1: \begin{abstract}

2: 

3: 

4:   We construct a continuous domain for temporal discretization of

5:   differential equations. By using this domain, and the domain of

6:   Lipschitz maps, we formulate a generalization of the Euler operator,

7:   which exhibits second-order convergence. We prove computability of

8:   the operator within the framework of effectively given domains. The

9:   operator only requires the vector field of the differential equation

10:   to be Lipschitz continuous, in contrast to the related operators in

11:   the literature which require the vector field to be at least

12:   continuously differentiable. Within the same framework, we also

13:   analyze temporal discretization and computability of another variant

14:   of the Euler operator formulated according to Runge-Kutta theory. We

15:   prove that, compared with this variant, the second-order operator

16:   that we formulate directly, not only imposes weaker assumptions on

17:   the vector field, but also exhibits superior convergence rate. We

18:   implement the first-order, second-order, and Runge-Kutta Euler

19:   operators using arbitrary-precision interval arithmetic, and report

20:   on some experiments. The experiments confirm our theoretical

21:   results. In particular, we observe the superior convergence rate of

22:   our second-order operator compared with the Runge-Kutta Euler and

23:   the common (first-order) Euler operators.

24:   

25: \end{abstract}

26: