Parabolic regularzation of a first order stochastic partial differential equation (Q4487015)

The authors use regularization methods for proving the existence and uniqueness of smooth solutions of a first-order semilinear stochastic partial differential equation of the following type \[ dX(t,x)=\left\{\sum_{i=1}^n\partial_{x_i}[b_i(t,x)X(t,x)]+A(t,x,X(t,x)) \right\}dt +B(t,x,X(t,x))dW(t) \] for \(t>0\) and \(x\in\mathbb R^n\). For the regularizations they add \(\varepsilon\Delta X\) to the equation such that the known theory of stochastic parabolic Itô equations can be applied to obtain results for the regularized problem. By taking the limit \(\varepsilon\rightarrow 0\) they obtain \(L^2(\mathbb R^n)\)-continuous generalized (or weak) solutions, which are pathwise unique. Moreover, they consider mean square boundedness of solutions. For this they establish a uniform bound for the mean value \(E\|X(t)\|_{L^2(\mathbb R^n)}^2\) on bounded intervals \(t\in[0,T]\). In case \(n=1\), this bound holds already for any \(t\in\mathbb R\).