Solvability of the Gleason problem on a class of bounded pseudoconvex domains

Let \chi\in C^{\infty}(\overline{\Omega}) so that \chi\equiv 1 on the support of W and \text{supp}(\chi)\cap F^{-1}(0,0,...,0)=\emptyset. Now define

and define

Now, \tau_F(X)=1 on the support of W and so we define Y:=X\wedge W\in \Gamma^{\infty}_{(r+1,s)}. Then

and the support of Y is away from F^{-1}(0,0,...,0). Furthermore, \overline{\partial}Y is bounded since \overline{\partial}W=0 and g_j\in C^{\infty}(\overline{\Omega}) for j\in \{1,...,n\}. Note that in [ST, Proof of Lemma 3] we needed X to have compact support. This is not needed in this proposition since we assumed F is smooth up to the closure of the domain (that is, \{f_1,...,f_n\}\subset C^{\infty}(\overline{\Omega})) and injective (so F^{-1}(0,0,...,0) is a singleton set). The rest of the proof follows the descending induction procedure seen in [ST, Proof of Lemma 3] with some minor modifications. Let s=n, 0\leq r\leq m-1 and also W\in \Gamma^{\infty}_{(r,n)} so that \text{supp}(W)\cap F^{-1}(0)=\emptyset and \tau_F(W)=0. It is clear that \overline{\partial}W=0 since any (0,n) form is \overline{\partial}-closed. Then there is Y_1\in \Gamma^{\infty}_{(r+1,n)} so that \tau_F(Y_1)=W. Also, it is clear that \overline{\partial}Y_1=0. Since \Omega is L^{\infty}-pseudoconvex, there exists Y\in \Gamma^{\infty}_{(r+1, n-1)} so that \overline{\partial}Y=Y_1. In other words, \tau_F\overline{\partial}(Y)=W. This is our base case. Now assume our proposition is true for s=k+1, k+2,...,n and r=0,1,...,m-1. Suppose 0\leq r\leq m-1 and suppose W\in \Gamma_{(r,k)}^{\infty} has the following properties:

1. \text{supp}(W)\cap F^{-1}(0,0,...,0)=\emptyset,

2. \overline{\partial}W=0 and \tau_F(W)=0.

Then there exists Y_1\in \Gamma^{\infty}_{(r+1,k)} so that

1. \overline{\partial}Y_1\in \Gamma^{\infty}_{(r+1,k+1)},

2. \text{supp}(Y_1)\cap F^{-1}(0,0,...,0)=\emptyset

Therefore, \overline{\partial}Y_1\in \Gamma^{\infty}_{(r+1,k+1)} satisfies the conditions of the proposition for s=k+1. In other words,

1. \text{supp}(\overline{\partial}Y_1)\cap F^{-1}(0,0,...,0)=\emptyset,

2. \overline{\partial}\overline{\partial}Y_1=0 and \tau_F\overline{\partial}(Y_1)=\overline{\partial}W=0.

By the induction hypothesis, we have \overline{\partial}Y_2\in \Gamma^{\infty}_{(r+2,k+1)} and \tau_F\overline{\partial}Y_2=\overline{\partial}Y_1 for some Y_2\in \Gamma^{\infty}_{(r+2,k)}. Thus we have

Let us define Y_3:=Y_1-\tau_FY_2. Then we have

and

Since \Omega is L^{\infty}-pseudoconvex, there exists Y\in \Gamma^{\infty}_{(r+1, k-1)}\cap\textbf{dom}(\overline{\partial}) so that \overline{\partial}Y=Y_3. Thus we have

as desired.

This proof is an application of a few results concerning the \overline{\partial}-Koszul complex. See [ST] for more information on the \overline{\partial}-Koszul complex. We will first consider n=2. Let \Omega\subset \mathbb{C}^2 be a bounded L^{\infty}-pseudoconvex domain. Let g\in H^{\infty}(\Omega) and without loss of generality, suppose g(0,0)=0. Since g has a power series expansion about (0,0) converging on some open set U\subset \subset \Omega, one can write

on U. Here, \lambda_1, \lambda_2\in H^{\infty}(U)\cap C^{\infty}(\overline{U}). Now let \chi_1\in C^{\infty}(U), supported in U, 0\leq \chi_1\leq 1 on U and \chi\equiv 1 on \widetilde{U}\subset \subset U. Also, (0,0)\in \widetilde{U}. Then on \Omega\setminus \overline{\widetilde{U}} we have

Therefore,

on \Omega. Now for the sake of notation, define

and

Notice that both L_1 and L_2 are bounded on \Omega and partial derivatives of L_1 and L_2 of all orders exist and are continuous on compact subsets of \Omega. In addition, \overline{\partial}L_1 and \overline{\partial}L_2 are bounded on \Omega. Now define

Notice that W_1 satisfies the hypothesis of Proposition 1 since the support of W_1 is away from F^{-1}(0,0) where F:=(z_1,z_2), W_1\in \Gamma^{\infty}_{(1,1)}\cap \textbf{dom}(\overline{\partial}), \overline{\partial}W_1=0, and \tau_F W_1=0. Thus by Proposition 1, there exists there exists H_1\in \mathcal{C}_{(0,0)}^{\infty}\cap\textbf{dom}(\overline{\partial}) so that

and

Thus we have that

and

Therefore,

and

Furthermore,

on \Omega. This completes the proof for n=2. We will demonstrate the proof works for n=3 and then show how it can be generalized. For \Omega\subset \mathbb{C}^n for n=3, we first use a power series argument similar to the n=2 case. That is, if g(0)=0, one can write g\equiv z_1L_1+z_2L_2+z_3L_3 where L_j are holomorphic on a neighborhood of 0, L_j are smooth and bounded on \Omega, and \overline{\partial}L_j\in L^{\infty}_{(0,1)}(\Omega) for j=1,2,3. We define

Clearly, \overline{\partial}W=0 and \tau_F(W)=0 where F=(z_1,z_2,z_3). Furthermore, the support of W is away from F^{-1}(0,0,0). Therefore, by Proposition 1, there exists Y\in \Gamma^{\infty}_{(2,0)} so that

Now, Y\in \Gamma^{\infty}_{(2,0)}, therefore, Y has the form

for \{y_1, y_2, y_3\}\subset \mathcal{C}_{(0,0)}^{\infty}\cap\textbf{dom}(\overline{\partial}). Now we apply \tau_F\overline{\partial} to Y and collect terms together. We get

Thus we have

and

This implies

and

Furthermore, we can write

The proof for n>3 is an analog of the previous proofs with more terms involved. We write g as a linear combination of bounded functions in the domain of \overline{\partial} that are holomorphic on a neighborhood of 0. Then we use Proposition 1 to ’correct’ these bounded functions to be holomorphic and bounded on \Omega. Furthermore, the linear combination of these functions still gives g.