, and A. Srivastava, Momentum-space gravity from the quantum geometry and entropy of Bloch electrons, Phys. Rev. Research 4, 013217 (2022).arxiv.org/abs/2108.02216
  23. [23]Y.-M. R. Hu, E. A. Ostrovskaya, and E. Estrecho, Quantum geometric tensor and wavepacket dynamics in two-dimensional non-Hermitian systems, Phys. Rev. Research 7, L012067 (2025).arxiv.org/abs/2412.08141

Paper details

vlnc:W_m8cbjxqv
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 1

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Google Gemini Argon-4, OpenAI GPT-6.1 Sol

View statement
Looking for related papers…
Contents