Collaborative Research: Data Fusion for Characterizing and Understanding Water Flow Systems in Karst Aquifers

Approximately 15% of the ice-free land surface globally is underlain by karst terrain and approximately 20 – 25% of the global populations depends largely or entirely on the associated aquifers with many large population centers relying on these aquifers as primary water supplies. While these systems are of critical importance, they have a variety of characteristics that make them highly vulnerable to both climate change and contamination. Karst systems have distinct physiographic features that result from dissolution of bedrock, typically carbonate rocks, and as a result have a wide range of permeability values and preferential flow paths that connect surface and subsurface systems. Because of the complex nature of karst systems, many methods (e.g., direct conduit mapping, tracer tests, geophysical investigations, etc) have been implemented in characterizing karst aquifer systems. These methods have provided valuable information but with significant limitations, especially in characterizing human-inaccessible small fractures and conduits that are critical to water flow and contaminant transport. We propose a collaborative research using a complementary data fusion approach. The approach infuses data collected from hydraulic tomography, river stage tomography, electrical resistivity tomography, and tracer tests to yield a more reliable map of fractures and conduits in karst aquifers in a cost-effective manners and in turn, better understanding and prediction of flow and solute transport in the karst aquifers. The overarching goal of this project is to develop, test, and validate an innovative approach and technology for characterizing 3-D karst aquifers in detail based on the data fusion concept. To achieve this goal, we will conduct two innovative field surveys: surface and subsurface river stage tomography and electrical resistivity tomography with moving current sources. We will also explore the feasibility of natural lightning tomography for large-scale resistivity surveys. The data collected from these surveys will be integrated into a geostatistical-based inversion framework to characterize distribution and morphology of fractures and conduits in the karst aquifer with great details. The fusion approach will be tested and validated at the Cane Run Royal Spring Basin in central Kentucky. The validity of the characterized karst aquifers will be evaluated using a separate model with field collected water level, water chemistry, tracer, and stable isotope data.