Tropical cyclones (TCs), locally known as hurricanes or typhoons, are one of the deadliest and costliest natural hazards, causing widespread havoc in coastal areas when they make landfall. Their primary hazards include high wind speeds, storm surge, and precipitation, but these can, in turn, trigger other hazards, such as landslides or the spread of water-borne diseases. The 2017 Atlantic Hurricane Season has been the costliest to date, with Hurricanes Harvey, Irma, and Maria’s combined overall losses estimated around US$ 220 billion. To protect coastal communities from these powerful storms and to reduce the future loss of life and property, it is crucial to support risk mitigation efforts with reliable TC risk assessments. Achieving this goal, however, requires adequate understanding of the characteristics of TCs (e.g., intensity, frequency, etc.) and of how these characteristics change under (near-) future climate change. The goal of this thesis is therefore to develop a novel method to derive and assess global-scale TC activity and wind speed probabilities, now and under climate change.