Item

Abstract

Data integration and analytics platforms play a critical role in better understanding natural systems and laying the foundations for modeling them as complex adaptive systems—for example, how carbon and hydrological cycles can track the functionalities of (agro)ecosystems. Also, the resultant databases and other digital resources are the precursors of knowledge graphs. Nonetheless, data-driven models have become widely used in water resource planning and management; these models require datasets that incorporate the diversity and variability of patterns that occur in nature, including those driven by land use, crop management, and climate variability and change, for model generalization. Further, the use and integration of data can be constrained to a few sources due to the availability of datasets and the time, skills, and efforts required to blend multi-disciplinary heterogeneous datasets. The main goal of this study is to integrate biophysical variables (crop, land cover and soil information) to the recently designed open-source web platform. The objectives are to (1) design a job schedular for data retrieval at the regular temporal interval, (2) data preprocessing, quality control, and data integration pipelines, (3) design of NoSQL (MongoDB) database for a flexible, efficient, and reliable data storage for heterogeneous and semi-structured datasets that support parallel processing and complex geospatial queries for data analytics. We expect to foster the exploration of the underlying biophysical processes associated with the intensity of the intertwined carbon and water cycle and eventually a better management of carbon budgets by developing and training data-driven models and analytics for stakeholders and decision-makers.