In austral spring 2016 the Antarctic region experienced anomalous sea ice retreat in all sectors, with sea ice extent in October and November 2016 being the lowest in the Southern Hemisphere over the observational record (1979–present). The extreme sea ice retreat was accompanied by the wettest and warmest spring on record, over large areas covering the Indian ocean, the Ross Sea, and the Weddell Sea.

The present study aims at explaining how intermittency (i.e., the alternation of dry and rainy periods) affects the rate at which precipitation extremes increase with temperature. Using high-resolution rainfall data from 99 stations in the United States, we show that at scales beyond a few hours, intermittency causes rainfall extremes to deviate substantially from Clausius–Clapeyron. A new model is proposed to better represent and predict these changes across scales.

The prediction of the El Niño phenomenon, an increased sea surface temperature in the eastern Pacific, fascinates people for a long time. El Niño is associated with natural disasters, such as droughts and floods. Current methods can make a reliable prediction of this phenomenon up to 6 months ahead. However, this article presents a method which combines network theory and machine learning which predicts El Niño up to 1 year ahead.

We find that CMIP5 models show more significant improvement in predicting zonal winds with initialisation than without initialisation based on the knowledge that zonal wind indices can be used as potential predictors for the EASM. Given the initial conditions, two models improve the seasonal prediction skill of the EASM, while one model decreases it. The models have different responses to initialisation due to their ability to depict the EASM–ESNO coupled mode.

State-of-the-art climate models sometimes differ in their prediction of key aspects of climate change. The technique of emergent constraints uses observations of current climate to improve those predictions, using relationships between different climate models. Our paper first classifies the different uses of the technique, and continues with proposing a mathematical justification for their use. We also highlight when the application of emergent constraints might give biased predictions.

The 100-year GWP is the most widely used metric for comparing the climate impact of different gases such as methane and carbon dioxide. However, there have been recent arguments for the use of different timescales. This paper uses straightforward estimates of future damages to quantitatively determine the appropriate timescale as a function of how society discounts the future and finds that the 100-year timescale is consistent with commonly used discount rates.

Using a dynamical climate model purely reduced from the conservation laws of ice-moving media, we show that ice-sheet physics coupled with a linear climate temperature feedback conceal enough dynamics to satisfactorily explain the system response over the full Pleistocene. There is no need, a priori, to call for a nonlinear response of, for example, the carbon cycle.

Earth system models provide simplified accounts of human–Earth interactions. Most current models treat CO₂ emissions as a homogeneously distributed forcing. However, this paper presents a new parameterization, POPEM (POpulation Parameterization for Earth Models), that computes anthropogenic CO₂ emissions at a grid point scale. A major advantage of this approach is the increased capacity to understand the potential effects of localized pollutant emissions on long-term global climate statistics.

The El Niño–Southern Oscillation phenomenon is a slow dynamics present in the coupled ocean–atmosphere tropical Pacific system which has important teleconnections with the northern extratropics. These teleconnections are usually believed to be the source of an enhanced predictability in the northern extratropics at seasonal to decadal timescales. This question is challenged by investigating the causality between these regions using an advanced technique known as convergent cross mapping.

We determine the point of no return (PNR) for climate change, which is the latest year to take action to reduce greenhouse gases to stay, with a certain probability, within thresholds set by the Paris Agreement. For a 67 % probability and a 2 K threshold, the PNR is the year 2035 when the share of renewable energy rises by 2 % per year. We show the impact on the PNR of the speed by which emissions are cut, the risk tolerance, climate uncertainties and the potential for negative emissions.

Results show that an additional 6.97 million people will be exposed to droughts in China under a 1.5 ºC target relative to reference period, mostly in the east of China. Demographic change is the primary contributor to exposure. Moderate droughts contribute the most to exposure among 3 grades of drought. Our simulations suggest that drought impact on people will continue to be a large threat to China under the 1.5 ºC target. It will be helpful in guiding adaptation and mitigation strategies.

Climate change projections of temperature extremes are particularly uncertain in central Europe. We demonstrate that varying soil moisture–atmosphere feedbacks in current climate models leads to an enhancement of model differences; thus, they can explain the large uncertainties in extreme temperature projections. Using an observation-based constraint, we show that the strong drying and large increase in temperatures exhibited by models on the hottest day in central Europe are highly unlikely.

Turbulent fluxes represent an efficient way to transport heat and moisture from the surface into the atmosphere. Due to their inherently highly complex nature, they are commonly described by semiempirical relationships. What we show here is that these fluxes can also be predicted by viewing them as the outcome of a heat engine that operates between the warm surface and the cooler atmosphere and that works at its limit.