Assessing impacts of global natural disasters on agricultural value chains and adaptation strategies

Natural disasters are intensifying under climate change, posing significant threats to globally interconnected agricultural supply chains. However, the cascading impacts of multi hazard events and the performance of adaptation portfolios remain poorly quantified. This study develops a high-resolution agent-based network model that embeds 419 disaster events across 9 hazard categories into a global supply chain framework. Model simulations indicate that direct agricultural losses amounted to 8349 M.USD, while net indirect losses contributed an additional 3314 M.USD, accounting for 28.4% of total losses and exhibiting a delayed temporal pattern. Furthermore, trade-network diagnostics suggest a feed-livestock cascade pathway, wherein upstream crop and feed disruptions propagate through rigid market linkages and jointly accounted for 64.7% of gross positive indirect agricultural losses. Single disaster simulations show that drought generated the largest total loss and the largest absolute adaptation benefits, indicating that adaptation efficacy varies across hazard types. Scenario evaluations reveal that engineering measures yield the most substantial aggregate benefits, whereas combined portfolios provided limited incremental benefits once the strongest individual measure was already included. These findings highlight the heterogeneous propagation of disaster risks and the need for targeted adaptation portfolios across regions, sectors, and hazard types.