Variability and Predictability of Blocking Regimes from Idealised to General Circulation Models
(2025)
Files
Anupama_Thesis_Final.pdf
Open Access - Adobe PDF
- 40.09 MB
Details
- Authors
- Supervisors
- Abstract
- The weather in the midlatitudes exhibits pronounced low frequency variability (LFV), characterised by persistent circulation patterns that evolve on timescales far longer than the typical two week forecasting limit. A central expression of this LFV is atmospheric blocking, whose onset, maintenance, and decay influence the occurrence of extreme weather such as heatwaves, cold spells, droughts, and prolonged precipitation events. Despite decades of study, blocking remains a major source of forecast uncertainty due to its nonlinear dynamics, sensitivity to initial conditions, and strong interactions with the underlying mean flow and synoptic scale disturbances. The overarching aim of this thesis is to investigate the variability and predictability of blocking using a hierarchy of reduced order and comprehensive general circulation models, with the goal of improving dynamical understanding of how blocking emerges, how its stability depends on geographical context. A first line of investigation employs a reduced order quasi geostrophic land - atmosphere coupled model to explore how blocking behaviour arises from idealised dynamical interactions. The model’s response to key environmental parameters, such as surface friction and radiative forcing, was systematically analysed to establish its regime structure. Distinct zonal, transition, and blocked states were identified, and their stability was quantified using backward Lyapunov exponents. The results show that blocked regimes are inherently more unstable than zonal flows and that their predictability varies strongly with location relative to idealised topography: blocks forming upstream of the imposed orographic feature were systematically less stable than downstream ones. This mechanistic result highlights the importance of geographic positioning in shaping blocking behaviour. To test whether such asymmetries persist in a more realistic context, blocking in the North Pacific was analysed using daily CMIP6 MIROC6 simulations. Blocking events were objectively classified into Western and Eastern regimes, and their intrinsic predictability was quantified through analogue based error growth. Contrary to the idealised model, Eastern North Pacific blocks displayed markedly lower predictability and faster error divergence, while Western blocks were comparatively more stable. The robustness of this result was confirmed through extensive sensitivity tests on detection thresholds and analogue definitions. This highlights that blocking predictability is not solely dictated by idealised dynamical principles but is also shaped by realistic basin geometry, background flow structure, and transient eddy activity. A complementary analysis examined blocking from an energetic viewpoint by quantifying interactions between kinetic energy, available potential energy, and near-surface temperature anomalies using Lorenz Energy Cycle diagnostics and an information transfer framework. These diagnostics reveal that the behavior of influence between energetic variables differ substantially between Eastern and Western blocking events, resulting in distinct atmospheric responses during blocked and non blocked periods. The results show that energetic pathways, and not only flow stability, contribute to the regional differences in blocking behaviour observed in realistic models. Finally, to contextualise blocking within broader LFV dynamics, the thesis also examines how model structure and scale interactions influence variability. Sensitivity tests with altered model resolution highlighted how reduced order systems can exhibit qualitatively different dynamical behaviour depending on the number of spatial modes included. This underscores the importance of choosing model complexity carefully when studying blocking dynamics and LFV. In the essence, this thesis demonstrates that atmospheric blocking variability and predictability are strongly modulated by both dynamical stability and geographical setting, and that different model hierarchies reveal complementary aspects of this behaviour. The results provide a unified framework for understanding blocking across idealised and comprehensive models, and they lay the groundwork for future studies on how blocking and midlatitude LFV more broadly may evolve and interact with different atmospheric components.
- Affiliations
Citations
Kuttikkat Xavier, A. (2025). Variability and Predictability of Blocking Regimes from Idealised to General Circulation Models. https://hdl.handle.net/2078.5/270901