{
  "id": "resolvent-operator",
  "title": "Resolvent Operator",
  "lang": "en",
  "source_lang": "en",
  "summary": "The resolvent operator is the bounded inverse of a parameter-shifted linear operator, encoding its spectrum and supporting spectral analysis, differential equations, and perturbation theory.",
  "category": "mathematics",
  "subcategory": "analysis",
  "type": "concept",
  "aliases": [
    "Operator resolvent",
    "resolvent"
  ],
  "generated_at": "2026-11-06",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/resolvent-operator/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/resolvent-operator/",
      "json": "https://aiwiki.page/zh/resolvent-operator.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "The resolvent operator of a linear operator \\(A\\) is the parameter-dependent inverse\n\n\\[R(z,A)=(zI-A)^{-1},\\]\n\ndefined for those complex numbers \\(z\\) for which this inverse exists as a bounded operator on the underlying space. It converts the equation \\((zI-A)u=f\\) into \\(u=R(z,A)f\\), while its dependence on \\(z\\) reveals the spectral properties of \\(A\\). Resolvents are fundamental tools in functional analysis, connecting finite-dimensional matrix theory with differential operators, evolution equations, and quantum mechanics."
    },
    {
      "heading": "Definition and conventions",
      "text": "Let \\(X\\) be a complex Banach space, and let\n\n\\[A:D(A)\\subseteq X\\longrightarrow X\\]\n\nbe a closed, densely defined linear operator. Here \\(D(A)\\) is its domain, and \\(I\\) is the identity operator.\n\nThe resolvent set \\(\\rho(A)\\) consists of all \\(z\\in\\mathbb C\\) such that\n\n\\[zI-A:D(A)\\longrightarrow X\\]\n\nis bijective and its inverse is bounded as a map from \\(X\\) to \\(X\\). For such \\(z\\),\n\n\\[R(z,A):X\\longrightarrow X\\]\n\nis the resolvent operator. Its range is exactly \\(D(A)\\), and\n\n\\[(zI-A)R(z,A)=I\\quad\\text{on }X,\\]\n\n\\[R(z,A)(zI-A)=I\\quad\\text{on }D(A).\\]\n\nFor a closed operator, bijectivity already implies boundedness of the inverse by the closed graph theorem. Nevertheless, boundedness is included explicitly in the standard definition.\n\nThe spectrum is the complement\n\n\\[\\sigma(A)=\\mathbb C\\setminus\\rho(A).\\]\n\nThus the resolvent exists precisely outside the spectrum. In infinite dimensions, spectral points need not be eigenvalues: the shifted operator may instead fail to be surjective, or may have an inverse on its range that is not bounded.\n\nA second common convention defines the resolvent as \\((A-zI)^{-1}\\), the negative of \\(R(z,A)\\). Both conventions contain the same information, but signs in derivative, contour-integral, and resolvent identities must be adjusted consistently."
    },
    {
      "heading": "Matrices and elementary examples",
      "text": "For a complex \\(n\\times n\\) matrix \\(A\\),\n\n\\[R(z,A)=(zI-A)^{-1}\\]\n\nexists exactly when\n\n\\[\\det(zI-A)\\ne0.\\]\n\nIts entries are rational functions of \\(z\\), with possible poles at eigenvalues.\n\nIf \\(A\\) admits a diagonalization\n\n\\[A=VDV^{-1}, \\qquad D=\\operatorname{diag}(\\lambda_1,\\ldots,\\lambda_n),\\]\n\nthen\n\n\\[R(z,A) = V\\operatorname{diag}\\left( \\frac1{z-\\lambda_1},\\ldots,\\frac1{z-\\lambda_n} \\right)V^{-1}.\\]\n\nThis shows directly how the resolvent encodes both eigenvalues and their associated spectral subspaces.\n\nNontrivial Jordan blocks produce higher-order poles. For example, if\n\n\\[A= \\begin{pmatrix} \\lambda&1\\\\ 0&\\lambda \\end{pmatrix},\\]\n\nthen\n\n\\[R(z,A)= \\begin{pmatrix} (z-\\lambda)^{-1}&(z-\\lambda)^{-2}\\\\ 0&(z-\\lambda)^{-1} \\end{pmatrix}.\\]\n\nThe pole structure therefore contains information beyond the locations of eigenvalues."
    },
    {
      "heading": "Resolvent identities and analytic dependence",
      "text": "For \\(z,w\\in\\rho(A)\\), the first resolvent identity is\n\n\\[R(z,A)-R(w,A) = (w-z)R(z,A)R(w,A).\\]\n\nIn particular, resolvents of the same operator commute:\n\n\\[R(z,A)R(w,A)=R(w,A)R(z,A).\\]\n\nThe resolvent set is an open set, and \\(z\\mapsto R(z,A)\\) is an operator-valued analytic function on it. Its derivatives satisfy\n\n\\[\\frac{d^n}{dz^n}R(z,A) = (-1)^n n!\\,R(z,A)^{n+1}.\\]\n\nThese properties follow locally from a Neumann series. If \\(z_0\\in\\rho(A)\\) and\n\n\\[|z-z_0|\\,\\|R(z_0,A)\\|<1,\\]\n\nthen\n\n\\[R(z,A) = R(z_0,A) \\sum_{n=0}^{\\infty} \\left[-(z-z_0)R(z_0,A)\\right]^n,\\]\n\nwith convergence in operator norm.\n\nFor two bounded operators \\(A\\) and \\(B\\), the second resolvent identity is\n\n\\[R(z,A)-R(z,B) = R(z,A)(A-B)R(z,B),\\]\n\nwhenever \\(z\\in\\rho(A)\\cap\\rho(B)\\). The same formula holds for suitable unbounded operators, in particular when \\(A\\) and \\(B\\) have the same domain and their difference extends to a bounded operator on \\(X\\). It is a basic tool for studying how solutions and spectra change under perturbations."
    },
    {
      "heading": "Norm estimates and spectral information",
      "text": "If \\(A\\) is bounded, then for \\(|z|>\\|A\\|\\),\n\n\\[R(z,A) = \\frac1z\\left(I-\\frac Az\\right)^{-1} = \\sum_{n=0}^{\\infty}\\frac{A^n}{z^{n+1}},\\]\n\nand\n\n\\[\\|R(z,A)\\|\\le\\frac1{|z|-\\|A\\|}.\\]\n\nThis expansion is the operator analogue of a geometric series.\n\nFor an operator with nonempty spectrum,\n\n\\[\\|R(z,A)\\| \\ge \\frac1{\\operatorname{dist}(z,\\sigma(A))}.\\]\n\nThus the resolvent norm necessarily becomes large when a point approaches the spectrum.\n\nFor a normal operator on a Hilbert space, the spectral theorem gives the exact formula\n\n\\[\\|R(z,A)\\| = \\frac1{\\operatorname{dist}(z,\\sigma(A))}.\\]\n\nFor nonnormal operators, the norm can be much larger. Consequently, eigenvalues alone may give an incomplete account of sensitivity or transient behavior.\n\nThis distinction motivates the pseudospectrum. One common convention defines the \\(\\varepsilon\\)-pseudospectrum by\n\n\\[\\sigma_\\varepsilon(A) = \\sigma(A)\\cup \\left\\{ z\\in\\rho(A): \\|R(z,A)\\|>\\varepsilon^{-1} \\right\\}.\\]\n\nIn finite-dimensional problems, this describes spectral sensitivity to small perturbations. If \\(E\\) is bounded and\n\n\\[\\|E R(z,A)\\|<1,\\]\n\nthen \\(z\\in\\rho(A+E)\\). A large resolvent norm therefore indicates that comparatively small perturbations may destroy invertibility."
    },
    {
      "heading": "Spectral projections and functions of operators",
      "text": "The resolvent permits constructions modeled on the Cauchy integral formula from complex analysis. For a bounded operator \\(A\\) and a function \\(f\\) holomorphic on a neighborhood of \\(\\sigma(A)\\),\n\n\\[f(A) = \\frac{1}{2\\pi i} \\int_\\Gamma f(z)R(z,A)\\,dz,\\]\n\nwhere \\(\\Gamma\\) is an appropriate positively oriented contour, or collection of contours, in \\(\\rho(A)\\) enclosing the spectrum. This is the holomorphic functional calculus.\n\nIf a contour encloses an isolated part of the spectrum and excludes the remainder, then\n\n\\[P= \\frac{1}{2\\pi i} \\int_\\Gamma R(z,A)\\,dz\\]\n\nis a Riesz projection. It satisfies \\(P^2=P\\), commutes with \\(A\\) in the appropriate domain sense, and separates the corresponding invariant spectral subspace.\n\nFor an isolated matrix eigenvalue \\(\\lambda\\), this projection selects the generalized eigenspace. The order of the resolvent pole at \\(\\lambda\\) equals the size of the largest Jordan block associated with that eigenvalue. For general infinite-dimensional operators, however, an isolated spectral point need not be a pole of the resolvent."
    },
    {
      "heading": "Self-adjoint operators and compact resolvent",
      "text": "For a self-adjoint operator \\(A\\), the spectrum is real, so every nonreal \\(z\\) belongs to \\(\\rho(A)\\). In particular,\n\n\\[\\|R(z,A)\\| \\le \\frac1{|\\operatorname{Im}z|}.\\]\n\nIf \\(E\\) is the projection-valued spectral measure of \\(A\\), then\n\n\\[R(z,A) = \\int_{\\sigma(A)} \\frac1{z-\\lambda}\\,dE(\\lambda).\\]\n\nThis representation connects resolvent boundary values near the real axis with spectral measures. It is especially important for Hamiltonians in quantum mechanics.\n\nAn operator has compact resolvent if its resolvent set is nonempty and \\(R(z_0,A)\\) is a compact operator for some \\(z_0\\in\\rho(A)\\). The resolvent identity then implies compactness of \\(R(z,A)\\) for every \\(z\\in\\rho(A)\\).\n\nFor such an operator, every spectral point is an isolated eigenvalue of finite algebraic multiplicity, and there is no finite accumulation point of the spectrum. This property occurs for many elliptic operators on bounded domains with appropriate boundary conditions. It explains why these infinite-dimensional problems can possess discrete spectral expansions resembling matrix decompositions."
    },
    {
      "heading": "Differential equations and Green's functions",
      "text": "For a differential operator \\(A\\), solving\n\n\\[(zI-A)u=f\\]\n\nis a parameter-dependent boundary-value problem. When the resolvent admits an integral-kernel representation,\n\n\\[(R(z,A)f)(x) = \\int G_z(x,y)f(y)\\,dy,\\]\n\nthe kernel \\(G_z\\) is a Green's function for the shifted operator with the specified boundary conditions.\n\nThe domain of \\(A\\) is essential: the same differential expression with different boundary conditions generally defines different operators, spectra, and resolvents. For example, Dirichlet and Neumann realizations of the Laplacian on the same bounded region have different spectral properties.\n\nResolvents also connect stationary equations with evolution equations. If \\(A\\) generates a strongly continuous operator semigroup \\(T(t)\\) satisfying\n\n\\[\\|T(t)\\|\\le Me^{\\omega t},\\]\n\nthen, for \\(\\operatorname{Re}z>\\omega\\),\n\n\\[R(z,A)x = \\int_0^\\infty e^{-zt}T(t)x\\,dt.\\]\n\nThus the resolvent is the Laplace transform of the evolution operators. This relation underlies semigroup approaches to the heat equation and other linear evolution problems."
    },
    {
      "heading": "Computational interpretation",
      "text": "In computation, applying \\(R(z,A)\\) to a vector usually means solving the linear system\n\n\\[(zI-A)u=f,\\]\n\nnot explicitly forming the inverse matrix. Factorizations and iterative methods can exploit sparsity or differential-operator structure.\n\nThe resolvent norm measures the worst-case amplification of perturbations in \\(f\\). The relative condition number of the shifted system additionally depends on \\(\\|zI-A\\|\\). Near spectral points, or in nonnormal regions with large resolvent norm, accurate solution may be difficult even when the shift is not especially close to an eigenvalue.\n\nA related but distinct use occurs in nonlinear analysis: for a possibly nonlinear or set-valued operator \\(T\\), the expression\n\n\\[J_{\\lambda T}=(I+\\lambda T)^{-1}\\]\n\nis also called a resolvent. Its single-valuedness and regularity require additional assumptions, such as maximal monotonicity; they do not follow from the linear spectral theory described above."
    }
  ],
  "links": [
    "functional-analysis",
    "banach-space",
    "linear-map",
    "eigenvalues-and-eigenvectors",
    "matrix",
    "matrix-diagonalization",
    "open-set",
    "analytic-function",
    "geometric-series",
    "hilbert-space",
    "spectral-theorem",
    "complex-analysis",
    "self-adjoint-operator",
    "quantum-mechanics",
    "boundary-value-problem",
    "greens-function",
    "laplace-operator",
    "laplace-transform",
    "heat-equation",
    "system-of-linear-equations",
    "condition-number"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}