Groundwater resources in hard rock coastal terrains
Time: Fri 2019-09-20 14.00
Location: F3, Lindstedtsvägen 26, Stockholm (English)
Subject area: Land and Water Resources Engineering
Doctoral student: Robert Earon , Vatten- och miljöteknik
Opponent: Professor Okke Batelaan, Flinders University, South Australia
Supervisor: Professor Bo Olofsson, Vatten- och miljöteknik
Abstract
Challenges regarding water security in hard rock coastal regions with limited soil cover are: a seasonal absence of recharge during times of peak residency, heterogeneity and variability of the fracture network, close proximity to saline water sources and spatially inconsistent storage and extraction. In areas where it is not feasible to connect residents to municipal water systems, a better understanding of the resilience of reservoirs is needed. The purpose of this study is to investigate and describe the spatial nature of hydraulic data in these types of terrains and present several novel GIS-based groundwater tools with the intent of increasing local water security and aiding in sustainable water resources management. Methods used in this study include groundwater balance modelling and conceptual groundwater storage modelling, as well as a combination of parametric and non-parametric statistical methods such as ANOVA, PCA, correlation and semivariogram analyses. Specific capacity estimates from the Geological Survey of Sweden’s well archive grouped by age or rock type showed very little autocorrelation and in assumed homogeneous geological regions showed statistically significant differences when arbitrarily grouped along a lineament. Estimates of kinematic porosity based on surface fracture data were found have statistically significant correlations with the well data. A GIS-based multivariate prediction tool for assessing Groundwater Resources Potential (GRP) was found to have statistically significant correlations with well data. The GRP method was then combined with a conceptual groundwater storage model and was subsequently found to have statistically significant correlations with chloride concentrations in well quality tests. The storage model was found to have a spatially-dependent sensitivity, meaning that different assumptions within the model had varying effects on the model depending on the geological settings. Incorporating the storage model into a spatial groundwater balance model was then compared with groundwater level time series data over a period of two years, where it was found to have a good explanative capacity and RMSE values of the storage ratio (0.06 to 0.34). Additionally, a soil depth model was developed, tested and found to produce promising results in regions with frequent rock outcrops, where up to 86% of estimates were within 2 m of actual soil depths. Conclusions from this study illustrate the need for a spatial approach to groundwater resources in these types of terrains, and demonstrate a strong potential of several new tools for quantity, capacity and vulnerability estimates to increase water security in a changing climate.