Geostatistical simulations of the spatial variability of hydraulic conductivity in an alluvial-marine sedimentary system in Beihai City, China

Abstract

Sustainable aquifer management in coastal cities is becoming critically important due to the challenges of climate variability amidst population growth. Study of the aquifer heterogeneity is effective in highlighting the aquifer behavior. In this study, we took the case of Beihai to explore the heterogeneity of its subsurface system. 8629 samples from 111 boreholes were divided into four hydrofacies units as discrete geological variables according to their hydraulic conductivities (K) and used as input datasets for the models. Sequential Indicator Simulation (SISIM) and Transition Probability Geostatistical Simulation (T-PROGS) were employed to construct 3D hydrofacies stochastic models. The simulated results realistically represent horizontal river-driven sediment distribution and vertical deposition processes of K in the aquifer, where the northeastern part is mainly composed of low K and the southern part forms the main multi-layered aquifer system with progressively increasing medium and high K. With a comparative analysis on the ability of capturing geological characteristics using both SISIM and T-PROGS, it highlighted that SISIM reduces initial data requirements without the need of Gaussian transformation and is more broadly applicable, while T-PROGS prefers considering the internal correlation of hydrofacies units and is applicable in environments where lateral accumulation is prevalent. This study demonstrates the credibility of stochastic simulation methods in recreating complex costal sedimentary environments, which provide environment for further groundwater modeling and exploration. Multi-interleaved source data and more accurate algorithms will be integrated in aquifer heterogeneity as subsequent research priorities.