KARST SCIENCE AND BIG DATA: ADVANCES AND CHALLENGES IN MONITORING AND STUDYING KARST SYSTEMS

In recent years, with the concurrent advancement of technology and understanding of karst system processes, the collection of, and need for, large, high-resolution datasets in karst research continues to evolve. The progression of environmental monitoring technology has ushered karst geoscience into the “Big Data” realm, enabling users to collect and aggregate millions of data points from a variety of sources, providing both new insights and data collection challenges. In karst areas, threats from geohazards include contaminated groundwater, flooding, sinkhole development, drought, and other hydrologically-driven issues. In order to remediate and mitigate these problems, the use of advanced monitoring equipment (e.g., data loggers, real-time data transmission, etc.) to be able to not only detect instantaneous changes in parameters, but also to build long-term, high-resolution baseline datasets from which to understand subtle changes in the system and develop predictive models is possible. In the field, sampling resolutions are also improving, due to increased capacity for field analyses of certain parameters, such as dye detection and analyzing bacterial loadings. Additionally, the study of karst processes can now generate enough data to inform complex hydrologic neural network and adaptive learning models, as well as calculate with high accuracy carbon flux, contaminant loadings, and volumetric flooding risks, among others. Collectively, these scenarios raise the question of the most practical, accurate, and affordable sampling resolution for different research and monitoring needs in karst research. Here, the advances and challenges of amassing large datasets in karst systems are discussed using several case studies, including the advantages and disadvantages of large datasets, types of software and applications, and their effect on interpreting karst systems.