A Long-Time-Step-Permitting Tracer Transport Model on the Regular Latitude–Longitude Grid

Abstract

If an explicit time scheme is used in a numerical model, the size of the integration time step is typically limited by the spatial resolution. This study develops a regular latitude–longitude grid-based global three-dimensional tracer transport model that is computationally stable at large time-step sizes. The tracer model employs a finite-volume flux-form semi-Lagrangian transport scheme in the horizontal and an adaptively implicit algorithm in the vertical. The horizontal and vertical solvers are coupled via a straightforward operator-splitting technique. Both the finite-volume scheme’s one-dimensional slope-limiter and the adaptively implicit vertical solver’s first-order upwind scheme enforce monotonicity. The tracer model permits a large time-step size and is inherently conservative and monotonic. Idealized advection test cases demonstrate that the three-dimensional transport model performs very well in terms of accuracy, stability, and efficiency. It is possible to use this robust transport model in a global atmospheric dynamical core.