Groundwater resources are limited and difficult to predict in crystalline bedrock due to heterogeneity and anisotropy in rock fracture systems. Municipal-level governments often lack the resources for traditional hydrogeological tests when planning for sustainable use of water resources. A new methodology for assessing groundwater resources potential (GRP) based on geological and topographical factors using principal component analysis (PCA) and analysis of variance (ANOVA) was developed and tested. ANOVA results demonstrated statistically significant differences in classed variable groups as well as in classed GRP scores with regard to hydrogeological indicators, such as specific capacity (SC) and transmissivity. Results of PCA were used to govern the weight of the variables used in the prediction maps. GRP scores were able to identify 79% of wells in a verification dataset, which had SC values less than the total dataset median. GRP values showed statistically significant correlations using both parametric (using transformed datasets) and non-parametric methods. The method shows promise for municipal or regional level planning in crystalline terrains with high levels of heterogeneity and anisotropy as a hydrogeologically and statistically based tool to assist in assessing groundwater resources. The methodology is executed in a geographic information systems environment, and uses often readily available data, such as geological maps, feature maps and topography, and thus does not require expensive and time-consuming aquifer tests.

Hydrogeology in crystalline rock aquifers is often problematic due to the heterogeneity and anisotropy in thefracture network. Kinematic porosity of the host rock is exceedingly important for municipal decision makers inassessing sustainably extractable water supply volumes and assessing contaminant transport behavior within thematrix. This study explores heteroscedasticity in the hydrogeological characteristics of the fracture network andestimation of kinematic porosity from superficial fracture measurements. Estimates were based on the geometricalproperties of the fractures including: fracture frequency, aperture and orientation. The estimates wereadjusted for aperture changes with depth, connectivity of the fracture network, fracture continuity and measurementorientation bias. The results were compared with well archive data and correlations were found to besignificant with more than 95% confidence. Erratic behaviour of well data relative to fracture measurementsindicates that well orientation with respect to the fracture network gives incomplete hydrogeological data.Spatial heterogeneity of the bedrock was examined using spatial statistics and geographic information systems.The results from the spatial statistical analyses of well data showed that the heterogeneity within the bedrock issufficiently high that spatial correlations cease to exist in nearly all investigated rock types at distances greaterthan 500 m, and in some rocks, particularly sedimentary gneisses, no spatial correlations were observed.Arbitrarily grouped samples with similar geology and topography showed evidence of non-stationary variance.Results indicate that regional generalizations based on sparse point measurements are highly error prone andpotential exists in complementary field-based estimates.

Knowledge about regolith thickness is important in several civil and environmental engineering fields. However, subsurface characteristics such as regolith thickness are difficult to determine through surface investigations and maps at regional scales. This paper presents four methods for estimating regolith thickness in a GIS environment for previously glaciated terrain with high frequency of rock outcrops: linear regression (LR) using topographical covariates; inverse distance weighting (IOW) interpolation of regolith thickness point data from well drillings: a trigonometrical approach (TA) developed for this study which uses outcrop slopes and distance between outcrops; and a simplified regolith model (SRM). The SRM is a model modified from TA which estimates the regolith thickness based on outcrops, slopes and the distance to outcrops in eight directions. The methods were compared for three study areas (Tyreso, Vallentuna and Osteraker) in Stockholm County, Sweden. Based on the results in this paper, LR proved to be the most accurate method for regolith thickness estimation, measured through root mean square error values. Whereas IDW was the most accurate method in terms of error within 2 m, which would make it a suitable model if and when large datasets of regolith point data are available. When drilling data is scarce then both the TA and SRM methods can be used for regolith estimations. However, the SRM proved to be a more accurate regolith thickness model compared to TA. SRM shows promising results and could be used at a preliminary stage in engineering projects where little or no data is available prior to detailed field investigations in previously glaciated terrain.

In terrains with limited soil cover and groundwater storage, groundwater resource management is governed by the spatial nature of storage, recharge and extraction. Local soils may act as important groundwater reservoirs for residents which have no other feasible water supply. A novel groundwater balance methodology is presented which accounts for the spatial distribution of storage and extraction. Existing topographical and geological databases as well as well data were used to construct a conceptual model of the groundwater system, assuming stratigraphy based on typical geology. The method is implemented in a geographic information systems environment and allows for variable climate and land use scenarios.

Several scenarios were examined with this method, demonstrating that on a regional scale average reservoir volumes meet demand but at the local levels depletion of reservoirs may be experienced. Groundwater level drawdown in excess of 50% of the projected reservoir storage were seen, particularly near the coast. Soil-filled valleys may act as local hydraulic barriers, preventing contamination from saline water provided no direct hydraulic connection is present. The method demonstrates the importance of a spatial approach in managing groundwater resources, and shows promise as a tool for planners in increasing water security.

Limited groundwater storage in recently glaciated terrains with frequent hard, crystalline bedrock outcrops poses challenges for groundwater resources management. Due to often-limited economic resources allocated to groundwater investigations in areas where drinking water is primarily supplied by private wells, heuristic solutions such as groundwater balances with built-in limitations of storage may serve a vital role in improving water security. This study investigates the use of a conceptual groundwater balance model and the use of a conceptual-statistical reservoir vulnerability model with this aim. A limited storage, GIS-based groundwater balance model using existing databases was applied to an area outside of Stockholm, Sweden with existing groundwater level measurement data. The spatial model showed improved performance over existing S-Hype model estimates of groundwater levels currently used in groundwater resources management by the Geological Survey of Sweden, even without local calibration. Differences between two wells with time series data showed evidence of strong influence due to in-situ geological conditions. Groundwater vulnerability estimates correlated significantly with chloride measurements from an existing chemistry database. The performance of the conceptually-based spatial groundwater balance supports the use of the approach as an aid for municipal planners and decision-makers in moving towards sustainable groundwater resources planning and improving water security in areas with limited storage and a large number of dispersed, private wells.