Search

Large-scale studies of the spatial and temporal variation of groundwater drought status require complete inventories of groundwater levels on regular time steps from many sites so that a standardised drought index can be calculated for each site. However, groundwater levels are often measured sporadically, and inventories include missing or erroneous data. A flexible and efficient modelling framework is developed to fill gaps and regularise data in such inventories. It uses linear mixed models to account for seasonal variation, long-term trends and responses to precipitation and temperature over different temporal scales. The only data required to estimate the models are the groundwater level measurements and freely available gridded weather products. The contribution of each of the four types of trends at a site can be determined and thus the causes of temporal variation of groundwater levels can be interpreted. Validation reveals that the models explain a substantial proportion of groundwater level variation and that the uncertainty of the predictions is accurately quantified. The computation for each site takes less than 130 s and requires little supervision. Hence, the approach is suitable to be upscaled to represent the variation of groundwater levels in large datasets consisting of thousands of boreholes.

Water resource management (WRM) practices, such as groundwater and surface water abstractions and effluent discharges, may impact baseflow. Here the CAMELS-GB large-sample hydrology dataset is used to assess the impacts of such practices on Baseflow Index (BFI) using statistical models of 429 catchments from Great Britain. Two complementary modelling schemes, multiple linear regression (LR) and machine learning (random forests, RF), are used to investigate the relationship between BFI and two sets of covariates (natural covariates only and a combined set of natural and WRM covariates). The LR and RF models show good agreement between explanatory covariates. In all models, the extent of fractured aquifers, clay soils, non-aquifers, and crop cover in catchments, catchment topography, and aridity are significant or important natural covariates in explaining BFI. When WRM terms are included, groundwater abstraction is significant or the most important WRM covariate in both modelling schemes, and effluent discharge to rivers is also identified as significant or influential, although natural covariates still provide the main explanatory power of the models. Surface water abstraction is a significant covariate in the LR model but of only minor importance in the RF model. Reservoir storage covariates are not significant or are unimportant in both the LR and RF models for this large-sample analysis. Inclusion of WRM terms improves the performance of some models in specific catchments. The LR models of high BFI catchments with relatively high levels of groundwater abstraction show the greatest improvements, and there is some evidence of improvement in LR models of catchments with moderate to high effluent discharges. However, there is no evidence that the inclusion of the WRM covariates improves the performance of LR models for catchments with high surface water abstraction or that they improve the performance of the RF models. These observations are discussed within a concep

This report presents the results and interpretation for Phase 5 of an integrated environmental monitoring programme that is being undertaken around two proposed shale gas sites in England – Preston New Road, Lancashire and Kirby Misperton, North Yorkshire. The report should be read in conjunction with previous reports freely available through the project website1 . These provide additional background to the project, presentation of earlier results and the rationale for establishment of the different elements of the monitoring programme.

This report describes the environmental baseline topsoil chemistry dataset collected in February-March 2018 (03-18) as part of the United Kingdom Geoenergy Observatories (UKGEOS) project. Ninety, samples were collected from the shallow coal-mine Glasgow Geothermal Energy Research Field Site (GGERFS). The report accompanies the GGERFS Soil Chemistry03-18 dataset. It provides valuable information on soil chemistry prior to installation of the GGERFS-facility boreholes, against which any future change during the development/ running of the facility can be assessed. This information is necessary to help understand and de-risk similar shallow geothermal schemes in the future, provide public reassurance, and inform sustainable energy policy.

This report details the steps taken in the process of producing risk (hazard) maps for chemical parameters listed in the European Commission Directive 98/83/EC on the quality of water intended for human consumption and the national Water Supply (Water Quality) Regulations that implement the requirements of the directive for drinking water in England and Wales. Amendments to 98/83/EC set out in Directive 2015/1787 provide the terms for reduced monitoring requirements by European Member States for drinking water where evidence indicates that waterquality risk is low. On the basis of the Water Safety Plan approach of the World Health Organization, DWI requires mapping of available data on raw-water sources in England and Wales to provide an evaluation of spatial distributions of the listed chemical parameters and their concentration ranges as evidence of risk for drinking water. An evaluation of temporal variability was also required to assess evidence for any trends to aid with decision making on future drinkingwater monitoring requirements. Data for an agreed list of 27 chemical parameters were collated, screened, evaluated and mapped, with surface water and groundwater being treated separately. This report details the data sources and steps taken to collate, evaluate, process and map them. Risk maps produced for individual parameters include expected values and 95th percentiles of measured values relative to the prescribed concentration or value (PCV) at any given location. The methodology employed required prediction of the entire statistical distribution of each parameter at each prediction location so that both expected value and percentile values for each parameter could be determined. This required the use of a statistical model to represent the variation of the data. The produced risk maps are produced for water-quality data analysed over the last three years, in line with the requirements of the 2015/1787 Directive. The correspondence between the two layers i

This report describes the results of activities carried out as part of the Environmental Monitoring Project (EMP) led by the British Geological Survey (BGS) in areas around two shale gas sites in England – Kirby Misperton (Vale of Pickering, North Yorkshire) and Preston New Road (Fylde, Lancashire). It focuses on the monitoring undertaken during the period April 2018–March 2019 but also considers this in the context of earlier monitoring results that have been covered in reports for earlier phases of the project (Phases I–IV) 2 . The EMP project is a multi-partner project involving BGS together with Public Health England (PHE), University of Birmingham, University of Bristol, University of Manchester, Royal Holloway University of London (RHUL) and University of York. The work has been enabled by funding from a combination of the BGS National Capability programme, a grant awarded by the UK Government’s Department for Business Energy & Industrial Strategy (BEIS) and additional benefit-in-kind contributions from all partners. The project comprises the comprehensive monitoring of different environment compartments and properties at and around the two shale-gas sites. The component parts of the EMP are all of significance when considering environmental and human health risks associated with shale gas development. Included are seismicity, ground motion, water (groundwater and surface water), soil gas, greenhouse gases, air quality, and radon. The monitoring started before hydraulic fracturing had taken place at the two locations, and so the results obtained before the initiation of operations at the shale-gas sites represent baseline conditions. It is important to characterise adequately the baseline conditions so that any future changes caused by shale gas operations, including hydraulic fracturing, can be identified. This is also the case for any other new activities that may impact those compartments of the environment being monitored as part of the project. In the per

Three-dimensional (3D) models are often utilised to assess the presence of sand and gravel deposits. Expanding these models to provide a better indication of the suitability of the deposit as aggregate for use in construction would be advantageous. This, however, leads to statistical challenges. To be effective, models must be able to reflect the interdependencies between different criteria (e.g. depth to deposit, thickness of deposit, ratio of mineral to waste, proportion of ‘fines’) as well as the inherent uncertainty introduced because models are derived from a limited set of boreholes in a study region. Using legacy borehole data collected during a systematic survey of sand and gravel deposits in the UK, we have developed a 3D model for a 2400 km2 region close to Reading, southern England. In developing the model, we have reassessed the borehole grading data to reflect modern extraction criteria and explored the most suitable statistical modelling technique. The additive log-ratio transform and the linear model of coregionalization have been applied, techniques that have been previously used to map soil texture classes in two dimensions, to assess the quality of sand and gravel deposits in the area. The application of these statistical techniques leads to a model which can be used to generate thousands of plausible realisations of the deposit which fully reflect the extent of model uncertainty. The approach offers potential to improve regional-scale mineral planning by providing an enhanced understanding of sand and gravel deposits and the extent to which they meet current extraction criteria.

A geostatistical survey for soil requires rational choices regarding the sampling strategy. If the variogram of the property of interest is known then it is possible to optimize the sampling scheme such that an objective function related to the survey error is minimized. However, the variogram is rarely known prior to sampling. Instead it must be approximated by using either a variogram estimated from a reconnaissance survey or a variogram estimated for the same soil property in similar conditions. For this reason, spatial coverage schemes are often preferred, because they rely on the simple dispersion of sampling units as uniformly as possible, and are similar to those produced by minimizing the kriging variance. If extra sampling locations are added close to those in a spatial coverage scheme then the scheme might be broadly similar to one produced by minimizing the total error (i.e. kriging variance plus the prediction error due to uncertainty in the covariance parameters). We consider the relative merits of these different sampling approaches by comparing their mean total error for different specified random functions. Our results showed the considerable benefit of adding close‐pairs to a spatial coverage scheme, and that optimizing with respect to the total error generally gave a small further advantage. When we consider the example of sampling for geostatistical survey of clay content of the soil, an optimized scheme based on the average of previously reported clay variograms was fairly robust compared to the spatial coverage plus close‐pairs scheme. We conclude that the direct optimization of spatial surveys was only rarely worthwhile. For most cases, it is best to apply a spatial coverage scheme with a proportion of additional sampling locations to provide some closely spaced pairs. Furthermore, our results indicated that the number of observations required for an effective geostatistical survey depend on the variogram parameters.

Increased groundwater abstraction is important to the economic development of Africa and to achieving many of the Sustainable Development Goals. However, there is little information on long-term or seasonal groundwater trends due to a lack of in situ monitoring. Here we used GRACE data from three products CSR, JPL-MCN and GRGS to examine Terrestrial Water Storage (TWS) changes in 12 African sedimentary aquifers, to examine relationships between TWS and rainfall from TRMM, and estimate Groundwater Storage (GWS) changes using four LSMs CLM2.0, VIC, MOSAIC and NOAH. We find that there are no substantial continuous longterm decreasing trends in groundwater storage from 2002 to 2016in any of the African basins, however, consistent rising groundwater trends amounting to ~1 km3/year and 1.5 km3/year are identified in the Iullemmeden and Senegal basins, respectively and longer term variations in ΔTWS in several basins associated with rainfall patterns. Discrete seasonal ΔTWS responses of ±1‒5 cm/year are indicated by GRACE for each of the basins, with the exception of the Congo, North Kalahari and Senegal basins which display larger seasonal ΔTWS equivalent to approx. ±11‒20 cm/year. The different seasonal responses in ΔTWS provide useful information about groundwater, including the identification of 5 to 9 month accumulation periods of rainfall in many semi-arid and arid basins as well as differences in ΔTWS responses between Sahelian and southern African aquifers to similar rainfall, likely reflecting differences in landcover. Seasonal ΔGWS estimated by combining GRACE ΔTWS with LSM outputs compare inconsistently to available in situ measurements of groundwater recharge from different basins, highlighting the need to further develop the representation of recharge process in LSMs and the need for more in situ observations from piezometry.

High-quality environmental baseline monitoring data are being collected in areas around two proposed shale gas sites near Kirby Misperton, North Yorkshire and Little Plumpton Lancashire. Monitoring has now been on-going for over two years and has produced an internationally unique data set that will allow any future changes that arise from industrial activities at either or both shale gas sites to be detected and characterised, as well as providing a significant resource for future research. The monitoring includes: water quality, air quality, seismicity, ground motion, soil gas and radon in air. This report presents the results of monitoring in the Vale of Pickering, within which the Kirby Misperton shale gas site (KM8) is located, for the period April 2017–March 2018. It also includes the results of atmospheric composition measurements made near the Little Plumpton (Preston New Road) site. Earlier results and other monitoring in Lancashire are reported elsewhere and can be accessed from the British Geological Survey’s website1. As well as providing valuable insight into the importance of establishing robust information on the conditions before shale gas operations start, it also highlights the challenges in establishing effective monitoring and producing reliable results. For groundwater, this includes the importance of: developing and flushing newly installed boreholes; the spatial variation in water quality and; the selection of monitoring and measuring techniques. Having two years of data has allowed comparison between years. The preliminary analysis reported here has shown that sample populations were not significantly different between the two years. This is directly relevant to the duration of monitoring required by legislation, with the evidence supporting a baseline monitoring period of at least 12 months before any site operations start. The seismic monitoring network installed for measuring background seismicity has operated successfully throughout the r

Even at low concentrations, the presence of arsenic and mercury in soils can lead to ecological and health impacts. The recent European-wide LUCAS Topsoil Survey found that the arsenic concentration of a large proportion of French soils exceeded a threshold which indicated that further investigation was required. A much smaller proportion of soils exceeded the corresponding threshold for mercury but the impacts of mining and industrial activities on mercury concentrations are not well understood. We use samples from the French national soil monitoring network (RMQS: Réseau de Mesures de la Qualité des Sols) to explore the variation of topsoil arsenic and mercury concentrations across mainland France at a finer spatial resolution than was reported by LUCAS Topsoil. We use geostatistical methods to map the expected concentrations of these elements in the topsoil and the probabilities that the legislative thresholds are exceeded. We find that, with the exception of some areas where the geogenic concentrations and soil adsorption capacities are very low, arsenic concentrations are generally larger than the threshold which indicates that further assessment of the area is required. The lower of two other guideline values indicating risks to ecology or health is exceeded in fewer than 5% of RMQS samples. These exceedances occur in localised hot-spots primarily associated with mining and mineralization. The probabilities of mercury concentrations exceeding the further assessment threshold value are everywhere less than 0.01 and none of the RMQS samples exceed either of the ecological and health risk thresholds. However, there are some regions with elevated concentrations which can be related to volcanic material, natural mineralizations and industrial contamination. These regions are more diffuse than the hot-spots of arsenic reflecting the greater volatility of mercury and therefore the greater ease with which it can be transported and redeposited. The maps provide a basel

The Chalk is an important water supply aquifer, yet ecosystems within it remain poorly understood. Boreholes (198) in seven areas of England (UK) were sampled to determine the importance of the Chalk aquifer as a habitat, and to improve understanding of how species are distributed. Stygobitic macro-invertebrates were remarkably common, and were recorded in 67 % of boreholes in unconcealed Chalk, although they were not recorded in Chalk that is concealed by low-permeability strata and thus likely to be confined. Most species were found in shallow boreholes (<21 m) and boreholes with deep (>50 m) water tables, indicating that the habitat is vertically extensive. Stygobites were present in more boreholes in southern England than northern England (77 % compared to 38 %). Only two species were found in northern England compared to six in southern England, but overall seven of the eight stygobitic macro-invertebrate species found in England were detected in the Chalk. Two species are common in southern England, but absent from northern England despite the presence of a continuous habitat prior to the Devensian glaciation. This suggests that either they did not survive glaciations in the north where glaciers were more extensive, or dispersal rates are slow and they have never colonised northern England. Subsurface ecosystems comprising aquatic macro-invertebrates and meiofauna, as well as the microbial organisms they interact with, are likely to be widespread in the Chalk aquifer. They represent an important contribution to biodiversity, and may influence biogeochemical cycles and provide other ecosystem services.

The fertilizer-nitrogen (N) requirement for wheat grown in the UK varies from field to field. Differences in the soil type, climate and cropping history result in differences in (i) the crops’ demands for N, (ii) the supply of N from the soil (SNS) and (iii) the recovery of the fertilizer by the crops. These three components generally form the basis of systems for N recommendation. Three field experiments were set out to investigate the variation of the N requirement for wheat within fields and to explore the importance of variation in the crops’ demands for N, SNS and fertilizer recovery in explaining the differences in the economic optima for N. The N optima were found to vary by >100 kg N/ha at two of the sites. At the other site, the yield response to N was small. Yields at the optimum rate of N varied spatially by c. 4 t/ha at each site. Soil N supply, which was estimated by the unfertilized crops’ harvested N, varied spatially by 120, 75 and 60 kg/ha in the three experiments. Fertilizer recovery varied spatially from 30% to >100% at each of the sites. There were clear relationships between the SNS and the N optima at all the three sites. The expected relationship between the crop's demand for N and N optima was evident at only one of the three sites. There was no consistent relationship between the N recovery and the N optima. A consistent relationship emerged, however, between the optimal yield and SNS; areas with a greater yield potential tending to also supply more N from the soil. This moderated the expected effect of the SNS and the crop's demand for N on the N optima.

Groundwater drought is a spatially and temporally variable phenomenon. Here we describe the development and application of a method to regionalize and quantify groundwater drought based on categorisation of Standardised Groundwater level Index (SGI) time series. The categorisation scheme uses non-hierarchical k-means cluster analysis. This has been applied to 74 SGI time series for the period January 1983 to August 2012 for a case study from Lincolnshire, UK. Six SGI time series clusters have been identified. For each cluster a correlation can be established between the mean SGI and a mean Standardised Precipitation Index (SPI) associated with an optimal SPI accumulation period, qmax. Based on a comparison of SPI time series for each cluster and SPI estimated for the whole study area, it is inferred that the clusters are largely independent of heterogeneity in the diving meteorology across the study region and are primarily a function of catchment and hydrogeological factors. This inference is supported by the observation that the majority of sites in each cluster are associated with one of three principal aquifers in the study region. The groundwater drought characteristics of the three largest clusters (CL1, CL2 and CL4 that constitute ~80% of the sites) have been analyzed. There is a common linear relationship between drought magnitude and duration for each of three clusters. However, there are differences in the character of the groundwater drought events between the three clusters as a function of autocorrelation of the mean SGI time series for each cluster. For example, CL1 has a relatively short period of significant SGI autocorrelation compared with CL2 (15 and 23 months respectively); CL1 has more than twice the number of drought episodes (39 episodes) than CL2 (15 episodes), and the average and maximum duration of droughts in CL1 (4.6 and 27 months) are less than half those of CL2 (11.3 and 61 months). The drought characteristics of CL4 are intermediate be

The English Lowlands is a relatively dry, densely populated region in the south-east of the UK in which water is used intensively. Consequently, parts of the region are water-stressed and face growing water resource pressures. The region is heavily dependent on groundwater and particularly vulnerable to long, multi-annual droughts primarily associated with dry winters. Despite this vulnerability, the atmospheric drivers of multi-annual droughts in the region are poorly understood, an obstacle to developing appropriate drought management strategies, including monitoring and early warning systems. To advance our understanding, we assess known key climate drivers in the winter half-year (October–March) and their likely relationships with multi-annual droughts in the region. We characterise historic multi-annual drought episodes back to 1910 for the English Lowlands using various meteorological and hydrological data sets. Multi-annual droughts are identified using a gridded precipitation series for the entire region, and refined using the Standardized Precipitation Index (SPI), Standardized Streamflow Index (SSI) and Standardized Groundwater level Index (SGI) applied to regional-scale river flow and groundwater time series. We explore linkages between a range of potential climatic driving factors and precipitation, river flow and groundwater level indicators in the English Lowlands for the winter half-year. The drivers or forcings include El Niño–Southern Oscillation (ENSO), the North Atlantic tripole sea surface temperature (SST) pattern, the Quasi-Biennial Oscillation (QBO), solar and volcanic forcing and the Atlantic Multi-decadal Oscillation (AMO). As expected, no single driver convincingly explains the occurrence of any multi-annual drought in the historical record. However, we demonstrate, for the first time, an association between La Niña episodes and winter rainfall deficits in some major multi-annual drought episodes in the English Lowlands. We also show signif

Soils act as a sink or a source of atmospheric carbon, and great efforts are made to monitor soil organic carbon stocks, but soil inorganic carbon (SIC) stocks are not measured by many national- and continental-scale soil monitoring networks. Topsoil (0–30 cm) SIC concentrations were determined for > 2000 sites on a regular 16-km grid as part of the French, Réseau de Mesures de la Qualité des Sols (RMQS). We used design-based statistical methods to calculate unbiased estimates of the mean SIC concentration and total stocks across France. Model-based methods were used to determine the uncertainty of these estimates and to map the spatial distribution of these quantities. Observations of inorganic carbon were highly skewed and did not conform to standard statistical models. Data were normalized using a nonparametric transformation. The estimates and predictions of inorganic carbon are baselines against which the results of future phases of the network can be compared. We found that the total topsoil inorganic carbon stocks in France amount to 1070 ± 61 Tg, ca. one-third of the corresponding organic carbon stocks. Spatial distribution of SIC was strongly linked to the underlying geology. We tested the reliability of estimating SIC concentrations and stocks from the French Soil Test Database, which contains the results of 280 000 soil analyses requested by farmers between 1990 and 2004. A biased estimate of soil inorganic carbon concentrations resulted, presumably because soil samples were selected according to concerns of farmers rather than by a statistical design.

Groundwater drought is a spatially and temporally variable phenomenon. Here we describe the development of a method to regionally analyse and quantify groundwater drought. The method uses a cluster analysis technique (non-hierarchical k-means) to classify standardised groundwater level hydrographs (the standardised groundwater level index, SGI) prior to analysis of their groundwater drought characteristics, and has been tested using 74 groundwater level time series from Lincolnshire, UK. Using the test data set, six clusters of hydrographs have been identified. For each cluster a correlation can be established between the mean SGI and a mean standardised precipitation index (SPI), where each cluster is associated with a different SPI accumulation period. Based on a comparison of SPI time series for each cluster and for the study area as a whole, it is inferred that the clusters are independent of the driving meteorology and are primarily a function of catchment and hydrogeological factors. This inference is supported by the observation that the majority of sites in each cluster are associated with one of the principal aquifers in the study region. The groundwater drought characteristics of the three largest clusters, which constitute ~ 80 % of the sites, have been analysed. There are differences in the distributions of drought duration, magnitude and intensity of groundwater drought events between the three clusters as a function of autocorrelation of the mean SGI time series for each cluster. In addition, there are differences between the clusters in their response to three major multi-annual droughts that occurred during the analysis period. For example, sites in the cluster with the longest SGI autocorrelation experience the greatest-magnitude droughts and are the slowest to recover from major droughts, with groundwater drought conditions typically persisting at least 6 months longer than at sites in the other clusters. Membership of the clusters is shown to be r

Open defecation is practised by over 600 million people in India and there is a strong political drive to eliminate this through the provision of on-site sanitation in rural areas. However, there are concerns that the subsequent leaching of excreta from subsurface storage could be adversely impacting underlying groundwater resources upon which rural populations are almost completely dependent for domestic water supply. We investigated this link in four villages undergoing sanitary interventions in Bihar State, India. A total of 150 supplies were sampled for termotolerant (faecal) coliforms (TTC) and tryptophanlike fluorescence (TLF): an emerging real-time indicator of faecal contamination. Sanitary risk inspections were also performed at all sites, including whether a supply was located within 10 m of a toilet, the recommended minimum separation. Overall, 18% of water supplies contained TTCs, 91% of which were located within 10 m of a toilet, 58% had TLF above detection limit, and sanitary risk scores were high. Statistical analysis demonstrated TLF was an effective indicator of TTC presence-absence, with a possibility of TTCs only where TLF exceeded 0.4 mg/L dissolved tryptophan. Analysis also indicated proximity to a toilet was the only significant sanitary risk factor predicting TTC presence-absence and the most significant predictor of TLF. Faecal contamination was considered a result of individual water supply vulnerability rather than indicative of widespread leaching into the aquifer. Therefore, increasing faecal contamination of groundwater-derived potable supplies is inevitable across the country as uptake of onsite sanitation intensifies. Communities need to be aware of this link and implement suitable decentralised low-cost treatment of water prior to consumption and improve the construction and protection of new supplies.

Purpose-built soil monitoring networks have been established in many countries to identify where soil functionality is threatened and to target remediation initiatives. An alternative to purpose-built soil monitoring networks is to use legacy soils information. Such information yields almost instant assessments of soil change but the results should be interpreted with caution since the information was not collected with monitoring in mind. We assess the threat of soil acidification in Victoria using two legacy datasets: (i) the Victorian Soils Information System (VSIS) which is a repository of the results of soil analyses conducted for scientific purposes since the 1950s and (ii) a database of 75 000 routine soil test results requested by farmers between 1973 and 1993. We find that the VSIS measurements are clustered in space and time and are therefore suitable for local rather than broad-scale assessments of soil change. The farmers' results have better spatial and temporal coverage and space-time models can be used to quantify the spatial and temporal trends in the pH measurements. However, careful validation of these findings is required since we do not completely understand how the measured paddocks were selected and we cannot be certain that sampling or laboratory protocols have not changed with time.

We briefly describe three methods of seabed characterization which are ‘fit for purpose’, in that each approach is well suited to distinct objectives e.g. characterizing glacial geomorphology and shallow glacial geology vs. rapid prediction of seabed sediment distribution via geostatistics. The methods vary from manual ‘expert’ interpretation to increasingly automated and mathematically based models, each with their own attributes and limitations. We would note however that increasing automation and mathematical sophistication does not necessarily equate to improve map outputs, or reduce the time required to produce them. Judgements must be made to select methodologies which are most appropriate to the variables mapped, and according to the extent and presentation scale of final maps.

A new index for standardising groundwater level time series and characterising groundwater droughts, the Standardised Groundwater level Index (SGI), is described. The SGI builds on the Standardised Precipitation Index (SPI) to account for differences in the form and characteristics of groundwater level and precipitation time series. The SGI is estimated using a non-parametric normal scores transform of groundwater level data for each calendar month. These monthly estimates are then merged to form a continuous index. The SGI has been calculated for 14 relatively long, up to 103 yr, groundwater level hydrographs from a variety of aquifers and compared with SPI for the same sites. The relationship between SGI and SPI is site specific and the SPI accumulation period which leads to the strongest correlation between SGI and SPI, qmax, varies between sites. However, there is a consistent positive linear correlation between a measure of the range of significant autocorrelation in the SGI series, mmax, and qmax across all sites. Given this correlation between SGI mmax and SPI qmax, and given that periods of low values of SGI can be shown to coincide with previously independently documented droughts, SGI is taken to be a robust and meaningful index of groundwater drought. The maximum length of groundwater droughts defined by SGI is an increasing function of mmax, meaning that relatively long groundwater droughts are generally more prevalent at sites where SGI has a relatively long autocorrelation range. Based on correlations between mmax, average unsaturated zone thickness and aquifer hydraulic diffusivity, the source of autocorrelation in SGI is inferred to be dependent on dominant aquifer flow and storage characteristics. For fractured aquifers, such as the Cretaceous Chalk, autocorrelation in SGI is inferred to be primarily related to autocorrelation in the recharge time series, while in granular aquifers, such as the Permo–Triassic sandstones, autocorrelation in SGI is in

A new index for standardising groundwater level time series and characterising groundwater droughts, the Standardised Groundwater level Index (SGI), is described. The SGI is a modification of the Standardised Precipitation Index (SPI) that accounts for differences in the form and characteristics of precipitation and groundwater level time series. The SGI is estimated using a non-parametric normal scores transform of groundwater level data for each calendar month. These monthly estimates are then merged to form a continuous index. The SGI has been calculated for 14 relatively long, up to 103 yr, groundwater level hydrographs from a variety of aquifers and compared with SPI for the same sites. The SPI accumulation period which leads to the strongest correlation between SPI and SGI, qmax, varies between sites. There is a positive linear correlation between qmax and a measure of the range of significant autocorrelation in the SGI series, mmax. For each site the strongest correlation between SPI and SGI is in the range 0.7 to 0.87, and periods of low values of SGI coincide with previously independently documented droughts. Hence SGI is taken to be a robust and meaningful index of groundwater drought. The maximum length of groundwater droughts defined by SGI is an increasing function of mmax, meaning that relatively long groundwater droughts are generally more prevalent at sites where SGI has a relatively long autocorrelation range. Based on correlations between mmax, average unsaturated zone thickness and aquifer hydraulic diffusivity, the source of autocorrelation in SGI is inferred to be dependent on aquifer flow and storage characteristics. For fractured aquifers, such as the Cretaceous Chalk, autocorrelation in SGI is inferred to be primarily related to autocorrelation in the recharge time series, while in granular aquifers, such as the Permo-Triassic Sandstones, autocorrelation in SGI is inferred to be primarily a function of intrinsic aquifer characteristics. These

The urban area of Swansea was a centre of metal processing and production from the 16th until the 20th century and was subject to a substantial amount of diffuse atmospheric and point-source metal pollution. We map the concentrations of four heavy metals across the urban region based upon a soil geochemical survey undertaken by the British Geological Survey in the summer of 1994 covering an area of 93 km2 at a density of 4 samples per km2. The non-aligned sampling grid yielded a total of 373 samples. Analyses of these samples by X-ray fluorescence spectromety revealed that throughout the urban area the concentrations of As, Cu, Pb and Sn were greater than the ambient background concentrations. Several hotspots of each of these heavy metals were observed at former industrial sites. These hotspots hinder the mapping of these metal concentrations by conventional geostatistical methods in two ways. Firstly when models of spatial variation are fitted to the data the hotspots lead to the variance of the metal concentrations being over-estimated. This can in turn lead to details within our maps being smoothed out. Secondly conventional interpolation techniques such as kriging exaggerate the size of the hotspots. Therefore the components of metal concentrations due to point-source and diffuse pollution were mapped separately by robust geostatistical methods. We assume that the dominant pattern of spatial variation of metal concentrations across the urban region results from diffuse pollution whereas the hotspots are statistical outliers. Robust models of spatial variation were used to represent the diffuse component of pollution. The parameters of these models are not unduly influenced by the outliers. The diffuse pollution was mapped by robust median kriging. This method is able to identify and truncate the outliers. Statistical validation confirmed the appropriateness of the model of the spatial variation of metal concentrations due to diffuse pollution. Historical record

Soil scientists require cost-effective methods to make accurate regional predictions of soil organic carbon (SOC) content. We assess the suitability of airborne radiometric data and digital elevation data as covariates to improve the precision of predictions of SOC from an intensive survey in Northern Ireland. Radiometric data (K band) and, to a lesser extent, altitude are shown to increase the precision of SOC predictions when they are included in linear mixed models of SOC variation. However the statistical distribution of SOC in Northern Ireland is bimodal and therefore unsuitable for geostatistical analysis unless the two peaks can be accounted for by the fixed effects in the linear mixed models. The upper peak in the distribution is due to areas of peat soils. This problem may be partly countered if soil maps are used to classify areas of Northern Ireland according to their expected SOC content and then different models are fitted to each of these classes. Here we divide the soil in Northern Ireland into three classes, namely mineral, organo-mineral and peat. This leads to a further increase in the precision of SOC predictions and the median square error is 2.2 %2. However a substantial number of our observations appear to be mis-classified and therefore the mean squared error in the predictions is larger (30.6 %2) since it is dominated by large errors due to mis-classification. Further improvement in SOC prediction may therefore be possible if better delineation between areas of large SOC (peat) and small SOC (non-peat) could be achieved.