Extreme event attribution is a rapidly growing field of climate science with important implications for public and government understanding of human-induced climate change. However, there is substantial variation in how well events can be attributed to human-induced climate change, depending on the nature of the event. Focusing on Australia: at one end of the scale, large-scale heat events on both the land and in the ocean are well suited to attribution studies because climate models simulate them reasonably well, there are high-quality observations available and our understanding of the processes that lead to extreme heat events is reasonably well developed. At the other end of the scale, very important phenomenon such as changes in east coast lows, severe convective storms and long-term droughts are less well observed, are beyond our current capability to robustly simulate in climate models and the complex mechanisms that lead to intensification are not well understood. Thus, some important extreme events can be linked to human-induced climate change, with a high degree of confidence, while others cannot. We review the state of the science relevant to event attribution with a focus on Australia. We highlight where progress can be made, focusing on observations, physical understanding, and realistic climate modelling.

极端事件归因是气候科学中一个快速发展的领域，对公众和政府对人为气候变化的理解具有重要意义。然而，根据事件的性质，将事件归因于人类引起的气候变化的程度存在很大差异。以澳大利亚为重点：一方面，陆地和海洋上的大规模高温事件都非常适合归因研究，因为气候模型可以很好地模拟它们，有高质量的观测数据以及我们对事件的理解导致极端高温事件的过程相当发达。另一方面，东海岸低压变化、强对流风暴和长期干旱等非常重要的现象却很少被观测到，超出了我们目前在气候模型中稳健模拟的能力以及导致这些现象的复杂机制。强化程度还没有被很好地理解。因此，一些重要的极端事件可以高度可信地与人类引起的气候变化联系起来，而另一些则不能。我们回顾了与事件归因相关的科学状况，重点关注澳大利亚。我们强调可以取得进展的领域，重点关注观测、物理理解和现实的气候建模。