We investigate methods that improve the estimation of indoor gamma ray dose rates at locations where measurements had not been made. These new predictions use a greater range of modelling techniques and larger variety of explanatory variables than our previous examinations of this subject. Specifically, we now employ three types of machine learning models in addition to the geostatistical, nearest neighbour and other earlier models. A large number of parameters, mostly describing the characteristics of dwellings in the area in question, have been added to the set of explanatory variables. The use of machine learning methods results in significantly improved predictions over earlier models. The machine learning models are noisy and there is some instability in the relative importance of particular explanatory variables although there are general and consistent tendencies supporting the importance of certain classes of variable. However, the range of predicted indoor gamma ray dose rates is much smaller than that of the measurements. It is probable that epidemiological studies using such predictions will have lower statistical power than those based on direct measurements.

Gamma-rays from naturally occurring radionuclides are a major component of background radiation. They are an important tool for geology and are also important for radiation protection. In this paper we use over a quarter of a million geochemical measurements of concentrations of potassium, thorium and uranium in soils and in stream sediments to estimate outdoor gamma-ray dose rates across Great Britain. The soil concentrations are generally at a depth of 5-20 cm with some at 35-50 cm. Soil measurements will give spatially relatively precise estimates, but as soil data are not available for much of Scotland, stream sediment data are used there. Kriging methods are used to estimate surface concentrations of K, Th and U and dose rates are imputed from these concentrations. Our results are compared with measurement surveys of both outdoor and indoor gamma-ray dose rates. Recently there has been interest in exploring the carcinogenic risks of low dose radiation by investigating associations between childhood cancer rates and doses from natural background gamma radiation. To achieve adequate statistical power, such studies must be so large that it is impractical to assess exposures by direct measurements in the homes of study subjects. Instead the exposures must be modelled. The results presented here will be an important input to such work.

Abstract Radon ( 222 Rn) is a natural radioactive gas that occurs in rocks and soils and can only be detected with special equipment. Radon is a major cause of lung cancer. Therefore, early detection is essential. The British Geological Survey and Public Health England have produced a series of maps showing radon affected areas based on underlying geology and indoor radon measurements, which help to identify radon-affected buildings. Many factors influence how much radon accumulates in buildings. Remedial work can be undertaken to reduce its passage into homes and workplaces and new buildings can be built with radon preventative measures.

Gamma radiation from naturally occurring sources (including directly ionizing cosmic-rays) is a major component of background radiation. An understanding of the magnitude and variation of doses from these sources is important, and the ability to predict them is required for epidemiological studies. In the present paper, indoor measurements of naturally occurring gamma-rays at representative locations in Great Britain are summarized. It is shown that, although the individual measurement data appear unimodal, the distribution of gamma-ray dose-rates when averaged over relatively small areas, which probably better represents the underlying distribution with inter-house variation reduced, appears bimodal. The dose-rate distributions predicted by three empirical and geostatistical models are also bimodal and compatible with the distributions of the areally averaged dose-rates. The distribution of indoor gamma-ray dose-rates in the UK is compared with those in other countries, which also tend to appear bimodal (or possibly multimodal). The variation of indoor gamma-ray dose-rates with geology, socio-economic status of the area, building type, and period of construction are explored. The factors affecting indoor dose-rates from background gamma radiation are complex and frequently intertwined, but geology, period of construction, and socio-economic status are influential; the first is potentially most influential, perhaps, because it can be used as a general proxy for local building materials. Various statistical models are tested for predicting indoor gamma-ray dose-rates at unmeasured locations. Significant improvements over previous modelling are reported. The dose-rate estimates generated by these models reflect the imputed underlying distribution of dose-rates and provide acceptable predictions at geographical locations without measurements.

Systematic mapping of the chemical environment of urban areas from around the world has demonstrated the strong impact of urbanisation on topsoil geochemical distributions originally controlled by the underlying parent material (PM). The variance of some elements including As, Ba, Ca, Cr, Cu, Mo, P, Pb, Sb, Se, Sn and Zn in urban domains appears to be impacted by a mixture of geogenic and anthropogenic controls. This study evaluates how soil chemistry has been influenced by different eras of urbanisation within London and other UK urban areas using (a) the pre-1940 Dudley Stamp First Land Utilisation Survey data and (b) the modern urban domain principally defined by the aggregate classes of the 2007 Land Cover Map. In the London area, calcium, and possibly a substantial proportion of Cu, Pb, Sn and Zn enrichment observed in soils impacted by pre-1940 urbanisation relative to soils impacted only by post-1940 urbanisation, may be partly related to the destruction of buildings during the period 1940–1941 rather than from the disposal or aerial dispersion of coal ash from domestic fires. Some Pb, Cu, Sb, Sb, Sn and Zn contamination appears to be caused by road traffic (leaded petrol and brake dust). The relationships between pre- and post-1940 urbanised areas in London also characterise most of 20 other urban centres in England and Wales for which BGS holds soil chemistry data.

The London Region Atlas of Topsoil Geochemistry (LRA) is a further step towards understanding the chemical quality of soils in London, following a previous project called London Earth carried out by the British Geological Survey (BGS) (Johnson et al., 2010[1]). The main advantage of the LRA is that it includes soil geochemical data from the counties surrounding London; placing the city within the context of its rural hinterland, allowing assessments of the impact of urbanisation on soil quality. The London Region Atlas of Topsoil Geochemistry is a product derived from the BGS Geochemical Baseline Survey of the Environment (G-BASE[2]) project. The London Region Geochemical Dataset (LRD, n=8400), on which the atlas is based, includes TOPSOIL data from two complementary surveys: i) the urban London Earth (LOND) and ii) the rural South East England (SEEN). The LRA covers the Greater London Authority (GLA) and its outskirts in a rectangular area of 80x62 km. This extends from British National Grid coordinates Easting 490000–570000, and Northing 153000–215000. The urban LOND and the rural SEEN surveys contribute with 6801 and 1599 samples respectively to the LRD. The concentrations of 44 inorganic chemical elements (Al2O3, CaO, Fe2O3, K2O, MgO, MnO, Na2O, P2O5, SiO2, TiO2, Ag, As, Ba, Bi, Br, Cd, Ce, Co, Cr, Cs, Cu, Ga, Ge, Hf, I, La, Mo, Nb, Nd, Ni, Pb, Rb, Sb, Sc, Se, Sn, Sr, Th, U, V, W, Y, Zn and Zr), loss on ignition (LOI) and pH in topsoil are included in the LRA. For each element, a map showing the distribution in topsoil across the atlas area and a one-page sketch of descriptive statistics and graphs are presented. Statistics and graphs for whole dataset (LRD), London urban subset (LOND) and London surroundings rural subset (SEEN), as well as graphs of topsoil element concentrations over each simplified geology unit are shown. The LRD has been used already in a study aiming to detect geogenic (geological) signatures and controls on soil chemistry in the London region (Appleton et al., 2013[3]). It includes maps showing the distribution of Al, Si, La and I (and Th, Ca, Mn, As, Pb and Zr in supplementary material) and it is concluded that the spatial distribution of a range of elements is primarily controlled by the rocks from where soil derives, and that these geogenic patterns are still recognisable inside the urban centre. Other studies have been done that are based on data in the LRD, namely using the LOND subset or part of it. The main focus of these studies was the mercury content (Scheib et al., 2010[4]), the influence of land use on geochemistry (Knights and Scheib, 2011[5]; Lark and Scheib, 2013[6]); the bioaccessibility of pollutants such as As and Pb (Appleton et al., 2012[7]; Appleton et al., 2012[8]; Cave, 2012[9]; Appleton et al., 2013[10]; Cave et al., 2013[11]) and the lability of lead in soils (Mao et al., 2014[12]); the determination of normal background concentrations of contaminants in English soil (Ander et al., 2013[13]) and the contribution of geochemical and other environmental data to the future of the cities (Ludden et al., 2015[14]). The London Region Atlas of Topsoil Geochemistry formally presents detailed information for all chemical elements in the LRD. This information can be easily visualised and elements compared as its production and layout is standardised. Differences in topsoil element concentrations between the centre of the city and its outskirts can be assessed by observing the map and comparing statistics and graphs reported for the LOND and SEEN subsets respectively. This urban/rural contrast is particularly evident for elements such as Pb, Sb, Sn, Cu and Zn, for which mean concentrations in the urban environment are two to three times higher than those observed in the rural environment. This is a typical indicator suite of urban soil pollution reported in several other cities in the UK also (Fordyce et al., 2005[15]).

Moderate and high radon potential in Northern Ireland is associated mainly with (i) the Neoproterozoic psammites, semipelites, meta-limestones, volcanics and mafic intrusives of Counties Londonderry and Tyrone; (ii) Silurian Hawick Group greywackes and, to a much more limited extent Gala Group greywackes, in the southern sector of Counties Armagh and Down; (iii) Ordovician and Silurian acid intrusives and volcanics in eastern Counties Londonderry and Tyrone; (iv) Middle-Late Devonian conglomerates in County Tyrone; (v) Lower Carboniferous (Dinantian) limestone in the western sector of Northern Ireland, especially in County Fermanagh; (vi) Palaeogene and Late Caledonian acid intrusive rocks of the Mourne Mountains Complex, Slieve Gullion Complex and Newry Granodiorite Complex in the SE sector in County Down and County Armagh.Moderate to high radon potential is sometimes associated with glacio-fluvial sand and gravel deposits where these overlie a range of bedrocks, some of which have relatively low radon potential. In this latter case the enhanced radon potential is probably caused by the high permeability of superficial deposits. Radon potential tends to be lower when bedrocks characterised by moderate or high radon potential are overlain by relatively impermeable silt-clay alluvium, glaciolacustrine, and lacustrine deposits; peat; and glacial till and moraine. Redistribution of rock debris derived from uranium-rich bedrocks, such as the Mourne Mountains granites, through glacial, alluvial and other processes can also result in higher radon potential being associated with superficial deposits relative to underlying bedrocks. (C) 2015 Natural Environment Research Council. Published by Elsevier Ltd on behalf of The Geologists' Association. All rights reserved.

A joint mapping method based on the results of radon measurements and geological information was used to estimate radon ‘Affected Area’ potential in Northern Ireland. The method allows variation of radon potential both between and within geological units and is based on measurements of radon in more than 23,000 homes. This report presents information about elevated radon potential in Northern Ireland as an indicative map, based on the highest radon potential for each 1 km square of the Irish grid. The full definitive detail is published as a digital dataset for geographical information systems, which can be licensed. The estimated radon potential for an individual home can be obtained through the Public Health England radon website, www.ukradon.org. The work was partially funded by the Northern Ireland Department of the Environment and was prepared jointly by Public Health England and the British Geological Survey.

Predictive linear regression (LR) modelling indicates that total Pb is the only highly significant independent variable for estimating Pb bioaccessibility in "mineralisation domains" located in limestone (high pH) and partly peat covered (low pH) shale-sandstone terrains in England. Manganese is a significant minor predictor in the limestone terrain, whilst organic matter and sulphur explain 0.5% and 2% of the variance of bioaccessible Pb in the peat-shale-sandstone terrain, compared with 93% explained by total Pb. Bootstrap resampling shows that LR confidence limits overlap for the two mineralised terrains but the limestone terrain has a significantly lower bioaccessible Pb to total Pb slope than the urban domain. A comparison of the absolute values of stomach and combined stomach-intestine bioaccessibility provides some insight into the geochemical controls on bioaccessibility in the contrasting soil types.

Synopsis 222 Rn, a natural radioactive gas produced by the radioactive decay of 238 U, accounts for about 50% of the total radiation dose to the average person in the UK. Geology is the most important factor controlling the source and distribution of radon; which has been linked to an increased risk of lung cancer. In order to prevent the public receiving high exposures to radon, it is necessary to identify those areas most at risk. We present results of new mapping of radon potential for Scotland using a method that allows the spatial variation in radon potential to be delineated both within and between geological groupings. The main geological and geochemical associations with moderate to high radon potential areas are described. The highest radon potential values in Scotland are associated with U-rich, highly evolved Siluro-Devonian biotite granite intrusions, notably those clustered within a zone to the west of Aberdeen and at Helmsdale, in Caithness. U mineralization plays a role in areas including the Helmsdale granite and the Middle Old Red Sandstone of the Orcadian Basin. Elevated radon potential is also associated with limestones – where fracture permeability is influential – and with Ordovician–Silurian greywackes. The radon potential of unconsolidated deposits, and how this affects the radon potential of the underlying bedrock, reflects both their permeabilities and their compositions.

Systematic mapping of the chemical environment of urban areas from around the world have shown varying degrees of control of element distributions by the underlying parent material (PM). The purpose of the study reported here is to assess whether geogenic signatures that dominate soil chemistry in rural domains of Eastern England and which are not strongly impacted by human activities, can also be detected in the London urban region. A PM soil chemistry mapping method is used to determine the spatial variation of topsoil chemistry data in London and the surrounding rural areas. Analysis of variance (ANOVA) of the soil data for the London region indicates that 26-33% of the variance of Al, Ce, Cs, Ga, K, La, Mg, Mn, Nb, Nd, Rb, Ti, V and Y is explained by soil PM (surface geology), and a slightly lesser proportion (19-25%) of the variance for Ca, Co, Fe, I, Ni, Sc, Sr and Th. In comparison, soil PM explains only 5% of the variance of Cd. The variance of some other elements appears to be influenced by a mixture of geogenic and anthropogenic controls, including As, Ba, Cr, Cu, Mo, P, Pb, Sb, Se, Sn and Zn for which PM controls 12-16% of the variance. Geogenic soil chemistry patterns observed for the elements strongly influenced by PM in the rural areas surrounding London can be quite clearly followed into and through the London urban area. Spatial patterns of a range of elements primarily controlled by PM have not been destroyed even in a major urban centre with a recorded history dating back over 2000 years and which has been subjected to extensive urban development, destruction and redevelopment especially during the last 200-300 years. (C) 2013 Natural Environment Research Council. Published by Elsevier Ltd. All rights reserved.

Radon is a natural radioactive gas that you cannot see, smell, or taste and that can only be detected with special equipment. It is produced by the radioactive decay of radium, which in turn is derived from the radioactive decay of uranium. Uranium is found in small quantities in all soils and rocks, although the amount varies from place to place. Radon decays to form radioactive particles that can enter the body by inhalation. Inhalation of the short-lived decay products of radon has been linked to an increase in the risk of developing cancers of the respiratory tract, especially of the lungs. Breathing radon in the indoor air of homes contributes to about 15,000 lung cancer deaths each year in the United States and 1,100 in the UK (HPA 2009). Only smoking causes more lung cancer deaths.

Predictive linear regression (LR) modelling between bioaccessible Pb and a range of total elemental compositions and soil properties was executed for the Glasgow, London, Northampton and Swansea urban areas in order to assess the potential for developing a national urban bioaccessible Pb dataset for the UK. LR indicates that total Pb is the only highly significant independent variable for estimating the bioaccessibility of Pb. Bootstrap resampling shows that the relationship between total Pb and bioaccessible Pb is broadly the same in the four urban areas. The median bioaccessible fraction ranges from 38% in Northampton to 68% in London and Swansea. Results of this study can be used as part of a lines of evidence approach to localised risk assessment but should not be used to replace bioaccessibility testing at individual sites where local conditions may vary considerably from the broad overview presented in this study.

This report presents a description and review of the methodologies developed by the British Geological Survey (BGS) to produce a national scale assessment of the concentrations selected potentially harmful elements (arsenic, cadmium, chromium, nickel and lead) in topsoils. The methodologies are described briefly in this report and in four scientific papers.

Predictive linear regression (LR) modelling between bioaccessible arsenic (B-As) and a range of total elemental compositions and soil properties was executed in order to assess the potential for developing a national B-As dataset for the UK. LR indicates that total arsenic (As) is the only highly significant independent variable for estimating B-As in urban areas where it explains 75-92% of the variance. The broad compatibility of the London, Glasgow and Swansea regression models suggests that application of these models to estimate bioaccessible As in UK soils impacted by diffuse anthropogenic urban contamination and non-ferrous metal processing should be relatively accurate. In areas dominated by Jurassic ironstones and associated clays and limestones, total As, P and pH are significant, accounting for 53, 14 and 5%, respectively, of the B-As variance. Models based on total As as the sole predictor in the combined Jurassic and Cretaceous sedimentary ironstones datasets explain about 40% of the B-As variance. The median As bioaccessible fraction (%As-BAF) is 19 to 28% in the anthropogenic contamination impacted urban domains, but much lower (5-9%) in geogenic terrains dominated by ironstones. Results of this study can be used as part of a lines of evidence approach to localised risk assessment but should not be used to replace bioaccessibility testing at individual sites where local conditions may vary considerably from the broad overview presented in this study.

The purpose of the study reported here is to assess whether it may in some circumstances be useful and appropriate to use a parent material (PM) soil chemistry mapping method developed for national soil chemistry data to portray spatial variation in urban soil chemistry data in Great Britain. Analysis of variance (ANOVA) of the urban soil data suggests that spatial interpolation of soil ambient background concentrations (ABCs) using PM classified soil data may be justified for those elements with strong geogenic control. The PM soil chemistry mapping method for urban soil data is demonstrated using data from the Northampton urban area, in the English Midlands. Geometric mean (GM) and inverse distance weighting (IDW) interpolations based on the nearest four topsoil samples were evaluated. Independent validation indicated that for As, Cr, Fe and to a lesser extent K, which all exhibit relatively strong geogenic control in the Northampton urban area: (i) the PM soil chemistry mapping method is more accurate and effective than the conventional IDW grid mapping and (ii) PM soil chemistry mapping based on the average of the nearest four Ln(e) element concentrations is more accurate than mapping based on IDW values calculated from the nearest four Ln(e) element concentrations. The variation in effectiveness of the methods can be explained by the fact that PM exerts a significant control on As, Cr, Fe and K in the Northampton area whereas anthropogenic inputs appear to be the dominant control on the spatial variation of Pb, especially at high concentrations. The PM mapping method would be expected to work efficiently in other urban areas and for those elements where a significant proportion of the variation can be explained by PM. (C) 2011 Natural Environment Research Council. Published by Elsevier Ltd. All rights reserved.

Erratum to “Comparison of Northern Ireland radon maps based on indoor radon measurements and geology with maps derived by predictive modelling of airborne radiometric and ground permeability data” [Sci. Total Environ. 409 (2011) 1572-1583] J.D. Appleton⁎, J.C.H. Miles, M. Young a British Geological Survey, Keyworth, Nottingham, NG12 5GG, UK b Health Protection Agency (HPA), Radiation Protection Division, Chilton, Didcot, Oxon OX11 ORQ, UK c Geological Survey of Northern Ireland, Department of Enterprise, Trade and Investment, Colby House, Stranmillis Court, Stranmillis Road, Belfast BT9 5BJ, Northern Ireland, UK

This report presents an overview of the results of detailed mapping in Scotland of radon potential, defined as the estimated percentage of homes in an area at or above the radon Action Level. The work was carried out jointly by the Health Protection Agency and the British Geological Survey and was based on the results of measurements of radon in over 19,000 homes. The method allows variations in radon potential both between and within geological units to be mapped. The resulting digital map, which defines radon Affected Areas in Scotland, includes much more detail than could be shown in an atlas. The full detail is published as a dataset for Geographical Information Systems, which can be licensed. The estimated radon potential for an individual home can be obtained through a website, www.UKradon.org. The atlas presented here is a simplified version of the map, so is indicative rather than definitive: that is, each 1-km grid square is coloured according to the highest radon potential found within it. © Health Protection Agency and British Geological Survey