CarboFarm: Data Integration and Knowledge Generation for Agricultural GHG Inventories

Context : The agricultural sector suffers the consequences of global warming and climate change. However, it is also one of the top global emitters of GHG. There is a need to propose new solutions that provide more sustainable agriculture, and an important step in this direction is the generation of GHG (Greenhouse Gas) inventories. Aims : This work presents an architectural proposal, called CarboFarm, to integrate (syntactically and semantically) heterogeneous datasets related to agriculture. The aim is to support the generation of greenhouse gas inventories on farms. Integrated data can also contribute to generating knowledge to support rural landowners' decision-making and generating carbon credits. Methods : We use the Design Science Research (DSR) method to develop CarboFarm. Using machine learning and semantic modeling techniques, we generate knowledge to support rural property owners' decision-making. In addition, CarboFarm aims to provide information to decision support applications for rural landowners. Results : Data analysis through machine learning techniques could identify patterns, trends, and provided insights that can be useful for decision-making. To support our approach, we carried out a case study integrating datasets of GHG emissions and stocks for Brazilian rural properties. Conclusion : The proposal offers alternatives for using land focusing on a positive GHG balance, which can contribute to generating carbon credit. Significance : There are many challenges in building systems that generate GHG inventories for rural properties. A comprehensive solution requires expertise from many areas of knowledge. From a technological perspective, a software solution must integrate data, generate knowledge, and provide decision support for rural producers. The literature review revealed the lack of studies that address these issues in an integrated manner. Furthermore, we did not find any studies that promote semantic integration and analysis of agricultural data to support decisions to generate GHG inventories.