1: \begin{abstract}

2: %We created a new thing to approximate another new thing, and then we created a more general new thing to approximate another more general new thing. There was a good reason, but we don't remember. It is about the journey, and not the destination. The journey was good. You should have come along. 

3: {Flip-flop processes refer to a family of stochastic fluid

4:   processes which converge to either a standard Brownian motion (SBM)

5:   or to a Markov modulated Brownian motion (MMBM). In recent years, it

6:   has been shown that complex distributional aspects of the univariate

7:   SBM and MMBM can be studied through the limiting behaviour of

8:   flip-flop processes. Here, we construct two classes of bivariate

9:   flip-flop processes whose marginals converge strongly to SBMs and

10:   are dependent on each other, which we refer to as \emph{alternating

11:     two-dimensional Brownian motion processes}. While the limiting

12:   bivariate processes are not Gaussian, they possess desirable

13:   qualities, such as being tractable and having a time-varying

14:   correlation coefficient function.}

15: 

16: 

17: \keywords{ Strong convergence, two-dimensional Brownian motion, stochastic fluid processes, flip-flops, time-varying correlation coefficient} 

18: 

19: \emph{2010 Mathematics Subject Classification:} Primary: 60F15, 60G50; Secondary: 60G15

20: 

21: %60F15   	Strong theorems

22: %60G15   Gaussian processes

23: %60G50   Sums of independent random variables; random walks

24: \end{abstract}