Airborne Geophysical and Borehole Data Fusion to Improve Mississippi River Valley Alluvial Aquifer Characterization

Airborne electromagnetic (AEM) data often fills the lithologic gap between boreholes and produces large‐scale structure and heterogeneity of aquifers. However, due to the nature of AEM data, there are unavoidable uncertainties in its nonunique interpretation. This study introduces a novel framework to seamlessly interpret and integrate AEM resistivity data with boreholes for high‐resolution aquifer characterization. Hierarchical agglomerative clustering (HAC) is employed to optimize depth‐dependent zonation thresholds by nearly collocated borehole data for AEM interpretation to lithology. Then, indicator cokriging (ICK) correlates borehole data (primary data) with AEM resistivity data (secondary data) for data fusion and estimates the probabilities of sand and clay facies. The ICK eliminates the mismatch between interpreted AEM data and borehole data and reduces interpretation uncertainty and blurry subsurface characterization using only AEM. The framework is applied to Mississippi River Valley alluvial aquifer (MRVA) characterization, and results are compared with a regional groundwater model conceptualization. The results show that many landforms are potential recharge zones and indicate that MRVA is highly accessible and prolific. MRVA has significant hydraulic connections with the carbonate Ozark aquifer and the sedimentary Mississippi embayment aquifer system. Moreover, MRVA is well connected to the Mississippi River, represented by high riverbed resistivity and high sand percentage at the riverbed. The data fusion framework maximizes the utilization of the AEM data and can significantly improve the regional groundwater model with aquifer characteristics.