Simulating the Characteristics of Tropical Cyclones over the South-West Indian Ocean: Sensitivity to Model Horizontal Resolution

Donkin, Paige Tayla

Simulating the Characteristics of Tropical Cyclones over the South-West Indian Ocean: Sensitivity to Model Horizontal Resolution

Master Thesis

2022

Abstract

Tropical cyclones (TCs) are the deadliest and most economically devastating of all natural hazards. Those that occur in the South-West Indian Ocean (SWIO) often strike vulnerable nations and can have serious socio-economic impacts, yet there is a dearth of information on how well these systems are simulated by contemporary climate models. This study investigates the performance of four global climate models and one variable-resolution climate model in simulating the characteristics of SWIO TC activity through both a climatological and individual event analysis, and investigates the influence of horizontal resolution on the simulated TC characteristics. The climatological analysis makes use of a multi-model, multi-resolution set of simulations from four modelling groups included in the CMIP6 High Res MIP project, covering the period 1980-2009. The individual event analysis focuses on the simulation of two highly destructive TC events which occurred in the SWIO in early 2019 – namely TC Idai and TC Kenneth. The simulation of individual events makes use of the atmospheric Model for Prediction Across Scales (MPAS-A), a non-hydrostatic variable-resolution model which is employed at multiple resolutions ranging from 60km–3km. In both analyses, systems were identified using a novel resolution dependent tracking algorithm that includes wind speed, vorticity, and warm core thresholds. All model simulations were validated against the Reunion Specialised Meteorological Centre (RSMC) best track observations and the European Centre for Medium-Range Weather Forecasts (ECMWF) fifth-generation reanalysis dataset (ERA5). The results of the CMIP6 analysis indicate that all models produce realistic representations of TCs over the SWIO, but underestimate TC intensity, and feature a westward shift bias in the spatial distributions of TC activity. Increasing model horizontal resolution enhances the quality of the simulated TC frequency and intensity, with higher-resolution models showing more realistic storm counts and an increased ability to represent the full range of observed TC intensities. However, increasing horizontal resolution does not improve the spatial distribution of cyclogenesis locations or TC tracks, nor the ability of the models to capture the long-term trends in TC activity. Similarly, the results of the MPAS-A analysis indicate that MPAS-A produces realistic TC tracks, but shows considerable southward track and landfall biases, and severely underestimates the intensities of the systems. A local increase in horizontal resolution over the SWIO improves all aspects of the TC simulation, including the TC track, intensity, wind-pressure relationship, and vertical structure. However, the level of improvement differs based on the physics parameterization suite employed in the model, the characteristic of interest, and the event in question. The present study illustrates the extent to which an increase in model horizontal resolution can improve simulated TC characteristics and offers insights on how to improve future projections of TC activity over the SWIO region.

Keywords

Environmental and Geographical Science

Reference:

Collections

Masters

Full item page