Martin is investigating extreme rainfall events in the tropics and sub-tropics by applying state of the art machine learning algorithms to observational data and by conducting high resolution cloud resolving model experiments.
In his previous post-doctoral fellowship within the Center of Excellence for Climate System Sciences at Monash University he developed a sea-breeze parametrization method for global climate models, the first one of its kind. This new parametrization has been implemented into the UK Met Office's Unified Model to improve the prediction of clouds, rainfall and convection in coastal areas.
During his PhD Martin developed a data driven computer vision algorithm that objectively detects rainfall patterns related to coastal land sea-interactions. With help of the results of this pattern recognition he developed a Markov chain based stochastic modelling approach that is able to capture the main characteristics of convection and clouds in coastal regions of the tropics.