Water footprint variations in major food crops triggered by different drought scenarios from a probabilistic perspective

Drought is an intermittent disturbance of the water cycle that threatens food security. Understanding the impact of drought on water use in crop production provides valuable insights for effective water resource management. The crop water footprint (WF) captures the type and quantity of water utilization throughout the growth cycle. Previous studies have used mainly deterministic approaches to assess the effects of drought on crop WFs, with less focus on probabilistic methods. This study utilized probabilistic methods, choosing WF and standardized precipitation evapotranspiration index (SPEI) as the indices, to evaluate the responses of major crop WFs to drought in Northeast China from 2001 to 2021. The results revealed that most crop WFs had a positive correlation with the SPEI during the growing season, and the response timescale of crop WFs to drought was primarily at the 1-month scale. A greater likelihood of deteriorating drought conditions resulted in a lower WF_green. The probabilities of WF_green falling below the 40th percentile reached 68.5 %, 95.9 % and 41.7 % under extreme drought for maize, rice and soybean, respectively. The probability response of WF_total to drought differed across various crops. With increasing drought severity, the probability increased at all higher WF_total statuses for rice and soybean, whereas the probability increased only in the WF_total > WF_80th scenario for maize. This study is projected to offer valuable insights for decision-makers and stakeholders in mitigating the impacts of drought on water resources.