On normal forms of gradient Ricci 4-solitons

Amir Babak Aazami · 2026

Read On normal forms of gradient Ricci 4-solitons by Amir Babak Aazami (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-03894.txt.

Introduction

Amir Babak Aazami analyzed the normal form of a curvature-related operator for gradient Ricci 4-solitons in the work of Cao and Tran. He showed that the curvature operator of the Koiso-Cao soliton inherits that normal form. Work by Johnson then yields a normal form for that operator relative to algebraic Kähler curvature operators. The note is short and technical, pinning down the algebraic shape of a key example.

Catalog record

CREATORAmir Babak Aazami
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE29817
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.03894

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

On normal forms of gradient Ricci 4-solitons Amir Babak Aazami Abstract . In this note we analyze the normal form of the operator R ^ + 1 2 ​ H ^ \hat{R}+\frac{1}{2}\hat{H} of a gradient Ricci 4-soliton in [ 5 ] . In particular, we show that the curvature operator R ^ \hat{R} of the Koiso–Cao soliton inherits this normal form. By [ 10 ] , this yields a normal form for the curvature operator of the Koiso–Cao soliton relative to the space of algebraic Kähler curvature operators. 1. Introduction The concept of a normal form for a Riemannian manifold [ 2 , 14 ] is motivated by a well known fact from linear algebra, namely, that on an inner product space (V,\langle{\,},{}\rangle) , a self-adjoint linear map T\colon V\longrightarrow V is completely determined by the critical points and values of its quadratic form v\mapsto\langle{Tv},{v}\rangle restricted to the unit sphere. The relevance of this fact for Riemannian geometry is that the quadratic form of the curvature operator \hat{R}\colon\Lambda^{2}\longrightarrow\Lambda^{2} of a Riemannian manifold (M,g) is (up to sign) precisely the sectional curvature: For any orthonormal pair x,y\in T_{p}M , x\wedge y\mapsto\langle{\hat{R}(x\wedge y)},{x\wedge y}\rangle=\underbrace{\,\text{Rm}_{\scalebox{0.6}{$g$}}(x,y,x,y)\,}_{\text{$-\text{sec}_{\scalebox{0.6}{$g$}}(x\wedge y)$}}. Because the sectional curvature \text{sec}_{\scalebox{0.6}{$g$}} determines the curvature tensor \text{Rm}_{\scalebox{0.6}{$g$}} , a natural question arises: When is \hat{R} determined by just the critical points and values of \text{sec}_{\scalebox{0.6}{$g$}} (relative to some class of algebraic curvature operators)? Here the latter is to be thought of as a smooth function \text{sec}_{\scalebox{0.6}{$g$}}\colon\text{Gr}_{2}(T_{p}M)\longrightarrow\mathbb{R} on the Grassmannian of 2-planes at T_{p}M . If so, then we say that the curvature operator has a “normal form” at p\in M ; this is a pointwise condition. Space forms, locally conformally flat manifolds, any Riemannian submanifold of a space form with a flat normal bundle — all of these have normal forms. By [ 10 ] , so do Kähler 4-manifolds with positive-definite Ricci tensor (relative to the space of algebraic Kähler curvature operators). But the most famous instances of normal forms are the class of Einstein curvature operators in dimension 4, because here the normal form is derived from the fact that \hat{R} commutes with the Hodge star operator * : \text{Ric}=\lambda g\iff*\hat{R}=\hat{R}*. Hence \hat{R},* have a common orthonormal basis of eigenvectors. This fact, together with the splitting of \Lambda^{2}=\Lambda^{+}\oplus\Lambda^{-} into the self-dual (\Lambda^{+}) and anti-self-dual (\Lambda^{-}) eigenspaces of * , and the fact that a 2-plane P is a critical point of \text{sec}_{\scalebox{0.6}{$g$}} if and only if \hat{R}P\in\text{span}\{P,*P\} , was used to derive an orthonormal basis \{e_{1},e_{2},e_{3},e_{4}\}\subseteq T_{p}M all of whose corresponding 2-planes e_{i}\wedge e_{j} are critical points of \text{sec}_{\scalebox{0.6}{$g$}} , with respect to which (1) \displaystyle\hat{R}=\begin{bmatrix}A_{0}&B_{0}\\ B_{0}&A_{0}\end{bmatrix}\hskip 7.22743pt,\hskip 7.22743ptA_{0}\mathrel{\hbox to0.0pt{\raisebox{1.07639pt}{$\cdot$}\hss}\raisebox{-1.07639pt}{$\cdot$}}=\text{diag}(\lambda_{1},\lambda_{2},\lambda_{3})\hskip 7.22743pt,\hskip 7.22743ptB_{0}\mathrel{\hbox to0.0pt{\raisebox{1.07639pt}{$\cdot$}\hss}\raisebox{-1.07639pt}{$\cdot$}}=\text{diag}(\mu_{1},\mu_{2},\mu_{3}). Here the \lambda_{i} are critical values of \text{sec}_{\scalebox{0.6}{$g$}} , and the \mu_{i} are not critical values but can be derived from them (with \mu_{1}+\mu_{2}+\mu_{3}=0 by the algebraic Bianchi identity). Furthermore, two Einstein operators with the same critical points and values are necessarily pointwise isometric; hence, within the Einstein class, the operator is determined up to orthogonal change by this data. The purpose of this note is to analyze another very interesting and recent example of a normal form in dimension 4. Namely, [ 5 ] showed that a normal form also exists for gradient Ricci 4-solitons , (2) \displaystyle\text{Ric}+\text{Hess}f=\lambda g, although not for the curvature operator \hat{R} itself, but rather for the modified operator \hat{S}\mathrel{\hbox to0.0pt{\raisebox{1.07639pt}{$\cdot$}\hss}\raisebox{-1.07639pt}{$\cdot$}}=\hat{R}+\frac{1}{2}\hat{H}\colon\Lambda^{2}\longrightarrow\Lambda^{2} , where H\mathrel{\hbox to0.0pt{\raisebox{1.07639pt}{$\cdot$}\hss}\raisebox{-1.07639pt}{$\cdot$}}=\text{Hess}f\raisebox{0.39993pt}{\,${\scriptstyle\mathchoice{\mathbin{\vtop{\halign{#\cr$\displaystyle\bigcirc$\cr$\displaystyle\wedge$\cr}}}{}}{\mathbin{\vtop{\halign{#\cr$\textstyle\bigcirc$\cr$\textstyle\wedge$\cr}}}{}}{\mathbin{\vtop{\halign{#\cr$\scriptstyle\bigcirc$\cr$\scriptstyle\wedge$\cr}}}{}}{\mathbin{\vtop{\halign{#\cr$\scriptscriptstyle\bigcirc$\cr$\scriptscriptstyle\wedge$\cr}}}{}}}$\,}g is given by \displaystyle\langle{\hat{H}(e_{i}\wedge e_{j})},{e_{k}\wedge e_{l}}\rangle\!\! \displaystyle\mathrel{\hbox to0.0pt{\raisebox{1.07639pt}{$\cdot$}\hss}\raisebox{-1.07639pt}{$\cdot$}}= \displaystyle\!\!H(e_{i},e_{j},e_{k},e_{l}) \displaystyle= \displaystyle\!\!\text{Hess}f(e_{i},e_{l})\,g_{jk}+\text{Hess}f(e_{j},e_{k})\,g_{il} \displaystyle\hskip 14.45377pt-\,\text{Hess}f(e_{i},e_{k})\,g_{jl}-\text{Hess}f(e_{j},e_{l})\,g_{ik}. In [ 5 , Remark 2.6] , it was asked whether \hat{R} inherits the normal form of \hat{S} . We show that in at least one instance, the answer is yes: Theorem . The curvature operator of the Koiso–Cao soliton on \mathbb{CP}^{2}\#(-\mathbb{CP}^{2}) inherits the normal form of \hat{S} . As a consequence, the curvature operator of the Koiso–Cao soliton has a normal form relative to the space of algebraic Kähler curvature operators. The Koiso–Cao soliton [ 11 , 4 ] will be analyzed in Theorem 1 and Corollary 3 in Section 4 below. 2. Overview of the normal form of \hat{S}=\hat{R}+\frac{1}{2}\hat{H} First, let \{e_{1},e_{2},e_{3},e_{4}\}=\{\partial_{1},\partial_{2},\partial_{3},\partial_{4}\} arise from normal coordinates centered at p\in M , so that \text{Hess}f(e_{i},e_{j})\big|_{p}=f_{ij}\big|_{p} and hence (3) \displaystyle H_{ijkl}\big|_{p}=f_{il}\delta_{jk}+f_{jk}\delta_{il}-f_{ik}\delta_{jl}-f_{jl}\delta_{ik}\,\big|_{p}. What makes the normal form of \hat{S} particularly interesting is that it is “Einstein-like,” namely, the normal form arises from the fact that \hat{S} commutes with * , *\hat{S}=\hat{S}*, and so the normal form basis \{e_{i}\wedge e_{j}\}\subseteq\Lambda^{2}(T_{p}M) of critical 2-planes is derived just as it was for \hat{R} in [ 2 , 14 ] in the Einstein case (once again, a 2-plane P will be a critical point of this quadratic form if and only if \hat{S}P\in\text{span}\{P,*P\} , by the same proof as in [ 14 ] ). Of course, the quadratic form here is not the sectional curvature \text{sec}_{\scalebox{0.6}{$g$}} , but rather that of \hat{S} : \displaystyle e_{i}\wedge e_{j}\mapsto\langle{\hat{S}(e_{i}\wedge e_{j})},{e_{i}\wedge e_{j}}\rangle\!\! \displaystyle= \displaystyle\!\!\langle{\hat{R}(e_{i}\wedge e_{j})},{e_{i}\wedge e_{j}}\rangle+\frac{1}{2}\langle{\hat{H}(e_{i}\wedge e_{j})},{e_{i}\wedge e_{j}}\rangle \displaystyle= \displaystyle\!\!\text{sec}_{\scalebox{0.6}{$g$}}(e_{i}\wedge e_{j})-f_{ik}\delta_{jl}-f_{jl}\delta_{ik}. We would like to investigate what \hat{S} ’s normal form says about \hat{R} . To do so, let us use the Hodge basis \{\xi_{1}^{+},\xi_{2}^{+},\xi_{3}^{+},\xi_{1}^{-},\xi_{2}^{-},\xi_{3}^{-}\} , where \Big\{\underbrace{\,\frac{1}{\sqrt{2}}(e_{1}\wedge e_{2}\pm e_{3}\wedge e_{4})\,}_{\text{$\xi_{1}^{\pm}$}},\underbrace{\,\frac{1}{\sqrt{2}}(e_{1}\wedge e_{3}\pm e_{4}\wedge e_{2})\,}_{\text{$\xi_{2}^{\pm}$}},\underbrace{\,\frac{1}{\sqrt{2}}(e_{1}\wedge e_{4}\pm e_{2}\wedge e_{3})\,}_{\text{$\xi_{3}^{\pm}$}}\Big\} comprise the self-dual and anti-self-dual eigenvectors of * ( *\xi_{i}^{\pm}=\pm\xi_{i}^{\pm} ) . With respect to this basis, \hat{R}=\begin{bmatrix}\hat{W}^{+}+\frac{\text{scal}}{12}I&K\\ K^{t}&\hat{W}^{-}+\frac{\text{scal}}{12}I\end{bmatrix}\hskip 14.45377pt,\hskip 14.45377pt\hat{H}=\begin{bmatrix}-\frac{\Delta f}{2}I&C\\ C^{t}&-\frac{\Delta f}{2}I\end{bmatrix}, where I is the 3\times 3 identity matrix, scal is the scalar curvature of g at p , \hat{W}^{\pm} are the self-dual and anti-self dual portions of the Weyl curvature operator \hat{W} that leave invariant the \pm 1 -eigenspaces of * , and -\frac{1}{2}K=C\mathrel{\hbox to0.0pt{\raisebox{1.07639pt}{$\cdot$}\hss}\raisebox{-1.07639pt}{$\cdot$}}=\begin{bmatrix}\frac{-f_{11}-f_{22}+f_{33}+f_{44}}{2}&f_{14}-f_{23}&-f_{13}-f_{24}\\ -f_{14}-f_{23}&\frac{-f_{11}+f_{22}-f_{33}+f_{44}}{2}&f_{12}-f_{34}\\ f_{13}-f_{24}&-f_{12}-f_{34}&\frac{-f_{11}+f_{22}+f_{33}-f_{44}}{2}\end{bmatrix}\cdot (See [ 5 , Lemma 2.3] , though note that in the latter the opposite sign convention for {\scriptstyle\mathchoice{\mathbin{\vtop{\halign{#\cr$\displaystyle\bigcirc$\cr$\displaystyle\wedge$\cr}}}{}}{\mathbin{\vtop{\halign{#\cr$\textstyle\bigcirc$\cr$\textstyle\wedge$\cr}}}{}}{\mathbin{\vtop{\halign{#\cr$\scriptstyle\bigcirc$\cr$\scriptstyle\wedge$\cr}}}{}}{\mathbin{\vtop{\halign{#\cr$\scriptscriptstyle\bigcirc$\cr$\scriptscriptstyle\wedge$\cr}}}{}}} is being used.) When \hat{S}=\hat{R}+\frac{1}{2}\hat{H} is expressed with respect to this Hodge basis, we have (4) \displaystyle\hat{S}=\begin{bmatrix}A+B&O\\ O&A-B\end{bmatrix}\hskip 14.45377pt\Rightarrow\hskip 14.45377pt\hat{R}=\begin{bmatrix}A+B+\frac{\Delta f}{4}I&-\frac{1}{2}C\\ -\frac{1}{2}C^{t}&A-B+\frac{\Delta f}{4}I\end{bmatrix}, where A\mathrel{\hbox to0.0pt{\raisebox{1.07639pt}{$\cdot$}\hss}\raisebox{-1.07639pt}{$\cdot$}}=\text{diag}(a_{1},a_{2},a_{3}) and B\mathrel{\hbox to0.0pt{\raisebox{1.07639pt}{$\cdot$}\hss}\raisebox{-1.07639pt}{$\cdot$}}=\text{diag}(b_{1},b_{2},b_{3}) . (If \hat{S} had been expressed with respect to \{e_{i}\wedge e_{j}\} directly, then \hat{S}=\begin{bmatrix}A&B\\ B&A\end{bmatrix} , similarly to ( 1 ).) With these preliminaries established, we now explore some properties of this \hat{S} -normal form basis \{e_{i}\wedge e_{j}\} and its relation to \hat{R} . 3. Properties of the normal form of \hat{S}=\hat{R}+\frac{1}{2}\hat{H} Unfortunately, the existence of \{e_{i}\wedge e_{j}\} does not in general yield any topological information that was not otherwise known directly from the Chern-Gauss-Bonnet formula for \chi(M) in dimension 4 (even though a normal form-argument was used in deriving the Hitchin-Thorpe inequality in [ 9 ] , it was not essential; see Section 5 below). Having said that, here is one case in which \hat{S} ’s normal form basis does yield topological information: Proposition 1 . If the normal form \{e_{i}\wedge e_{j}\} for \hat{S} is “pure”, meaning that the e_{i}\wedge e_{j} are eigenvectors of \hat{S} , then \tau(M)=0 . Proof. As we saw above, the normal form basis \{e_{i}\wedge e_{j}\} yields \hat{S}=\begin{bmatrix}A&B\\ B&A\end{bmatrix}\hskip 14.45377pt,\hskip 14.45377ptA=\text{diag}(a_{1},a_{2},a_{3})\hskip 14.45377pt,\hskip 14.45377ptB=\text{diag}(b_{1},b_{2},b_{3}). In terms of these, the signature is \tau(M)=\frac{1}{3\pi^{2}}\int_{M}\sum_{i=1}^{3}b_{i}\Big(a_{i}+\frac{\Delta f}{4}-\frac{\text{scal}}{12}\Big)dV_{\scalebox{0.6}{$g$}}. (For the derivation of this formula, see ( 10 ) below.) But if the e_{i}\wedge e_{j} are eigenvectors of \hat{S} , then b_{1}=b_{2}=b_{3}=0 . ∎ (The Koiso–Cao soliton on \mathbb{CP}^{2}\#(-\mathbb{CP}^{2}) has \tau=0 .) Rather, the more interesting question is whether the normal form for \hat{S} is sufficient to yield a normal form

Read the complete text