If u ∈ H 1 (M, N ) is a weakly J-holomorphic map from a compact without boundary almost hermitian manifold (M, j, g) into another compact without boundary almost hermitian manifold (N, J, h
Let (M, j, g) (respectively (N, J, h)) be a smooth hermitian almost complex manifold with dimension 2m (respectively 2n). Assume further that(N, h) is compact without boundary and isometrically embedded into some euclidean R k via the Nash's embedding theorem. Denote the Sobolev spaceis said to be weakly (j, J)-holomorphic map (or J-holomorphic map for abbreviation) if du preserves the almost complex structures in the sense: In this paper we are interested in regularity for weakly J-holomorphic Now we state our first theoremOur idea to prove theorem A follows from the two new observations: (1) Under the assumption that u(B r (x)) is contained in a coordinate chart U of N , we can use the local coordinate frame on U to express J as SO (2n) For m ≥ 2, observe that M 2,2m−2 is a scaling-invariant subspace of where F t : M → M is a parameter family of diffeomorphisms generated by X. It follows from Proposition 3.2 that any stationary J-holomorphic map satisfies the energy monotonicity inequality: there is a C 0 = C 0 (M, j, g) > 0such that e C0r r
2−2mBr (x)for any x ∈ M and 0 < r ≤ R ≤ R 0 = R 0 (M, j, g). A direct consequence of (1.5) is: for any x ∈ M and 0 < r ≤ R 0 ,Hence theorem A yields the partial regularity for stationary J-holomorphic maps, which is an analogy to the partial regularity for stationary harmonic maps. More precisely,Based on both the energy monotonicity inequality (1.5) and the small energy regularity theorem A, we find that the blow-up techniques for stationary harmonic maps developed by Lin [L] can be modified to study the convergence issues for sequences of stationary J-holomorphic maps.From now on, we call a nonconstant smooth J-holomorphic map ω :as a pseudo-holomorphic S 2 , here j 0 is the standard complex structure on S 2 . We prove
stationary J-holomorphic maps which converges weakly to u ∈ H 1 (M, N ). Then u is a weakly J-holomorphic map and there exists a closedas convergence of Radon measures, for some nonnegative Radon measure νThe main difference between our proof of theorem D and §2 of [L] is that we need to verify that the concentration set Σ is j-holomorphic (2m − 2)-rectifiable set. Once we achieve this, then both the conformality and the removablity of isolated singularity for pseudo-holomorphic curves (cf. [Y] [PW]) guarantee that the restriction of a bubble on (T x Σ) ⊥ can be lifted to be a pseudo-holomorphic S 2 .It is a very important problem to quantify the density function θ for the defect measure ν in the content of blow-up analysis for stationary harmonic The ideas to prove theorem E are based on the observations that on The paper is written as follows. In §2, we prove theorem A and C. In §3, we prove theorem D, E and F. In §4, we discuss the relationship between J-holomorphic maps and harmonic maps in the case either (M, j, g) and (N, J, h) are almost Kähler manifo...