Tropical Cyclones and Equatorial Waves in Global Storm-Resolving Models

Since the inception of numerical weather and climate prediction, numerical weather prediction models have struggled to accurately simulate the tropical atmosphere. In FY2019, MMM scientists found that that global storm resolving models (global models with 5 km grid spacing or less) eliminate some issues that previous models had, mostly by foregoing the need to parametrize deep convection. In particular, global storm-resolving models improve the representation of tropical cyclones and equatorial waves. In theory, the weather has longer predictability in the tropics than in the middle latitudes. Notwithstanding their overall positive impact, global storm-resolving models still suffer from substantial biases. A model intercomparison reveals that all participating models over-predict tropical cyclone intensity, whereas each model struggles in its own way to predict the number of cyclones in a given ocean basin.

Tropical cyclone tracks for the period 1 Aug--10 Sep 2016. Top row (black) is observation from the IBTrACS database, the rows below (red) are from MPAS simulations with successively lower resolution
Figure: Tropical cyclone tracks for the period 1 Aug--10 Sep 2016. Top row (black) is observation from the IBTrACS database, the rows below (red) are from MPAS simulations with successively lower resolution. The 3.75-km MPAS captures tropical cyclone activity most realistically.

Acknowledgments

  • CISL