Geometric Analysis on Non-Smooth Spaces with Lower Ricci Bounds

46 concepts10 papers1 proposal

In classical Riemannian geometry, smooth manifolds are studied using tools that rely on derivatives. At the heart of this approach is the Ricci curvature tensor, which measures how geometric volumes deviate from Euclidean space. When a manifold has a lower bound on its Ricci curvature, it constrains how functions behave. For instance, it limits how fast heat diffuses, bounds the eigenvalues of the Laplacian, and forces rigid structural symmetries on the space itself.

However, many geometric spaces of interest are not smooth. They may have sharp corners, conical singularities, or arise as the limits of sequences of smooth manifolds. On these singular spaces, classical calculus breaks down because we can no longer define a pointwise Ricci tensor or take standard second derivatives.

To bypass this limitation, a powerful research direction has emerged: translating geometric analysis from smooth manifolds to non-smooth spaces using synthetic curvature bounds. By redefining curvature through the macroscopic behavior of geodesics, optimal transport, and heat diffusion, researchers can recover the profound rigidity and analytical estimates of smooth geometry without requiring smoothness.

The Core Challenge: Replacing the Bochner Formula

To understand why this direction is challenging, consider one of the foundational pillars of smooth geometric analysis: the Bochner formula. For a smooth function $f$ on a Riemannian manifold, the formula states:

$$\frac{1}{2} \Delta |\nabla f|^2 = |Hess f|^2 + \langle \nabla \Delta f, \nabla f \rangle + Ric(\nabla f, \nabla f)$$

If the Ricci curvature ($Ric$) is bounded below, this formula yields powerful differential inequalities. For example, Yau’s gradient estimates for harmonic functions and the Li-Yau estimates for the heat equation both flow directly from applying the maximum principle to variations of the Bochner formula.

On a singular space, such as an Alexandrov space (a metric space with a lower bound on sectional curvature via triangle comparison), the terms $|Hess f|^2$ and $Ric(\nabla f, \nabla f)$ lack pointwise meaning. To overcome this, Xi-Ping Zhu developed a generalized, weak Bochner-type formula tailored for Alexandrov spaces in Yau's Gradient Estimates on Alexandrov Spaces. By integrating against test functions and leveraging the space's intrinsic metric properties, the author bypassed the need for smooth second derivatives, successfully extending Yau's classical gradient estimates to settings with explicit singularities.

Synthetic Curvature Frameworks

To scale these insights to even more general settings, mathematicians rely on two primary synthetic frameworks that capture lower Ricci bounds:

  1. Alexandrov Spaces with Ricci Bounds: While Alexandrov spaces natively possess sectional curvature bounds via triangle geometry, defining a pure Ricci bound requires global tracking of geodesics. In Ricci curvature on Alexandrov spaces and rigidity theorems, Xi-Ping Zhu adapted smooth Riemannian techniques—such as Busemann functions and gradient flows—to establish a working definition of lower Ricci bounds. This framework proved robust enough to recover classical global rigidity results like the Cheeger-Gromoll splitting theorem and Cheng's maximal diameter theorem.
  2. The $RCD^*(K,N)$ Framework: For general metric measure spaces, the Riemannian Curvature-Dimension ($RCD$) condition provides a measure-theoretic generalization of Ricci curvature bounded below by $K$ and dimension bounded above by $N$. It uses the language of optimal transport, dictating that an entropy functional is strongly convex along Wasserstein geodesics, combined with a linearity condition on the heat flow to enforce "Riemannian" rather than "Finslerian" behavior.

Heat Diffusion as a Regularizer

Why do synthetic bounds work so well when analyzing heat equations? The intuition lies in the regularizing nature of diffusion. If you start with a rough, non-smooth function on a space, running the heat flow smooths it out instantly. The rate of this smoothing is fundamentally governed by the geometry of the space. If a space pinches sharply or fractures, heat traps or disperses erratically. A lower Ricci bound prevents these pathological geometries.

Using the $RCD^*(K,N)$ framework, Renjin Jiang demonstrated in Hamilton’s gradient estimates and a monotonicity formula for heat flows on metric measure spaces that classical Li-Yau-Hamilton gradient estimates and Perelman's $W$-entropy monotonicity formulas remain perfectly valid on non-smooth spaces. The heat flow acts as a bridge, allowing smooth differential inequalities to be interpreted weakly via Dirichlet forms and function regularity.

Smooth Manifold          ---> Pointwise Derivatives ---> Classical Bochner Formula
                                                                |
                                                                v
Non-Smooth Space (RCD)   ---> Optimal Transport     ---> Weak/Synthetic Gradient Estimates

While global estimates capture the macroscopic behavior of the space, local estimates are far harder to secure because they require isolating regions without the help of smooth coordinate charts. In Local Li–Yau’s estimates on RCD $^*(K,N)$ metric measure spaces, Xi-Ping Zhu proved that synthetic lower Ricci bounds are localized enough to support fine, regional geometric analysis, filling a long-standing gap in the study of limits of manifolds.

Sharp Bounds and Global Structure

Once gradient estimates for the heat equation are secured, they unlock precise control over the heat kernel—the fundamental solution to the heat equation. Renjin Jiang and Huaiqian Li established sharp upper and lower bounds for the heat kernel and its gradient on $RCD^*(K,N)$ spaces in Heat Kernel Bounds on Metric Measure Spaces and Some Applications, overcoming the lack of classical tools like the Gauss-Green integration-by-parts formula.

These precise heat kernel bounds have immediate consequences for the global structure of the space, notably through the spectral gap. The spectral gap is the first non-zero eigenvalue ($\lambda_1$) of the Laplacian, which physically represents the fundamental vibrational frequency of the space or the rate at which gas diffuses to equilibrium.

In Sharp spectral gaps on metric measure spaces, Yin Jiang proved that the sharp lower bounds for spectral gaps from smooth manifolds carry over entirely to $RCD^*(K,N)$ spaces. Similarly, Zhongmin Qian and Xi-Ping Zhu established sharp spectral gaps and an analogue of Obata's theorem for Alexandrov spaces in Sharp spectral gap and Li–Yau’s estimate on Alexandrov spaces. Obata's theorem is a rigidity result: it shows that if the spectral gap hits its absolute geometric minimum, the underlying space must possess a highly symmetric, rigid structure analogous to a round sphere.

Broader Horizons: Large Deviations and Algebraic Spaces

The success of using heat flows to probe non-smooth geometry extends beyond finite-dimensional metric spaces. For instance, when studying infinite-dimensional or non-locally-compact spaces under the $RCD(K, \infty)$ condition, standard local compactness arguments fail. To bridge this, Nicola Gigli and Luca Tamanini in Viscosity Solutions of Hamilton–Jacobi Equation in $\textsf{RCD}(K,\infty )$ Spaces and Applications to Large Deviations utilized viscosity solutions of the Hamilton-Jacobi equation. By proving that viscous approximations converge to the Hopf-Lax formula, the authors verified Varadhan's formula, linking small-time heat kernel asymptotics directly to the underlying metric distance via large deviations principles.

Concurrently, these analytical techniques help map the boundaries of where classical geometric intuition fails. In On the H.-Q. Li inequality on step-two Carnot groups, Ye Zhang studied spaces where the classical lower Ricci bound cannot be satisfied uniformly. By analyzing the behavior of the heat semigroup gradient constant, the author demonstrated that certain algebraic structures inherently resist non-negative Ricci curvature, highlighting the precise boundary where smooth and singular geometries diverge. Even on non-compact complex spaces, modified or perturbed heat flows remain the primary tool for proving the existence of canonical structures, as shown by Shi-Fan Cai et al. in the context of twisted quiver bundles.

Ultimately, this research direction demonstrates that the deep relationships between curvature, diffusion, and topology are not accidental artifacts of smooth calculus. Instead, they are fundamental geometric truths that persist even in the presence of severe singularities, provided one views curvature not as a collection of derivatives, but as a structural constraint on space and measure.

Key Foundational Work

Ricci curvature on Alexandrov spaces and rigidity theorems (2010) — contributed 13 concept elements to this direction.
Sharp Spectral Gap and Li-Yau's Estimate on Alexandrov Spaces (2012) — contributed 6 concept elements to this direction.
Yau's Gradient Estimates on Alexandrov Spaces (2012) — contributed 2 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_adaptation_of_riemannian_techniques_fa6357a9,
  title = {adaptation of Riemannian techniques},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/adaptation-of-riemannian-techniques/rdr_fa6357a9df4c40a8b593bd0a2460ecd6},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_adaptation_of_riemannian_techniques_fa6357a9_v1,
  title = {adaptation of Riemannian techniques},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/adaptation-of-riemannian-techniques/rdr_fa6357a9df4c40a8b593bd0a2460ecd6/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake