Carbon storage plays a pivotal role in global initiatives aimed at curbing carbon emissions. Among the diverse reservoirs for carbon sequestration, agricultural land emerges as a substantial and underutilised carbon sink. In this context, nature-based solutions, especially the practice of carbon farming, stand out as a strategic means to enhance climate change mitigation efforts. This project explores the untapped potential of agricultural landscapes as a carbon storage mechanism and examines the transformative impact of carbon farming practices. Firstly, DS-CHANGES aims to quantitatively assess and identify the key factors influencing the adoption of carbon-farming techniques by local farmers. Our approach involves creating a unique georeferenced dataset by combining soil characteristics and climate data from Fondazione Cariplo with RICA longitudinal data on Lombardy – and Italian – farms. This dataset will serve as the foundation for our analysis, which consists in employing state-of-the-art econometric and machine learning techniques to understand the contribution of each driver to carbon-farming adoption. Additionally, we will assess the profitability of carbon farming compared to alternative practices, thereby generating estimates for carbon credit pricing that would render carbon farming financially sustainable. This analysis will further provide valuable insights into the effectiveness of existing policies, reveal the need for potential new policies, and estimate behavioural parameters influencing farmers' adoption decisions in response to various drivers, including climate and institutional factors. In a subsequent stage, DS-CHANGES will capitalise on the data and behavioural insights acquired earlier to construct a fully data-driven agent-based model. The model will be initialised using real-world data, and agent behaviour will be calibrated entirely based on empirically derived behavioural rules. This model will be dynamically integrated with a bio-physical submodule, enabling precise crop yield predictions under diverse climatic conditions. It will be employed for conducting mid-century projections, taking into account various climatic patterns, agricultural market dynamics, and policy scenarios.