Research software developed by domain experts enables scientific research by addressing domain-specific problems tailored to the domains and constraints of their respective fields. As such software evolves and its functional and quality requirements increase, applying software engineering practices becomes essential to manage the software system. An example of such research software is OptiGob, a decision-support tool to project CO2e emissions among other properties for Ireland's agriculture, forestry, and other land use (AFOLU) sector. In this paper, we present the software engineering practices applied during the evolution of OptiGob to handle newly emerging requirements and a model-driven approach to establish a foundation for future extensions such as additional AFOLU systems and characteristics in the dataset and new application features. Our solution is based on MontiGem, a code generator for data-centric applications. This new version of OptiGob enables users to build and evaluate more complex AFOLU scenario configurations while managing the growing complexity of the software through model-driven engineering.