---
title: Chen-Ricci Inequalities in Geometric Analysis
url: https://www.emergentmind.com/topics/chen-ricci-inequalities
type: topic
---

# Chen-Ricci Inequalities in Geometric Analysis

Searching arXiv for recent and foundational papers on Chen-Ricci inequalities and closely related Chen-type inequalities.
Chen-Ricci inequalities are curvature pinching inequalities that relate intrinsic curvature quantities—most prominently Ricci curvature, scalar curvature, sectional curvature, or Chen’s $\delta$-invariants—to extrinsic quantities such as the mean curvature vector, the second fundamental form, or, in mapping-theoretic settings, the mean curvature of fibers and the tension or second fundamental tensors of a map. In the classical submanifold setting, they measure the gap between intrinsic and extrinsic geometry; in later generalizations, the same pattern appears for curvature-like tensors, CR-warped products, contact and quaternionic geometries, Riemannian submersions, and Riemannian maps [1104.3439], [2606.18725]. A related but distinct line of work studies “Chen-Ricci type” inequalities among Ricci invariants in Lorentzian geometry, where the extrinsic-submanifold interpretation is replaced by algebraic inequalities for traces of powers of the Ricci tensor [2509.22187].

## 1. Classical form and the curvature-like tensor framework

For an $n$-dimensional submanifold $M$ of a real space form $\widetilde{M}(c)$, the classical Chen-Ricci inequality recorded in the literature cited by the 2011 paper is
\[
|H|^2 \geq \frac{1}{n(n-1)} \left\{ \mathrm{Ric}(X) - (n-1)c \right\}
\qquad \forall~\text{unit}~X \in T_p M,
\]
where $|H|^2$ is the squared norm of the mean curvature vector, $\mathrm{Ric}(X)$ is the Ricci curvature in the direction $X$, and $c$ is the constant sectional curvature of the ambient space [1104.3439]. In the summary accompanying that work, this inequality is described as expressing a deep relationship between intrinsic and extrinsic invariants.

A major abstraction of the subject is the curvature-like tensor formulation. In that setting one considers a Riemannian manifold $(M,g)$, a Riemannian vector bundle $(B,g_B)$ over $M$, a $B$-valued symmetric $(1,2)$-tensor field $\zeta$, and a curvature-like tensor $T$ satisfying the algebraic Gauss equation
\[
T(X, Y, Z, W) = g_B(\zeta(X, W), \zeta(Y, Z)) - g_B(\zeta(X, Z), \zeta(Y, W)).
\]
The corresponding Chen-Ricci inequality takes the form
\[
\mathrm{Ric}_T(X) \leq |\operatorname{trace} \zeta|^2
\]
for any unit vector $X \in T_pM$ [1104.3439]. Here $\operatorname{trace}\zeta$ plays the role of a mean-curvature-type vector.

The equality case is already rigid at this general level. It is attained if and only if $\zeta(X,Y)=0$ for all $Y\perp X$ and $\zeta(X,X)=\operatorname{trace}(\zeta)$ [1104.3439]. This algebraic characterization is one of the reasons the curvature-like tensor framework became useful: it isolates the mechanism behind Chen-Ricci inequalities from any single ambient geometry and makes possible a unified treatment of Kähler, Sasakian, and related contexts. This suggests that a substantial part of the theory is driven less by a specific ambient curvature tensor than by the Gauss-type algebraic structure itself.

## 2. Improved inequalities and sharpness phenomena

The main advance of "Improved Chen-Ricci inequality for curvature-like tensors" is an improved inequality under certain algebraic conditions on $\zeta$:
\[
\mathrm{Ric}_T(X) \leq \frac{n}{n-1} |\operatorname{trace}\zeta|^2
\qquad \text{for all unit } X \in T_pM,\ n\geq 2
\]
[1104.3439]. The associated equality statement is highly restrictive: equality holds if and only if either $\zeta=0$ or $n=2$ and $\zeta$ has a special algebraic form [1104.3439].

The same paper emphasizes sharpness and rigidity in its applications. In the summary, the improved bound is described as setting a finer upper bound and characterizing equality typically only in highly constrained cases such as the totally geodesic case or very special $H$-umbilical forms [1104.3439]. This is representative of Chen-type inequalities more broadly: the numerical coefficient is significant, but the geometry of the equality case is often even more important.

Later work transferred this sharpness paradigm to other settings. For Riemannian maps, a 2025 paper gives a general Chen-Ricci inequality
\[
{\rm Ric}^{\mathcal{H}}(X) \leq {\rm Ric}^{\mathcal{R}}(F_*X) + \frac{1}{4}\|{\rm trace}\, B^{\mathcal H}\|^2,
\]
and an improved Chen-Ricci inequality
\[
{\rm Ric}^{\mathcal{H}}(X) \leq {\rm Ric}^{\mathcal{R}}(F_*X) + \frac{r-1}{4r}\|{\rm trace}\, B^{\mathcal H}\|^2,
\]
where $r$ is the rank of the map [2509.15281]. The 2026 paper on generalized Chen inequalities for submersions and maps likewise states that it derives optimal inequalities involving generalized $\delta$-invariants and characterizes equality cases through precise algebraic conditions on the corresponding shape operators [2606.18725]. In these mapping-theoretic generalizations, the role played classically by the second fundamental form of an immersion is assumed by O’Neill tensors, second fundamental tensors of the submersion, or the second fundamental form of the Riemannian map.

A recurring misconception is to read every “improved” Chen-Ricci inequality as merely a better coefficient. The cited works show that the improvement is inseparable from the algebraic hypotheses under which it is valid and from the resulting rigidity statements [1104.3439], [2509.15281].

## 3. Submanifolds of complex and contact manifolds

The 2011 paper applies the improved inequality to several special classes of submanifolds [1104.3439]. For Lagrangian submanifolds $M^n \subset \widetilde{M}^{2n}(c)$ of complex space forms, it recovers and sharpens the inequality
\[
\mathrm{Ric}(X) \leq c + n|H|^2.
\]
Equality holds if either $p$ is a geodesic point or, for $n=2$, $M$ is $H$-umbilical with
\[
\sigma(e_1, e_1) = 3\mu Je_1,\quad
\sigma(e_2, e_2) = \mu Je_1,\quad
\sigma(e_1, e_2) = \mu Je_2
\]
for suitable local frames and function $\mu$ [1104.3439]. The same summary notes the connection of this equality case with Lagrangian $H$-umbilical surfaces such as the Whitney 2-sphere.

For Kaehlerian slant submanifolds of complex space forms with slant angle $\theta$, the inequality becomes
\[
\mathrm{Ric}(X) \leq (n-1)\left( n|H|^2 + c + 3c\cos^2\theta \right),
\]
so the ambient holomorphic structure enters explicitly through $\cos^2\theta$ [1104.3439]. For $C$-totally real submanifolds of Sasakian space forms, the bound is
\[
\mathrm{Ric}(X) \leq (n-1)(c+3) + n|H|^2,
\]
with equality characterized by an $H$-umbilical-type condition
\[
\sigma(e_1,e_1)=3\mu\varphi e_1,\quad
\sigma(e_2,e_2)=\mu\varphi e_1,\quad
\sigma(e_1,e_2)=\mu\varphi e_2
\]
[1104.3439].

Subsequent papers extended Chen-Ricci inequalities in contact settings. For $C$-totally real submanifolds in generalized $(\kappa,\mu)$-contact space forms, one has
\[
\operatorname{Ric}(U) \leq \frac{1}{4} n^2 \| H \|^2 + (n-1) f_1
 + f_4 \big( \operatorname{tr}(h^T) + (n-2) g(h^T U, U) \big)
 + f_{5,1} \big( \operatorname{tr}(h^T) g(h^T U, U) - \| h^T U \|^2 \big)
 + f_{5,2} \big( \operatorname{tr}((\phi h)^T) g((\phi h)^T U, U) - \| (\phi h)^T U \|^2 \big),
\]
and in the generalized Sasakian case with $f_3=f_1-1$ and $h=0$, this simplifies to
\[
\operatorname{Ric}(U) \leq \frac{1}{4} n^2 \|H\|^2 + (n-1) f_1
\]
[1512.07647]. The same paper states that if $H=0$, equality for a unit vector occurs if and only if the vector lies in the relative null space; equality for all unit vectors at a point occurs only if the point is totally geodesic, or if $n=2$ and the point is totally umbilical [1512.07647].

A different contact-geometric generalization, in $(\kappa,\mu)$-contact space forms with generalized semi-symmetric non-metric connections, yields a Chen-Ricci inequality of the form
\[
\operatorname{Ric}(X) \leq \frac{(n-1)}{4}(c+3) + \frac{3}{4}(c-1)|PX|^2 + \frac{c+3-4\kappa}{4}\left[(n-2)\eta(X)^2 + |\xi^T|^2\right] + \cdots + \frac{n^2}{4}|H|^2,
\]
where the omitted terms depend on the structure tensors and the connection parameters [2003.00185]. This indicates that non-Levi-Civita ambient connections alter the classical pinching pattern by adding explicit correction terms while preserving the basic intrinsic–extrinsic scheme.

## 4. Warped products, CR geometry, and Kähler-type ambients

Chen-type inequalities for warped products usually place the second fundamental form, the warping function, and ambient structure tensors in a single formula. For CR-warped product submanifolds of a locally conformal Kähler space form $\tilde M^{2m}(c)$, the main inequality is
\[
\|h\|^2 \geq -2p(\Delta\ln f) + 4p \| \nabla \ln f \|^2 - 4pk(\lambda(\ln f)) - p|\theta|^2 + 2k|\theta|^2 + G + 2pF,
\]
where $p=\dim N_\perp$, $2k=\dim N_T$, $\theta$ is the Lee form, $\lambda$ the Lee vector field, and $G,F$ are the auxiliary terms specified in the paper [1404.3468]. In the Vaisman subclass one obtains
\[
\|h\|^2 \geq -2p(\Delta \ln f) + 4p\| \nabla \ln f \|^2 - 4pk(\lambda(\ln f)) - p|\theta|^2 + 2pG^*.
\]
Equality in the general LCK inequality holds if and only if $N_T$ is totally geodesic, $N_\perp$ is totally umbilical, and the mixed second fundamental forms satisfy the stated orthogonality conditions [1404.3468].

In Bochner Kähler geometry, the inequalities acquire additional ambient curvature terms. For an $n$-dimensional submanifold $W$ of a Bochner Kähler manifold, Theorem 3.1 in the summary gives
\[
K(\pi)  \geq \left( \frac{5n^2 + 31n + 26 + 3 \|T\|^2}{2(2n + 2)(2n + 4)} \right) \rho
- \frac{n^2(n-2)}{2(n-1)} \|H\|^2
+ \frac{6}{2(2n+4)} \sum_{i,j=1}^n \mathrm{Ric}(e_j, Je_j) g(e_i, Je_j),
\]
with analogous slant, invariant, anti-invariant, and Einstein specializations [1601.04130]. The same paper also establishes, for a warped product CR-submanifold $W=W_T\times_f W_\perp$,
\[
\| h \|^2 \geq \| h_{\mathcal{D}\mathcal{D}} \|^2 + q \|\nabla^{\mathcal{D}}\log f\|^2,
\]
and states that if $W$ is compact and orientable, $PD\mathcal D$ lies in $\mathcal D$, and certain symmetry holds for the shape operators, then
\[
\rho \leq 0,
\]
with equality if and only if $f$ is constant on the base [1601.04130].

A 2026 development moves from second-fundamental-form inequalities to a first Chen inequality for CR-warped product submanifolds of a complex space form and introduces the leaf-wise first Chen invariant
\[
\hat\delta(V)(x) := \tau^M(V) - \inf\{ K^M(\pi) : \pi\subseteq V,\, \dim\pi=2 \}
\]
[2605.19601]. For $M^n=N_T^{n_1}\times_f N_\perp^{n_2}$ in a complex space form $\tilde M^{2m}(c)$, the holomorphic factor satisfies
\[
\hat{\delta}(T_x N_T)(x) \leq \frac{n^2}{2} \|\vec{H}\|^2 - \frac{n_2\Delta f}{f} + \frac{n_1(n_1+2n_2+2)}{8} c - \tilde{K}_{\min}(T_x N_T),
\]
and the totally real factor satisfies
\[
\hat{\delta}(T_x N_\perp)(x) \leq \frac{n^2}{2} \|\vec{H}\|^2 - \frac{n_2\Delta f}{f} + \frac{n_2(n_2+2n_1-1)}{8} c - \frac{c}{4}
\]
[2605.19601]. The same paper stresses a distinction that is easily blurred: on $N_T$, $\hat\delta(T_xN_T)$ coincides with the intrinsic Chen invariant of $N_T$, whereas on $N_\perp$ it is related to the intrinsic Chen invariant by the Bishop–O’Neill formula [2605.19601]. This directly addresses a common confusion between leaf-wise and intrinsic $\delta$-invariants in warped products.

## 5. Riemannian submersions and Riemannian maps

Recent work extends Chen-type inequalities from immersions to smooth mappings with distinguished vertical and horizontal distributions. For a Riemannian submersion $F:(M_1^n,g_1)\to(M_2^m,g_2)$ with fiber dimension $r=n-m>2$, mean curvature vector field $H$ of the fibers, and a vertical $2$-plane $\Pi\subset\mathcal V_p$, the 2026 paper on Chen’s first inequalities proves
\[
\tau_{\mathcal V}^{\ker F_*}(p) - K_{\mathcal V}^{\ker F_*}(\Pi)
\geq
\tau_{\mathcal V}^{M_1}(p) - K_{\mathcal V}^{M_1}(\Pi)
-
\frac{r^2(r-2)}{2(r-1)}\|H\|^2
\]
[2602.15301]. Equality holds if and only if explicit algebraic conditions on the components of the O’Neill tensor ${\mathcal T}^{\mathcal H}$ are satisfied [2602.15301]. In a real space form $M_1(c)$ this simplifies to
\[
\tau_{\mathcal V}^{\ker F_*}(p) - K_{\mathcal V}^{\ker F_*}(\Pi)
\geq
\frac{1}{2} \left\{ c(r^2 - r - 2) - \frac{r^2(r-2)}{(r-1)} \|H\|^2 \right\}
\]
[2602.15301].

The 2025 paper "General Chen-Ricci inequalities for Riemannian submersions and Riemannian maps" gives distributional Ricci versions. For the vertical distribution of a Riemannian submersion,
\[
{\rm Ric}^{\ker F_*}_{\mathcal V}(\vee_1)
\geq
{\rm Ric}^{M_1}_{\mathcal V}(\vee_1) - \frac{1}{4}n^2\|H\|^2,
\]
while for the horizontal distribution,
\[
{\rm Ric}^{(\ker F_*)^\perp}_{\mathcal H}(h_1)
\leq
{\rm Ric}^{M_1}_{\mathcal H}(h_1)
\]
[2509.15281]. The same paper also states a combined vertical–horizontal inequality involving the mixed curvature terms, $\delta(N)$, $\|T^{\mathcal V}\|^2$, $\|A^{\mathcal H}\|^2$, and the mean curvature of the fibers [2509.15281].

These formulas become explicit in generalized complex and generalized Sasakian space forms. For example, for a Riemannian submersion from a generalized complex space form,
\[
{\rm Ric}^{\ker F_*}_{\mathcal V}(\vee_1)
\geq
c_1(n-1) + 3c_2\|Q\vee_1\|^2 - \frac{1}{4}n^2\|H\|^2,
\]
and
\[
{\rm Ric}^{(\ker F_*)^\perp}_{\mathcal H}(h_1)
\leq
c_1(r-1) + 3c_2\|Ph_1\|^2
\]
[2509.15281]. For invariant, anti-invariant, and slant submersions in complex space forms, the same paper records, respectively,
\[
{\rm Ric}^{\ker F_*}_{\mathcal V}(\vee_1) \geq \frac{c}{4}(n-1) + \frac{3c}{4} - \frac{1}{4}n^2\|H\|^2,
\]
\[
{\rm Ric}^{\ker F_*}_{\mathcal V}(\vee_1) \geq \frac{c}{4}(n-1) - \frac{1}{4}n^2\|H\|^2,
\]
and
\[
{\rm Ric}^{\ker F_*}_{\mathcal V}(\vee_1) \geq \frac{c}{4}(n-1) + \frac{3c}{4}\cos^2\theta - \frac{1}{4}n^2\|H\|^2
\]
[2509.15281].

For Riemannian maps, the same paper gives the general Chen-Ricci inequality
\[
{\rm Ric}^{\mathcal H}(X)\leq {\rm Ric}^{\mathcal R}(F_*X) + \frac{1}{4}\|{\rm trace}\,B^{\mathcal H}\|^2,
\]
and the improved version
\[
{\rm Ric}^{\mathcal H}(X)\leq {\rm Ric}^{\mathcal R}(F_*X) + \frac{r-1}{4r}\|{\rm trace}\,B^{\mathcal H}\|^2
\]
[2509.15281]. Related first-inequality formulations for Riemannian maps appear in [2510.10505], where, for every plane section $\mathbb P\subset T_pM$,
\[
K^{\mathcal H}(\mathbb P)
\geq
\frac{1}{2}\left\{
2\rho^{\mathcal H}
-\frac{r-2}{r-1}\|\tau^{\mathcal H}\|^2
-2\rho^{\mathcal R}
+2K^{\mathcal R}(\mathbb P)
\right\},
\]
with specializations to generalized complex and generalized Sasakian space forms [2510.10505]. The 2026 paper on generalized Chen inequalities for submersions and maps then recasts these developments in terms of generalized $\delta$-invariants for mutually orthogonal subspaces and states that, for special low-dimensional choices such as $k=1$ and $r_1=2$, these recover sharp bounds for Ricci curvature of vertical or horizontal distributions [2606.18725].

## 6. Related analogues, interpretation, and scope

The accumulated literature shows that “Chen-Ricci inequality” now names a family of related estimates rather than a single formula. In submanifold geometry, the common pattern is a comparison between intrinsic curvature and extrinsic shape, usually via $|H|^2$ or $\|h\|^2$ [1104.3439], [1404.3468], [1512.07647]. In the geometry of submersions and maps, the extrinsic terms become the mean curvature of fibers, O’Neill tensors, or the trace of the second fundamental form of the map [2602.15301], [2509.15281], [2606.18725]. In quaternionic Kähler geometry with a Ricci quarter-symmetric metric connection, the same pattern persists but with additional correction terms involving the quaternionic structure and the connection tensors. For instance, the paper on this topic gives
\[
\mathrm{Ric}(X) \leq c \left\{ (n-1) + 3\sum_{k=1}^3 \sum_{j=2}^n g^2(\varphi_k X, e_j) \right\}
- c \left\{ (n-1) + 3\sum_{k=1}^3  |P_k|^2 n \right\}[m + (n-2)M(X,X)] + \frac{n^2}{4} |H|^2
\]
[2011.08582].

A separate but explicitly acknowledged analogue appears in spacetime geometry. For a $D$-dimensional spacetime of Segre class $A1$, $A3$, or $B$, one has
\[
R_s^{2m} \leq D^{2m-s}(R_{2m})^s
\qquad (s<2m),
\]
and the special case
\[
R_1^2 \leq D R_2
\]
is described in the summary as the direct analogue of the Chen-Ricci inequality [2509.22187]. This is not a submanifold inequality and does not compare intrinsic and extrinsic data; rather, it is an algebraic inequality among Ricci invariants. The distinction is important, because the phrase “Chen-Ricci type” here identifies formal analogy rather than identity of geometric setting.

Across the cited literature, several themes remain stable. First, the inequalities are typically sharp, and the equality cases are characterized by strong algebraic conditions on the second fundamental form, the shape operators, or O’Neill tensors [1104.3439], [1512.07647], [2602.15301], [2606.18725]. Second, many later results are explicit specializations obtained by inserting the curvature tensors of complex, Sasakian, Kenmotsu, cosymplectic, generalized complex, generalized Sasakian, or Bochner Kähler ambient spaces into a general master inequality [1601.04130], [2509.15281], [2510.10505]. Third, the literature increasingly distinguishes several non-equivalent notions of intrinsic curvature defect—Ricci curvature, $k$-Ricci curvature, scalar-minus-sectional quantities, Chen’s $\delta$-invariants, and leaf-wise $\delta$-invariants—rather than treating “the” Chen inequality as a single invariant statement [1512.07647], [2605.19601], [2606.18725].

This suggests that the modern theory of Chen-Ricci inequalities is best understood as a flexible curvature comparison framework. Its central idea is stable—intrinsic curvature is constrained by extrinsic geometry—but its concrete realization depends strongly on the ambient structure, the type of geometric object under study, and the precise curvature invariant being estimated.

Source: https://www.emergentmind.com/topics/chen-ricci-inequalities