Research Highlights

Active and break spells are the primary drivers of regional flood and drought in the Indian summer monsoon, making their dynamics central to seasonal prediction and water resource planning. Flood and drought years are distinguished by contrasting intraseasonal oscillation modes: high-frequency (10–20 day) northwestward-propagating oscillations dominate in flood years, while low-frequency (20–60 day) poleward-drifting modes prevail in drought years. Stronger midlatitude dry-air intrusions intensify break phases in drought years. A moisture budget decomposition shows that advection of mean moisture gradients by mean winds is the leading term differentiating flood from drought years — providing new dynamical insight into how intraseasonal processes shape interannual extremes in the South Asian monsoon.

Whether the South Asian monsoon qualifies as an Earth system tipping element has been a matter of scientific debate. Low-order models of monsoon energetics show that despite destabilizing processes in the thermodynamic budget, a bifurcation is unlikely — the dominant stabilizing effect of atmospheric thermal stratification governs both linear and nonlinear stability and ensures the long-term persistence of the monsoon. Thus, while climate change can substantially alter the South Asian monsoon, it is no longer considered a tipping element.

The abrupt development of the Somali Jet and monsoon onset are tied to a nonlinear boundary layer regime — distinct from classical Ekman balance — that emerges near the equator where absolute vorticity vanishes. This regime produces a precise scaling between boundary layer kinetic energy and the north-south pressure gradient, and together these studies show how critical nonlinear processes in the atmospheric boundary layer give rise to abrupt monsoon onset and slower retreat, with implications for understanding the predictability and timing of the monsoon.

Climate feedbacks — lapse-rate, water vapour, albedo, and cloud — depend strongly on where a forcing is applied and the particular forcing agent, not just global magnitude. High-latitude forcing generates substantially larger warming than tropical forcing, primarily through lapse-rate differences; and interhemispheric forcing asymmetries shift tropical circulation and the ITCZ through cloud and heat-transport changes. Methane also has lower climate efficacy than CO₂ because its forcing is concentrated differently in latitude and altitude, with implications for greenhouse gas accounting and climate change mitigation. Together, these results bear directly on understanding the regional risks and trade-offs of mitigation and solar geoengineering.

This work introduces "minimal chaotic models" — a framework identifying the specific physical processes responsible for chaos in finite-dimensional climate and geophysical systems. Necessary conditions for forced dissipative chaos are derived for a broad class of low-order models: in the simplest models of Rayleigh–Bénard convection, for example, chaos requires external forcing of temperature to coexist with momentum dissipation. The Volterra gyrostat, a recurring nonlinear system shared by many atmospheric models, is analysed in detail. By identifying the minimum conditions for chaos in these models, the results illuminate the physical origins of irregular dynamics across the wide range of geophysical systems — from convection to quasi-geostrophic flow — whose low-order models share a gyrostatic structure.

Spatial heterogeneity in the timing of the seasonal-mean wind-speed diurnal cycle — "diurnal smoothing" — provides generation smoothing over and above the conventional distance-based geographic smoothing. For sites separated by more than 200 km, differences in diurnal peak timing explain more of the variation in wind-speed correlation than distance alone; sites whose peaks differ by several hours are more likely to be uncorrelated. This effect, quantified here for the first time using India as a case study, is relevant globally wherever strong wind-speed diurnal cycles exist, with implications for wind-farm siting and grid integration.

Wind energy droughts are substantially more frequent than solar droughts of comparable intensity; wind-solar hybridization reduces combined drought risk, but the benefits vary by region — pronounced in South India where wind and solar generation are negatively correlated, negligible in Rajasthan. When one major region faces a drought, there is only a 10% probability of a simultaneous drought elsewhere, highlighting the value of interregional grid connections. At the system level, a Pareto frontier analysis of wind-solar-storage tradeoffs shows that adding storage without expanding renewable capacity is inefficient, and that achieving high reliability within officially assessed renewable potential poses a fundamental challenge for fully decarbonized systems.

Seasonal mismatches between renewable energy supply and demand are a defining challenge for net-zero electricity systems. A global review of strategies — spanning long-duration storage, cross-sectoral flexibility, renewables overcapacity, and grid interconnection — finds that no single approach is adequate across all geographies and resource conditions. Each technology faces constraints of geography, cost, or self-discharge, and leveraging synergies among diverse flexibility options is essential for building resilient, deeply decarbonized grids.

The concept of "path independence" — that global temperature responds primarily to cumulative CO₂ emitted rather than to the rate or timing of emissions — underpins carbon budgets and net-zero targets. Ours are the first theoretical studies of its conditions: path independence requires the timescale over which cumulative emissions change to be short relative to the timescale over which the airborne fraction changes. When this fails, path independence breaks down — with implications for how carbon budgets should be constructed.