Extraordinary events such as large volcanic eruptions or nuclear war could cause sudden global climate disruptions and affect food security. Past volcanic eruptions, like the 1783 Laki eruption in Iceland and the 1815 Tambora eruption in Indonesia, caused significant global cooling via sulfuric acid aerosol injections into the stratosphere, resulting in famines and political instability. A nuclear war would cause similar, but potentially far more devastating, effects. Bombs detonated over urban and industrial areas would ignite firestorms, injecting massive amounts of soot into the upper atmosphere. This soot would spread globally, causing a rapid and significant cooling of the planet, lasting for decades. This “nuclear winter” would severely disrupt terrestrial and aquatic food production systems. Previous studies in the 1980s investigated nuclear winter's impact on agriculture, but limitations in data and models prevented comprehensive analysis. Recent advancements in climate, crop, and fishery models, coupled with updated data on nuclear arsenals and weapon yields, allow for more accurate estimations. A war between India and Pakistan could release 5–47 Tg of soot, while a conflict between the United States and Russia could release over 150 Tg, potentially triggering a full-scale nuclear winter. While smaller-scale nuclear conflicts would have less severe global impacts, the risk of escalation remains extremely high once a war begins. This study addresses the lack of integrated estimates of the impacts of a range of nuclear war scenarios on both land-based and ocean-based food production. We analyze six scenarios, ranging from 5 Tg to 150 Tg of soot injection, examining their consequences on food supply. The scenarios presented are illustrative examples, and similar soot injection levels could result from other nuclear conflicts involving countries like China, France, the United Kingdom, North Korea, and Israel. Recent large-scale wildfires (Canada 2017, Australia 2019-2020) produced 0.3-1 Tg of smoke, demonstrating that the soot injection and global dispersal modeled in this study are plausible. While localized contamination of soil and water near detonation sites is anticipated, the global spread of soot once it reaches the upper atmosphere renders the impacts globally relevant, regardless of the warring nations. Our focus here is on the climate disruption and its ensuing effects on global food production systems.

Several recent studies have individually analyzed the impact of regional nuclear war on major grain crops and marine wild-catch fisheries using climate, crop, and fishery models. However, a comprehensive, integrated assessment of the impact across the entire range of potential war scenarios on both land and ocean-based food production was missing. This study addresses this gap by using state-of-the-art models and considering the cascading effects of climate change on the entire food system, encompassing crops, fisheries, and livestock, as well as potential adaptations such as food waste reduction and shifting livestock feed to human consumption.

The study employed a multi-model approach to assess the impact of nuclear war on global food security. Six scenarios were considered, involving varying amounts of stratospheric soot injection (5 Tg to 150 Tg), each corresponding to different nuclear war scenarios. The Community Earth System Model (CESM) was utilized to simulate climatic changes (temperature, precipitation, solar radiation) resulting from soot injection. These climatic changes then served as inputs for two other models: the Community Land Model version 5 crop (CLM5crop) to assess crop productivity (maize, rice, spring wheat, soybean), and the BiOeconomic mArine Trophic Size-spectrum (BOATS) model to simulate marine wild-catch fisheries. CLM5crop was selected for its ability to simulate rainfed and irrigated crops separately, providing comprehensive assessments under various water availability conditions. The model was validated against FAO data (1991-2010 and 2006-2018) to ensure accuracy in simulating the yields of the four main crops. The model was driven by climate model outputs, without bias correction. This choice allows for a more realistic representation of the changing variance in climate variables under the extreme conditions of nuclear war scenarios. The BOATS model was chosen for its capacity to calculate the size-structured biomass of commercially targeted fish, accounting for factors like fishing effort and fish price. It was driven by climate model-generated sea surface temperature and net primary production changes. The combined results of CLM5crop and BOATS provided estimates of calorie production changes for major crops and marine fish. To evaluate overall food availability, additional factors were considered, including the proportion of crops used for non-food purposes (e.g., livestock feed), the composition of the global diet (FAO data from 2010), and potential societal responses. Three scenarios were explored: Livestock (minimal adaptation, continued livestock production), Partial Livestock (50% of livestock feed used for human consumption), and No Livestock (livestock production ceases after the first year). The impact of household food waste reduction was also assessed. Finally, the model assumed the cessation of international food trade, to reflect the likely disruption of global food markets after a nuclear war. Nation-level calorie availability was calculated for each scenario and each year post-war, combining production of simulated crops and marine fish with estimations of changes in other food types based on FAO data and assumptions about societal responses.

The study's key findings highlight the catastrophic consequences of nuclear war on global food security, with the severity increasing dramatically with the magnitude of soot injection. Even a relatively small soot injection (5 Tg) resulted in a 7% decrease in global average crop calorie production in years 1-5 post-war. Larger injections caused far more significant declines, with a nearly 90% reduction in the 150 Tg scenario. While marine fish catches also declined, the reduction was less severe than that of crop production due to the ocean's greater heat capacity and more moderate reductions in primary productivity. The dominance of terrestrial crop production in total food calorie production means that changes in crops outweigh the changes in fish. The combined impact of reduced crop and fish production is further exacerbated by the significant portion of crops used for animal feed (maize and soybean are primarily used for animal feed). Even with adaptations like shifting livestock feed to human consumption, these adaptations alone are insufficient to offset the massive reductions in overall crop and fish yields under larger soot injection scenarios. Regional differences in food production were observed, with mid- to high-latitude regions in the Northern Hemisphere experiencing the most significant percentage reductions in crop calorie production. The study projected massive food shortages leading to widespread starvation, even under moderate soot injection levels (27 Tg). In this scenario, mid- to high latitudes of the Northern Hemisphere showed reductions in crop calorie production of greater than 50%, while fish catches declined by 20–30%. Nuclear-armed nations in mid- to high latitudes faced calorie reductions ranging from 30% to 86%, while lower-latitude nations experienced less severe, though still significant, impacts. Analysis of calorie intake levels revealed alarming consequences. Under the 150 Tg scenario, most nations experienced calorie intake below the basal metabolic rate, leading to widespread death from starvation. Only Australia and New Zealand experienced less severe impacts due to their relatively high wheat production and favorable climatic conditions for wheat. However, even these regions would likely face an influx of refugees from other areas suffering from food insecurity. Even with optimistic assumptions regarding food waste reduction and equitable global distribution, the model predicted insufficient food for the global population under the larger soot injection scenarios. The most optimistic case, considering no food waste and equitable global distribution with a complete switch to human consumption of livestock feed, would still not support the global population under a 47 Tg scenario. The study estimates that a nuclear war between India and Pakistan could result in more than 2 billion deaths due to famine, while a war between the United States and Russia could lead to over 5 billion deaths.

The study's findings strongly underscore the catastrophic potential of nuclear war on global food security. The results demonstrate that even with potential compensatory behaviors (such as converting livestock feed to human consumption and reducing food waste), the substantial climate-driven reductions in crop and pasture production cannot be fully offset, particularly under scenarios involving large atmospheric soot injections. The analysis emphasizes the vulnerability of countries dependent on food imports, as illustrated by the severe impact on African and Middle Eastern countries under the no-trade assumption. The study's impact on protein supply is also highlighted, which further underscores the severity of the situation. The findings stress the fragility of the global food system and the interconnectedness of nations in the face of such a catastrophic event. The study's projections provide further support to the widely held view that a nuclear war is unwinnable and must be avoided at all costs.

This study demonstrates through state-of-the-art modeling that nuclear war would trigger a global food security catastrophe, resulting in massive loss of life. Even with potential adaptive measures, the reductions in crop and fish production are overwhelmingly large to offset. The results underscore the urgent need for international cooperation to prevent nuclear conflict. Further research could focus on more detailed modeling of adaptive strategies, including exploring alternative food sources and the impacts of radioactive contamination and changes in the insect community. Economic models would be needed to better understand food distribution under a no-trade scenario.

Several limitations exist within this study. The use of a single Earth system model, a single crop model, and a single fishery model could underestimate variability in the model outputs. The study assumes the cessation of international trade, and a more sophisticated model considering economic factors and the dynamics of food distribution systems is needed. Moreover, the study does not directly address the effects of reduced human population due to direct or indirect mortality following a nuclear war, and thus the available labor and calorie distribution dynamics are not fully captured. The impacts of elevated ultraviolet radiation and the role of adaptation strategies like shifts in cultivar selection or alternative food sources are only partially considered and warrant further investigation.