Global synthesis of peer-reviewed articles reveals blind spots in climate impacts research

Understanding how climate hazards such as floods, droughts, and storms turn into disasters requires detailed socioeconomic impact data. While extensive research documents these impacts, the rapid growth of the literature hinders effective synthesis. Here, we present the first global stocktake of scientific literature on the socioeconomic impacts of climate hazards by systematically screening over 59,000 open-access articles using machine learning. We find significant regional biases in how impacts are documented: disasters in low-income countries must cause about 16 times more deaths and affect 130 times more people to receive the same scientific attention as those in high-income countries. When adjusting for the population exposure to climate hazards, about 32\% of the countries in the Global South suffer from under-reporting, compared to only 16\% in the Global North. Moreover, indirect and cascading impacts-such as disruptions to communication systems, supply chain breakdowns, and societal conflicts-are rarely studied. These blind spots underscore the need for more inclusive and equitable research efforts that prioritize vulnerable regions to better understand and prevent the impacts of climate hazards.