As a consequence of the pandemic situation, a series of online seminars titled Perspectives on Climate Sciences: from historical developments to research frontiers was organized in October 2020. The seminars were such a success that the series was extended twice, finally running until July 2021.

The aim of this initiative was to allow senior researchers in the field of climate sciences to share their expertise and exchange ideas with early-career scientists. The ultimate goal was to raise awareness of the chain of arguments, discoveries, and technological breakthroughs that have enabled today's most interesting and relevant lines of research in climate sciences. The speakers were asked to give their perspective on what future developments they expect in the field and, last but not least, how their personal background, the historical situation at the time of their most seminal works, and the network of interactions they established, contributed to their career.

The seminar series, hosted on online platforms, was supported by the Nonlinear Processes in Geosciences (NP) division of the European Geosciences Union (EGU) at first and later by the EGU as a whole, as an official NP Campfire event. It featured 24 contributions that were livestreamed and uploaded on a Youtube channel, with more than 300 subscribed users. Overall, almost 1000 users registered for the seminars, and the livestream featured an average of about 80 participants, with peaks of 170. The abstracts from the seminars can be found here: https://sites.google.com/view/perspectivesonclimate/materials

The development of theories and tools in nonlinear sciences contributed crucially to the understanding and prediction of the climate system. One well-known example is the development of chaos theory, which allowed us to understand the apparently erratic behavior of climate fields and their sensitivity to initial and model errors, and it pushed forward the development of probabilistic predictions. Other important examples are the successes of (i) the theory of stochastic resonance for understanding the transitions between glacial and interglacial climate states, (ii) the development of optimal fingerprinting techniques in order to separate the forced changes from internal variability, and (iii) applications of stochastic methods, such as stochastic parameterizations, which pushed breakthrough achievements in the setup of efficient and manageable Earth system models for climate prediction. Some of these advancements have been recognized for their outstanding contribution to the detection of anthropogenic climate change and their impact on human societies by awarding the 2021 Nobel Prize in Physics to Syukuro Manabe, Klaus Hasselmann, and Giorgio Parisi for their studies on nonlinear dynamics in complex systems and their applications in climate sciences and beyond. Klaus Hasselmann was one of the speakers of this seminar series.

With this special issue, we want to build upon the outcomes of the seminar series, and we wish to highlight some of the nonlinear and interdisciplinary foundations on which modern climate sciences were built, by inviting a group of influential scientists to review their achievements and the perspectives of their work.