Search for: All records

Abstract Compound climate hazards, such as co-occurring temperature and precipitation extremes, substantially impact people and ecosystems. Internal climate variability combines with the forced global warming response to determine both the magnitude and spatial distribution of these events, and their consequences can propagate from one country to another via many pathways. We examine how exposure to compound climate hazards in one country is transmitted internationally via agricultural trade networks by analyzing a large ensemble of climate model simulations and comprehensive trade data of four crops (i.e. wheat, maize, rice and soya). Combinations of variability-driven climate patterns and existing global agricultural trade give rise to a wide range of possible outcomes in the current climate. In the most extreme simulated year, 20% or more of the caloric supply in nearly one third of the world’s countries are exposed to compound heat and precipitation hazards. Countries with low levels of diversification, both in the number of suppliers and the regional climates of those suppliers, are more likely to import higher fractions of calories (up to 93%) that are exposed to these compound hazards. Understanding how calories exposed to climate hazards are transmitted through agricultural trade networks in the current climate can contribute to improved anticipatory capacity for national governments, international trade policy, and agricultural-sector resilience. Our results highlight the need for concerted effort toward merging cutting-edge seasonal-to-decadal climate prediction with international trade analysis in support of a new era of anticipatory Anthropocene risk management. 

Abstract Seasonal‐to‐decadal climate prediction is crucial for decision‐making in a number of industries, but forecasts on these timescales have limited skill. Here, we develop a data‐driven method for selecting optimal analogs for seasonal‐to‐decadal analog forecasting. Using an interpretable neural network, we learn a spatially‐weighted mask that quantifies how important each grid point is for determining whether two climate states will evolve similarly. We show that analogs selected using this weighted mask provide more skillful forecasts than analogs that are selected using traditional spatially‐uniform methods. This method is tested on two prediction problems using the Max Planck Institute for Meteorology Grand Ensemble: multi‐year prediction of North Atlantic sea surface temperatures, and seasonal prediction of El Niño Southern Oscillation. This work demonstrates a methodical approach to selecting analogs that may be useful for improving seasonal‐to‐decadal forecasts and understanding their sources of skill. 

The observed increase in extreme weather has prompted recent methodological advances in extreme event attribution. We propose a machine learning–based approach that uses convolutional neural networks to create dynamically consistent counterfactual versions of historical extreme events under different levels of global mean temperature (GMT). We apply this technique to one recent extreme heat event (southcentral North America 2023) and several historical events that have been previously analyzed using established attribution methods. We estimate that temperatures during the southcentral North America event were 1.18° to 1.42°C warmer because of global warming and that similar events will occur 0.14 to 0.60 times per year at 2.0°C above preindustrial levels of GMT. Additionally, we find that the learned relationships between daily temperature and GMT are influenced by the seasonality of the forced temperature response and the daily meteorological conditions. Our results broadly agree with other attribution techniques, suggesting that machine learning can be used to perform rapid, low-cost attribution of extreme events.