1: \begin{abstract} %Reussite Combustion 2018

2:   The modeling of atmospheric processes in the context of

3:   weather and climate simulations is an important and

4:   computationally expensive challenge. The temporal integration

5:   of the underlying PDEs requires a very large number of time

6:   steps, even when the terms accounting for the propagation

7:   of fast atmospheric waves are treated implicitly. Therefore,

8:   the use of  parallel-in-time integration schemes to reduce

9:   the time-to-solution is of increasing interest, particularly

10:   in the numerical weather forecasting field.

11: 

12:   We present a multi-level parallel-in-time integration

13:   method combining the Parallel Full Approximation Scheme in

14:   Space and Time (PFASST) with a spatial discretization

15:   based on Spherical Harmonics (SH). The iterative algorithm

16:   computes multiple time steps concurrently by

17:   interweaving parallel high-order fine corrections and serial

18:   corrections performed on a coarsened problem. To do that, we design a

19:   methodology relying on the spectral basis of the SH

20:   to coarsen and interpolate the problem in space. 

21: 

22:   The methods are evaluated on the shallow-water

23:   equations on the sphere using a set of tests

24:   commonly used in the atmospheric flow community.

25:   We assess the convergence of PFASST-SH upon refinement

26:   in time. We also investigate the impact of the coarsening

27:   strategy on the accuracy of the scheme, and specifically

28:   on its ability to capture the high-frequency modes

29:   accumulating in the solution. Finally, we study the

30:   computational cost of PFASST-SH to demonstrate that

31:   our scheme resolves the main features of the solution

32:   multiple times faster than the serial schemes. 

33: \end{abstract}

34: