Application and development trend of geostatistics in the research of spatial variation of aquifer parameters

Abstract

Scientific and reasonable evaluation of groundwater resources is essential for overall planning, rational development and utilization of regional groundwater, and ensuring the safety of regional ecological environment.Obtaining spatial heterogeneous distribution information of aquifer characteristics is a critical first step in resolving a variety of groundwater issues, such as seepage, pollution transport, groundwater development and exploitation.The heterogeneity of aquifers, however, is difficult to properly define due to the limitations of traditional survey equipment.Two-point geostatistics determines the correlation of random variables through variogram, solves the spatial linear estimation of geological variables and characterizes their anisotropy.Multi-point geostatistics breaks through the limitation of spatial correlation between two points, and effectively reflects the spatial distribution characteristics of aquifer parameters through multi-point training image modeling, which is also more suitable for simulating complex geological bodies.Based on this, the paper briefly describes and discusses the commonly used two-point geostatistics in the assessment of the spatial variation of aquifer parameters.Furthermore, the hydraulic conductivity is utilized as a medium to summarize the restricted synergistic relationships between hydraulic conductivity and electrical resistivity, hydraulic gradient or hydraulic head in two-point geostatistics.Besides, the advantages of multi-point geostatistical modeling are summarized after being compared with traditional geostatistical modeling.The unsolved problems and future development direction by its own algorithms and modeling methods are also discussed.Meanwhile, it is also pointed out that under the background of the rapid development of satellite, radar and remote sensing technology, the arrival of geological big data era shows a general trend that multi-source, multi-spatial and multi-resolution spatial data can be integrated and scale-driven by data assimilation, machine learning and other methods to help geostatistics achieve numerical modeling.

Keywords

• aquifer parameters
• geostatistics
• spatial variability
• geological big data