Heat Flows with Prescribed Singularities from 3-dimensional Manifold

Jie Ji; Jingru Niu

REVIEW 2 major objections 6 minor 34 references

Heat Flows with Prescribed Singularities from 3-dimensional Manifold

T0 review · 2 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper proves that a 3D singular heat flow into hyperbolic space with prescribed curve singularity exists for all time, is regular in weighted Hölder spaces, and converges exponentially to a singular harmonic map.

desk verdict New parabolic extension of Li-Tian with two real proof gaps: invalid Poincaré step in Lemma 4.2 and short-time existence obstruction in Theorem 3.1; deserves revision, not publication as-is. read the letter →

arxiv 2412.14634 v1 pith:N3EHUAXK submitted 2024-12-19 math.AP

Jie Ji , Jingru Niu This is my paper · ORCID

classification math.AP MSC 80A1958E2035A2158J35

keywords singularheatflowprescribedsingularitieshyperbolicspaceharmonicmapweightedHölderestimatesexponentialconvergence3-dimensionalmanifold

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper treats the parabolic counterpart of harmonic maps with prescribed singularities: on a bounded 3-manifold, a map into hyperbolic space is forced to have a prescribed logarithmic singularity along a closed curve $\Gamma$. The authors claim that, for any admissible initial data, the singular heat flow exists for all time, gains weighted Hölder regularity up to an order determined by $\alpha$, and converges to a singular harmonic map at an exponential rate. The result matters because such maps arise from stationary axially symmetric Einstein equations, where the axis is the singularity curve, and the heat flow offers a canonical route from prescribed data to the stationary solution. The proof builds a weighted Schauder theory for the linearized parabolic system and then uses energy decay to extract the stationary limit.

What carries the argument

The carrying object is a weighted parabolic regularity theory adapted to the distance function $\rho(x)=\operatorname{dist}(x,\Gamma)$. The paper defines weighted Sobolev and Hölder spaces with weights $\rho^{-\alpha}$ and $\rho^{-\gamma}$, proves that the linearized operator $DP(\phi^0,\cdot)$ is an isomorphism $A\times B\to C\times D$ via Galerkin approximations and weighted Schauder estimates, then uses the inverse function theorem for short-time existence and barrier, Campanato, and Bochner-type energy arguments for global decay. The exponential convergence is driven by the estimate $\frac{d}{dt}\int_M \theta^2\,dx\leq -C_0\int_M\theta^2\,dx$ for $\theta=h^{-2\alpha}e^{-2\phi_2}|\partial_t\phi_1|^2+|\partial_t\phi_2|^2$, which is meant to produce the rate $C_0/4$.

What would settle it

Solve (1.2) numerically from a smooth admissible initial datum on a simple 3-manifold with $\Gamma$ an axis, and measure $\int_M \theta^2\,dx(t)$ with $\theta=h^{-2\alpha}e^{-2\phi_2}|\partial_t\phi_1|^2+|\partial_t\phi_2|^2$; the theorem predicts decay like $e^{-C_0 t}$, while the zero-mean issue in Lemma 4.2 predicts a possible nonzero plateau unless an additional argument removes the mean. If the measured decay is not exponential with the spectral-gap constant, the central claim as proven fails; if it is, the theorem survives despite the gap in the written proof.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.1: for any $0<T\leq\infty$ and any initial pair $(\phi^0_1,\phi^0_2)$ in the weighted Hölder spaces $A_0\times B_0$, the system (1.2) has a solution on $M\times[0,T]$; for every $0<\varepsilon<\min\{\frac12,\alpha-1\}$ the solution lies in $C^{k+\lambda,(k+\lambda)/2}(M\times[0,T])$ with $k=[2\alpha-2\varepsilon]$ and $\lambda=2\alpha-k-2\varepsilon$; and as $t\to+\infty$ the solution converges in the weighted $C^2_*(M)$ norm to a limit $\bar\phi$ satisfying the singular harmonic map system, with the weighted error bounded by $C e^{-(C_0/4)t}$ for $t\geq1$. The limit itself belongs to $C^{k,\lambda}(M)$ for any $0<\varepsilon<2\alpha$. In other words, the heat flow exists globally, regularizes singular initial data, and relaxes exponentially into the stationary singular harmonic map.

Load-bearing premise

The exponential-convergence argument in Lemma 4.2 applies the Poincaré inequality to the squared time derivative of the map as though that quantity had zero spatial average; since it is nonnegative and its average is generally not zero, the decay step is not justified as written.

Editorial extensions

If this is right

Any admissible initial data produces a global solution, so no finite-time singularity develops along $\Gamma$ during the flow.
The limit map is a singular harmonic map with the same prescribed singularity, giving a parabolic construction of such stationary solutions.
Convergence holds with an explicit exponential rate in the weighted $C^2_*$ norm, so the stationary regime is reached exponentially fast rather than only asymptotically.
The available regularity is tied to $\alpha$: the Hölder exponent is $[2\alpha-2\varepsilon]$, so larger $\alpha$ yields smoother behavior near $\Gamma$.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A testable consequence of the proof structure is that the exponential rate is controlled by the Poincaré constant of $M$; on manifolds whose spectral gap $C_0$ tends to zero, the predicted rate should deteriorate accordingly.
Because the weighted spaces $A_0\times B_0$ require initial regularity higher than that of the solution itself, reapplying the short-time existence argument from a later time slice needs the relaxed spaces introduced in Theorem 4.13; a different proof would be needed for very rough initial data.
For radially symmetric data with $\phi_1=0$, the system reduces to a scalar heat equation for $\phi_2$; that reduced case is a concrete laboratory for testing the claimed exponential decay numerically.
The paper itself notes that the Sobolev embedding and monotonicity-formula steps are 3-dimensional, so extending the result to higher dimensions would require replacing those two ingredients.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

This is a serious attempt to extend Li-Tian's elliptic theory of prescribed-singularity harmonic maps into H^2 to the parabolic heat-flow setting. The problem is natural, the result — existence, weighted regularity, and exponential convergence of singular heat flows along a closed curve in a 3-manifold — is genuinely new, and the technical scaffolding is appropriate: weighted Schauder estimates, Galerkin approximation, inverse function theorem, Campanato-style decay. Citations to Li-Tian, Weinstein, and Han-Khuri-Weinstein-Xiong are on point, with no self-citation padding. If the proof were correct, this would be a meaningful contribution.

The problem is that two load-bearing steps are wrong as written. Lemma 4.2 applies the Poincaré inequality to θ = h^{-2α}e^{-2φ2}|∂tφ1|^2 + |∂tφ2|^2, which is nonnegative and generally has nonzero spatial mean. The asserted inequality -2∫|∇θ|^2 ≤ -C0∫θ^2 is false for such θ; Poincaré gives the analogous bound for θ minus its mean. At t=0, ∂tφ(0) is pinned down by φ0 and is not zero, so the mean is positive and the claimed exponential decay of ∫θ^2 fails. That decay drives the pointwise bounds in Lemma 4.2 and feeds Proposition 4.6, Theorem 4.13, Lemma 5.1, and Theorem 5.2, so the advertised convergence rate is not established.

Second, in Theorem 3.1 the function p is set to 0 for t ≤ δ and to P(φ0) for t ≥ 2δ. The claim that p lies in a small neighborhood W of P(φ0) in C×D is not true: those spaces measure values for all t, including near t=0, where p-P(φ0) equals -P(φ0), which is not small. The short-time existence step is therefore also open. Both gaps look repairable — one could try to control the mean of θ separately, and one could replace p by a cutoff that actually converges in the relevant norm — but as written the proof is incomplete.

The manuscript also has drafting issues: LaTeX artifacts and a stray meta-comment ("these definitions are somewhat vague without more context") that suggest an unpolished early draft. These are minor relative to the mathematical gaps.

Bottom line: the paper deserves serious referee time because the problem and strategy are legitimate, but it needs major revision before it can be believed. I would not cite it in its current form.

Referee Report

2 major / 6 minor

Summary. The paper studies a parabolic analogue of the Li--Tian prescribed-singularity problem: for a compact 3-manifold M (without boundary), a closed curve Γ, and α>1, it considers the system (1.2) for (φ1,φ2) with initial data in weighted Hölder spaces A0×B0 and with φ1=0 on Γ. Theorem 1.1 asserts global existence and regularity in weighted parabolic Hölder spaces and exponential convergence in C^2_* to a solution of the singular harmonic map system. The proof is organized as follows: Section 2 develops a linear theory in weighted Sobolev/Hölder spaces via Galerkin approximation and weighted Schauder estimates; Section 3 uses the inverse function theorem to obtain short-time existence; Section 4 derives uniform energy and monotonicity estimates and proves long-time existence; Section 5 proves exponential convergence. The paper relies on results of Li--Tian, Weinstein, and Han--Khuri--Weinstein--Xiong.

Significance. If correct, the result would be a useful parabolic extension of the elliptic theory of harmonic maps with prescribed singularities and would provide quantitative convergence to a singular harmonic map, complementing the qualitative convergence results for harmonic map heat flow into nonpositively curved targets. The manuscript is proof-based and contains no fitted parameters; it engages seriously with weighted Poincaré and Schauder estimates and cites the relevant independent literature. However, the central proof contains two load-bearing gaps, one in the exponential decay estimate (Lemma 4.2) and one in the short-time existence argument (Theorem 3.1). Because these gaps occur in the main logical chain and are not mere presentational issues, the advertised theorem is not established in the present form.

major comments (2)

[Section 4, Lemma 4.2, Eq. (4.3)] The exponential decay step is invalid. From (∂t−Δ)θ≤0 the authors write d/dt∫θ² ≤ 2∫θΔθ = −2∫|∇θ|² and then apply Poincaré as −2∫|∇θ|² ≤ −C0∫θ². The Poincaré inequality on the closed manifold M applies to zero-mean functions and gives −2∫|∇θ|² ≤ −C0∫(θ−θ̄)². The function θ = h^{−2α}e^{−2φ2}|∂tφ1|²+|∂tφ2|² is nonnegative and is not zero mean for generic initial data; for example, a positive constant has zero gradient and positive L² norm. At t=0, θ(0) is determined by φ0 through the equations in (1.2) and is generally nonzero. Consequently the asserted inequality d/dt∫θ² ≤ −C0∫θ², and therefore the bounds ∫θ ≤ Ce^{−C0t/2} and ρ^{3/2−α}|∂tφ1|+ρ^{3/2}|∂tφ2| ≤ Ce^{−C0t/4}, are not established. These bounds feed Proposition 4.6, Lemma 5.1, and Theorem 5.2, and the exponential rate in Theorem 1.1 rests on them. Replacing θ by θ−θ̄ only controls the variance and does not yield exponential decay of ∫θ without an independent bound on θ̄, which is the desired conclusion.
[Section 3, Theorem 3.1, Eq. (3.1)] The inverse function theorem argument is not valid as stated. The theorem supplies a neighborhood W of P(φ0) in C×D. The function p defined by p=0 for 0≤t≤δ and p=P(φ0) for 2δ≤t≤T is claimed to lie in W for sufficiently small δ. But on (0,δ), p−P(φ0) equals −P(φ0), and the C^{β,β/2}(QT;ρ^{−γ}) component of the norm of this difference is sup_{0<t<δ} ρ^{2−γ}|P(φ0)(x,t)|, which does not tend to zero as δ→0 and is positive for generic φ0 (for instance P1(φ0) is generally not zero when φ0 is not already a singular harmonic map). Hence p is not close to P(φ0) in the norm that defines W. The conclusion that there is a unique φb∈V with P(φb+φ0)=0 on M×[0,δ) therefore does not follow. Since Theorem 4.13 uses this short-time existence step to extend the solution, the existence part of Theorem 1.1 is also not established.

minor comments (6)

[Definition 2.1] Definition 2.1 contains a sentence 'Regarding the additional notation ρ_X = ρ(x) and ρ_{X,Y} = max{ρ_X, ρ_Y}, these definitions are somewhat vague without more context.' This editorial note should be replaced by actual definitions of ρ_X and ρ_{X,Y}, and the note removed.
[Throughout] There are numerous LaTeX rendering artifacts in the text, including 'l /greaterorequalslant1', 't /greaterorequalslant1', and 'k /greaterorequalslant1'; these should be rendered as mathematical symbols.
[Lemma 4.2] The notation '∫_M ... dx(t)' is nonstandard; write dV or dx and evaluate the integrand at time t.
[Theorem 5.2, Eq. (5.13)] The intermediate limiting system (5.13) appears to mix the two equations: after multiplying the first equation of (1.2) by ρ^{7/2−α} and the second by ρ^{7/2}, the limit should involve ρ^{7/2−α}(−Δφ1+2(∇φ2+α∇h/h)∇φ1)=0 and ρ^{7/2}(−Δφ2−h^{−2α}e^{−2φ2}|∇φ1|²)=0. As written, (5.13) has the wrong exponents and terms; this should be corrected for clarity.
[Theorem 5.2] The notation for the weighted convergence norm switches between C^2_* and C^{2+α}_* in Theorem 5.2 and its proof; please define and use one norm consistently.
[Theorems 1.1 and 4.13] The statement 'M×[0,T]' for T=∞ should be interpreted as M×[0,∞); this should be stated explicitly.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the paper's existence and convergence proof rests on independent external results and contains no fitted inputs, self-citation chains, or inputs renamed as predictions.

full rationale

This is a proof-based analysis paper. Its central claims—existence of singular heat flows and exponential convergence to a singular harmonic map—are established by constructing solutions through Galerkin approximations, the inverse function theorem, a priori estimates, and convergence arguments. The external inputs are standard PDE/geometric tools and prior independent works by Li-Tian, Weinstein, Han-Khuri-Weinstein-Xiong, and others; none of these citations are authored by the present authors, and none already contain Theorem 1.1. The exponential convergence rate is not a fitted parameter but is supposedly derived from a decay estimate; even if the Poincaré inequality application in Lemma 4.2 is questionable because θ need not have zero mean, that is a correctness concern, not a circular one, since the inequality is not assumed to be the conclusion. No quantity is defined in terms of the result it predicts, no fitted input is relabeled as a prediction, and the cited regularity and uniqueness results are not used to forbid alternatives in a way that smuggles in the target theorem. Therefore the circularity score is 0.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claim rests on standard PDE regularity theory, the external elliptic theory of Li-Tian, and geometric properties of hyperbolic space. No new entities are postulated, and no parameters are fitted to data. The main unstated assumption is the validity of the Poincare step in Lemma 4.2, which is flagged in the red flags.

assumptions (5)

standard math Weighted Hardy and Poincare inequalities from Corollary 4.1 of Li-Tian [21]
Used in Lemma 2.6 and elsewhere to control the weighted L2 norms of k1 and its gradient.
standard math Regularity theory for singular harmonic maps from Theorem 1.1 of Li-Tian [21]
Invoked to assert the regularity of the limiting map phi_bar in Theorem 5.2.
standard math Standard elliptic and parabolic regularity estimates, including Sobolev embeddings and Schauder estimates
Used throughout Sections 2, 4, and 5 without proof.
domain assumption Existence of a positive solution h=rho e^u to Delta log h = 0 on M\Gamma with prescribed log growth
This is the prescribed singularity structure taken as given from Li-Tian [21], defining the ansatz (phi1, h^alpha e^{phi2}).
domain assumption Nonpositive curvature of the target H^2 and the Bochner identity
Used in Lemma 4.1 and Lemma 4.2 to derive bounds on the distance and the time derivative.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Heat Flows with Prescribed Singularities from 3-dimensional Manifold." pith.science (2026). https://pith.science/paper/N3EHUAXK

@misc{pith2026241214634,
  author       = {Pith},
  title        = {Pith review of: Heat Flows with Prescribed Singularities from 3-dimensional Manifold},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/N3EHUAXK}},
  note         = {Machine review of arXiv:2412.14634}
}

read the original abstract

In this paper, we study singular heat flows from a 3-dimensional complete bounded Riemannian manifold without boundary into the hyperbolic space with prescribe singularity along a closed curve. We prove the existence and regularity of the singular heat flows. Furthermore, we prove that the singular heat flows converge to a singular harmonic map at an exponential rate.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 32 canonical work pages

[1]

Indiana Univ

Adams, D., Simon, L.: Rates of asymptotic convergence ne ar isolated singularities of geo- metric extrema. Indiana Univ. Math. J. 37(2), 225-254 (1988)

work page 1988
[2]

Allard, W ., Almgren, F.: On the radial behavior of minima l surfaces and the uniqueness of their tangent cones. Ann. Math. 113(2), 215-265 (1981) 41

work page 1981
[3]

Springer-V erlag London Limited 2010, Cor- rected printing, (2012)

Andrew, P .: Elementary Differential Geometry. Springer-V erlag London Limited 2010, Cor- rected printing, (2012)

work page 2012
[4]

Nonlinear Anal

Bonforte, M., Figalli, A.: The Cauchy-Dirichlet Proble m for the fast diffusion equation on bounded domains. Nonlinear Anal. 239, Paper No. 113394, 55 pp. (2024)

work page 2024
[5]

Peking University Press, Beijing, (2003)

Chen, Y .: Parabolic Partial Differential Equation of Se cond Order. Peking University Press, Beijing, (2003)

work page 2003
[6]

Chen, Y ., Struwe, M.: Existence and Partial Regularity R esults for the Heat Flow for Har- monic Maps. Math.Z. 201(1), 83-103 (1989)

work page 1989
[7]

Chiarenza, F., Serapioni, R.: Degenerate parabolic equ ations and Harnack inequality. Ann. Mat. Pura Appl. 137(4), 139-162 (1984)

work page 1984
[8]

Springer , New Y ork, (2012)

Coleman, R.: Calculus on Normed V ector Spaces. Springer , New Y ork, (2012)

work page 2012

Show all 34 references

[9]

Eells, J., Sampson, J.: Harmonic Mappings of Riemannian Manifolds. Amer. J. Math. 86, 109-160 (1964)

1964
[10]

Wadsworth and Brooks/cole Math- ematics, 19(1), (2010)

Evans, L: Partial Differential Equations second editi on. Wadsworth and Brooks/cole Math- ematics, 19(1), (2010)

2010
[11]

Fioravanti, G.:The Dirichlet problem on lower dimensi onal boundaries: Schauder estimates via perforated domains.arXiv:2412.11294 [math.AP](2024 )

2024
[12]

Gilbarg, D., Trudinger, N.: Elliptic partial differen tial equations of second order. 2nd. Springer, Berlin, (1983)

1983
[13]

Lecture Notes in Mathematics, 471, Springer-V erlag, Berlin-New Y ork, (1975)

Hamilton, R.: Harmonic maps of manifolds with boundary . Lecture Notes in Mathematics, 471, Springer-V erlag, Berlin-New Y ork, (1975)

1975
[14]

and Xiong, J.: Asympto tic Analysis of Harmonic Maps with prescribed singularities

Han, Q., Khuri, M., Weinstein, G. and Xiong, J.: Asympto tic Analysis of Harmonic Maps with prescribed singularities. arXiv:2212.14826 [math.D G](2022)

2022 arXiv
[15]

Jendoubi, M.: A simple uniﬁed approach to some converge nce theorems of L. Simon. J. Funct. Anal. 153(1), 187–202 (1998)

1998
[16]

: Heat ﬂows on hyperbolic spaces

Lemm, M., Markovi´ c,V . : Heat ﬂows on hyperbolic spaces . J. Differential Geom. 108(3), 495–529 (2018)

2018
[17]

Li, P ., Tam, L.:The heat equation and harmonic maps of co mplete manifolds. Invent. Math. 105, 1–46 (1991)

1991
[18]

Li, P ., Tam, L.:Uniqueness and regularity of proper har monic maps. Ann. of Math. (2) 137(1), 167–201 (1993)

1993
[19]

Indiana Univ

Li, P ., Tam, L.:Uniqueness and regularity of proper har monic maps II. Indiana Univ. Math. J. 42(2), 591–635 (1993)

1993
[20]

Li, P ., Wang, J.:Harmonic rough isometries into Hadama rd space. Asian J. Math. 2(3), 419– 422 (1998)

1998
[21]

Li, Y ., Tian, G.: Regularity of Harmonic Maps with Presc ribed Singularities. Commun. Math. Phys. 149, 1–30 (1992)

1992
[22]

Paciﬁc J

Liao, G., Tam, L.:On the heat equation for harmonic maps from non-compact manifolds. Paciﬁc J. Math. 153(1), 129–145 (1992)

1992
[23]

World Scientiﬁc, (1996)

Lieberman, G.: Second Order Parabolic Differential Eq uations. World Scientiﬁc, (1996)

1996
[24]

World Scientiﬁc Publishing Co

Lin, F., Wang, C.: The Analysis of Harmonic Maps and thei r Heat Flows. World Scientiﬁc Publishing Co. Pte. Ltd., Hackensack, NJ (2008)

2008
[25]

Markovi´ c, V .: Harmonic maps between 3-dimensional hy perbolic spaces. Invent. Math. 199(3), 921–951 (2015)

2015
[26]

Lecture Notes in Pure and Appl

Schoen, R.:The role of harmonic mappings in rigidity an d deformation problems, in Com- plex geometry (Osaka, 1990). Lecture Notes in Pure and Appl. Math. 143, 179–200 (1990)

1990
[27]

Simon, L.: Asymptotics for a class of non-linear evolut ion equations, with applications to geometric problems. Ann. Math. 118, 525–571 (1983)

1983
[28]

Lectures in Mathematics, ETH Z¨ uric, Birkh¨ auser, Basel, (1996)

Simon, L.: Theorems on Regularity and Singularity of En ergy Minimizing Maps. Lectures in Mathematics, ETH Z¨ uric, Birkh¨ auser, Basel, (1996)

1996
[29]

L.: An introduction to Manifolds, Second Edition

Tu. L.: An introduction to Manifolds, Second Edition. S pringer New Y ork, NY , (2011) 42

2011
[30]

Wang, J.:The heat ﬂow and harmonic maps between complet e manifolds. J. Geom. Analy- sis. 8(3), 485–514 (1998)

1998
[31]

Weinstein, G.: On rotating black holes in equilibrium i n general relativity. Ph.D. thesis, at Courant Institute, New Y ork University (1989)

1989
[32]

Duke Math

Weinstein, G.: On the Dirichlet problem for harmonic ma ps with prescribed singularities. Duke Math. J. 77(1), 135–165 (1995)

1995
[33]

Weinstein, G.: Harmonic maps with prescribed singular ities into Hadamard manifolds. Math. Res. Lett. 3(6), 835–844 (1996)

1996
[34]

Weinstein, G.: Harmonic maps with prescribed singular ities and applications in general relativity, in the role of metrics in the theory of partial di fferential equations. Adv. Stud. Pure Math. 85, Math. Soc. Japan, Tokyo, 479–489 (2020) 43

2020

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Indiana Univ

Adams, D., Simon, L.: Rates of asymptotic convergence ne ar isolated singularities of geo- metric extrema. Indiana Univ. Math. J. 37(2), 225-254 (1988)

work page 1988

[2] [2]

Allard, W ., Almgren, F.: On the radial behavior of minima l surfaces and the uniqueness of their tangent cones. Ann. Math. 113(2), 215-265 (1981) 41

work page 1981

[3] [3]

Springer-V erlag London Limited 2010, Cor- rected printing, (2012)

Andrew, P .: Elementary Differential Geometry. Springer-V erlag London Limited 2010, Cor- rected printing, (2012)

work page 2012

[4] [4]

Nonlinear Anal

Bonforte, M., Figalli, A.: The Cauchy-Dirichlet Proble m for the fast diffusion equation on bounded domains. Nonlinear Anal. 239, Paper No. 113394, 55 pp. (2024)

work page 2024

[5] [5]

Peking University Press, Beijing, (2003)

Chen, Y .: Parabolic Partial Differential Equation of Se cond Order. Peking University Press, Beijing, (2003)

work page 2003

[6] [6]

Chen, Y ., Struwe, M.: Existence and Partial Regularity R esults for the Heat Flow for Har- monic Maps. Math.Z. 201(1), 83-103 (1989)

work page 1989

[7] [7]

Chiarenza, F., Serapioni, R.: Degenerate parabolic equ ations and Harnack inequality. Ann. Mat. Pura Appl. 137(4), 139-162 (1984)

work page 1984

[8] [8]

Springer , New Y ork, (2012)

Coleman, R.: Calculus on Normed V ector Spaces. Springer , New Y ork, (2012)

work page 2012