European farm households will face increasingly challenging conditions in the coming decades due to climate change, as the frequency and severity of extreme weather events rise. This study assesses the complex interrelations between external framework conditions such as climate change or adjustments in the agricultural price and subsidy schemes with farmers’ decision-making. As social aspects remain understudied drivers for agricultural decisions, we also consider value-based characteristics of farmers as internal factors relevant for decision-making. We integrate individual learning as response to extreme weather events into an agent-based model that simulates farmers’ decision-making. We applied the model to a region in Eastern Austria that already experiences water scarcity and increasing drought risk from climate change and simulated three future scenarios to compare the effects of changes in socio-economic and climatic conditions. In a cross-comparison, we then investigated how farmers can navigate these changes through individual adaptation. The agricultural trajectories project a decline of active farms between −27 and −37% accompanied by a reduction of agricultural area between −20 and −30% until 2053. The results show that regardless of the scenario conditions, adaptation through learning moderates the decline in the number of active farms and farmland compared to scenarios without adaptive learning. However, adaptation increases the workload of farmers. This highlights the need for labor support for farms.

由于气候变化，随着极端天气事件的频率和严重程度的上升，欧洲农户在未来几十年将面临日益严峻的条件。本研究评估了外部框架条件（例如气候变化或农产品价格和补贴计划的调整）与农民决策之间的复杂相互关系。由于社会因素仍然是农业决策的驱动因素，我们还没有充分研究，因此我们也将农民基于价值的特征视为与决策相关的内部因素。我们将个人学习作为对极端天气事件的响应整合到一个基于代理的模型中，该模型模拟农民的决策。我们将该模型应用于奥地利东部的一个地区，该地区已经经历了水资源短缺和气候变化导致的干旱风险增加，并模拟了三种未来情景，以比较社会经济和气候条件变化的影响。在交叉比较中，我们随后研究了农民如何通过个体适应来应对这些变化。农业轨迹预计，到 2053 年，活跃农场数量将减少 -27% 至 -37%，农业面积将减少 -20% 至 -30%。结果表明，无论情景条件如何，通过学习进行适应都会减缓农业面积的减少。与没有自适应学习的场景相比，活跃农场和农田的数量。然而，适应增加了农民的工作量。这凸显了农场对劳动力支持的需求。