1: \begin{abstract}

2: 	This paper studies  the long time stability  of  both  stochastic heat equations 

3: 	on a bounded domain  driven by a correlated noise and their approximations. 

4: 	It is   popular for researchers to prove the intermittency of the solution 

5: 	which means that the moments of solution to  stochastic heat equation 

6: 	usually  grow exponentially to infinite  and this   hints  that the 

7: 	solution to stochastic heat equation is generally not stable in long time.

8: 	However, quite surprisingly in this paper 

9: 	we show that when the domain is bounded and when the noise 

10: 	is  not singular in spatial variables, the system can be  long time stable

11: 	and we also prove that we can approximate the solution by its finite dimensional

12: 	spectral approximation  which  is also long time stable. 

13: 	The idea is to use  eigenfunction expansion of the Laplacian on   bounded domain.   

14: 	% Given the feasibility of numerical approximation, we truncate the eigenfunction expansion of the solution and noise, and thus get the convergence and the long time numerical stability of the stochastic heat equations. The discretization in time is a implicit scheme. Under the regular assumptions of the noise, we investigate the long time numerical stability. 

15: 	We also present numerical experiments which are consistent with our theoretical results.

16: \end{abstract}

17: