Understanding volcanic activity is crucial for uncovering the fundamental physical mechanisms governing this natural phenomenon. In this study, we show the application of multifractal and statistical analysis to investigate changes associated with volcanic activity. We aim to identify significant variations within the physical processes related to changes in volcanic activity. These methodologies offer the potential to identify pertinent changes preceding a high-energy explosion.

We train neural networks as denoising diffusion models for state generation in the Lorenz 1963 system and demonstrate that they learn an internal representation of the system. We make use of this learned representation and the pre-trained model in two downstream tasks: surrogate modelling and ensemble generation. For both tasks, the diffusion model can outperform other more common approaches. Thus, we see a potential of representation learning with diffusion models for dynamical systems.

The characterisation of abrupt transitions in palaeoclimate records allows understanding of millennial climate variability and potential tipping points in the context of current climate change. In our study an algorithmic method, the matrix profile, is employed to characterise abrupt warmings designated as Dansgaard–Oeschger (DO) events and to identify the most similar transitions in the palaeoclimate time series.

Our study investigates how turbulence behaves in stable conditions using direct numerical simulations. We found that rethinking how energy dissipates in these situations is crucial. By revising existing models, we uncovered limitations in understanding how temperature is transported vertically in very stable conditions. We focus on how turbulence works in extreme stability and offer new insights that could improve our understanding of natural phenomena affected by stable atmospheric conditions.

The intensification of tropical cyclones (TCs) is explored through a conceptual model derived from geophysical principals. Focus is put on the behaviour of the model with parameters which change in time. The rates of change cause the model to either tip to an alternative stable state or recover the original state. This represents intensification, dissipation, or eyewall replacement cycles (ERCs). A case study which emulates the rapid intensification events of Hurricane Irma (2017) is explored.

A methodology for directly predicting the time evolution of the assumed parameters for the distribution densities based on the Liouville equation, as proposed earlier, is extended to multidimensional cases and to cases in which the systems are constrained by integrals over a part of the variable range. The extended methodology is tested against a convective energy-cycle system as well as the Lorenz strange attractor.

A novel approach, optimal transport data assimilation (OTDA), is introduced to merge DA and OT concepts. By leveraging OT's displacement interpolation in space, it minimises mislocation errors within DA applied to physical fields, such as water vapour, hydrometeors, and chemical species. Its richness and flexibility are showcased through one- and two-dimensional illustrations.

Recently, scientists have been looking into ways to control the weather to lead to a desirable direction for mitigating weather-induced disasters caused by torrential rainfall and typhoons. This study proposes using the model predictive control (MPC), an advanced control method, to control a chaotic system. Through numerical experiments using a low-dimensional chaotic system, we demonstrate that the system can be successfully controlled with shorter forecasts compared to previous studies.