Atmospheric processes

Abstract:

Machine-learnt weather prediction (MLWP) models are now well established as being competitive with conventional numerical weather prediction (NWP) models in the medium range. However, there is still much uncertainty as to how this performance extends to longer timescales, where interactions with slower components of the earth system become important. We take GenCast, a state-of-the-art probabilistic MLWP model, and apply it to the task of seasonal forecasting with prescribed sea surface temperature (SST), by providing anomalies persisted over climatology (GenCast-Persisted) or forcing with observed SSTs (GenCastForced). The forecasts are compared to the European Centre for Medium-Range Weather Forecasts seasonal forecasting system, SEAS5. Our results indicate that, despite being trained at short timescales, GenCast-Persisted produces much of the correct precipitation patterns in response to El Ni˜no and La Ni˜na events, with several erroneous patterns in GenCast-Persisted corrected with GenCast-Forced. The uncertainty in precipitation response, as represented by the ensemble, compares favourably to SEAS5. Whilst SEAS5 achieves superior skill in the tropics for 2-metre temperature and mean sea level pressure (MSLP), GenCast-Persisted achieves higher skill in some areas in higher latitudes, including mountainous areas, with notable improvements for MSLP in particular; this is reflected in a slightly higher correlation with the observed NAO index. Reliability diagrams indicate that GenCast-Persisted has little skill relative to climatology, whilst GenCast-Forced produces forecasts with reliability comparable to SEAS5. These results provide an indication of the potential of MLWP models similar to GenCast for the ‘full’ seasonal forecasting problem, where the atmospheric model is coupled to ocean, land and cryosphere models.

Abstract:

Data driven parameterisations for sub-grid processes unlocks the ability to surpass the current computational constraints of Earth system models. However, machine learning (ML) can be brittle. State-of-the-art ML approaches can reliably perform on in-distribution data, exceeding human ability across a diverse range of tasks. Yet, when faced with shifts in data distribution, performance degrades. In climate modelling, when the task is predicting the state of a non-stationary system, this is evidently a substantial issue. We illustrate this with the ClimSim dataset, forming spatio-temporal groups and quantitatively show how even small shifts in distribution affect performance.Next, we use the theory of compositional generalisation to build models which are less susceptible to these shifts in distribution. Compositional generalisation is the formation of novel combinations of observed elementary components. That is, the ability to decompose data into building blocks that are reused across both the in- and shifted-domains, such that a model can capture a domain shifted state through a set of in-domain, learnt abstractions. Inspired by these concepts we propose various architectural and regularisation changes to standard ML parameterisations to improve generalisation. Preliminary results in sub-grid process emulators suggest new insights into if and how CG can reduce model sensitivity to domain shifts.

Abstract:

Understanding how fast atmospheric variability shapes slow climate variability and sensitivity is a central challenge in Earth-system science. Recent advances in machine-learned (ML) atmospheric models have demonstrated remarkable skill on weather timescales, but their emergent behaviour in a fully coupled climate system is largely unexplored. We present results from a new hybrid modelling framework that couples a machine-learned atmosphere to a dynamical ocean model. We report on a set of 70-year coupled simulations (1950–2020 historical forcing and fixed-1950s control) in which the ACE2 ML climate emulator is interactively coupled to the NEMO ocean model. These experiments represent, to our knowledge, the first multi-decadal integrations of a machine-learned atmosphere interacting with a full-depth dynamical ocean. We assess the behaviour of the coupled system, with particular focus on low-frequency tropical variability and the climate response to greenhouse-gas forcing. Preliminary results indicate realistic emergent El Nino-like variability and a physically plausible climate sensitivity, suggesting that key atmosphere–ocean feedbacks can be captured within a hybrid ML–dynamical framework. These results evaluate the possible role of entirely machine-learned components in next-generation Earth-system models.

Abstract:

Floods are hydro-climatic extremes with severe socioeconomic and environmental consequences. Many studies have examined how large-scale modes of climate variability (e.g., ENSO, NAO) influence floods, but many have relied on catchments influenced by anthropogenic activities, which obscure underlying climate-flood relationships. Here, we use the newly released ROBIN Reference Hydrometric Network, a global dataset of over 3,000 near-natural catchments with daily streamflow records, to provide an observational assessment of climate-flood relationships at the global scale. We first quantify long-term and multi-temporal trends in annual flood peaks and peak-over-threshold events and evaluate their connections with key modes of climate variability across different IPCC regions. Trend analysis reveals how flood metrics have evolved across regions and time periods, while correlation analysis reveals the modes of climate variability that are associated with year-to-year variations in flood peaks and frequencies. A signal-to-noise framework tests whether global mean surface temperature leaves a detectable fingerprint on high flow regimes. This analysis helps to clarify the extent to which climate variability influences flood occurrence and magnitude in near-natural catchments worldwide. Moreover, we propose a machine learning-based process attribution framework to identify climate and catchment controls on floods in near-natural catchments. Preliminary results indicate substantial spatial variability in dominant flood drivers across and within IPCC regions and suggest that large-scale atmospheric circulation modes exert strong, but regionally distinct, influence on seasonal flood frequency. Overall, our findings underscore the importance of regional climate modes in modulating floods and provide the first global baseline on climate-driven changes to floods in near-natural catchments.

Abstract:

The mesoscale atmospheric energy spectrum has puzzled scientists for decades, sitting between classical turbulence and wave theories. Using year-long ECMWF operational analyses of high resolution and a spherical coarse-graining framework (Flowsieve), we present the first consistent global maps of local mesoscale kinetic energy fluxes. At 200~hPa, we identify a striking band of upscale transfer aligned with the ITCZ, while storm tracks and orography leave distinct dynamical imprints at both 200 and 600~hPa. By decomposing divergent and rotational components, we show that divergent energy dominates in the tropics and stratosphere, while rotational energy dominates in the extratropical troposphere. Conditioning spectra on this balance reveals contrasting regimes: a Nastrom–Gage-like spectrum under divergent dominance, and a spectrum reminiscent of the classical dual cascade of textbook two-dimensional turbulence under rotational dominance at 600~hPa. These results demonstrate that mesoscale energy transfer is shaped by a patchwork of mechanisms, reconciling long-standing debates and providing new inspiration for parametrisations and predictability in weather and climate models.