Harnessing climate data to assess, anticipate and attribute climate impacts on agriculture

Human-induced climate change already impacts agriculture and food security, with varying impacts on regions, communities and individuals. There is an urgent need to increase agricultural resilience to ensure food security for a growing world population under humaninduced climate change. Climate data is a necessary resource not only to better understand climate impacts but also to manage and minimise current and future negative climate impacts on agriculture. However, there are large uncertainties in observations of past climate and simulations of climate under different emission scenarios leading to uncertainties in climate information for agricultural management. This is particularly the case in regions with complex climatology and low capacity for climate research. This dissertation presents three studies that demonstrate how publicly available observational and simulated climate data, in combination with crop models, can be robustly used to assess, anticipate and attribute climate impacts on agriculture, thereby contributing to increased resilience. The case studies focus on West Africa and Kazakhstan, two regions with distinct farming systems, both of which are critical for food security. The regions are at the same time highly vulnerable to climate change and lack climate information.