Radiometric analysis of Quaternary deposits from the southeastern Brazilian coast

Marine Geology 229 (2006) 29 – 43 www.elsevier.com/locate/margeo

Radiometric analysis of Quaternary deposits from the southeastern Brazilian coast R.M. Anjos ⁎, R. Veiga, K. Macario, C. Carvalho, N. Sanches, J. Bastos, P.R.S. Gomes Instituto de Física da Universidade Federal Fluminense, Av. Gal. Milton Tavares de Souza, s/n _o, Gragoatá, 24210-340, Niterói, RJ, Brazil Received 19 May 2005; received in revised form 28 January 2006; accepted 3 March 2006

Abstract Natural gamma radiation measurements of the beach sand deposits were performed with the aim of understanding the provenance and transport processes of sediments along the coastal zone of three Brazilian States: São Paulo (SP), Rio de Janeiro (RJ) and Espírito Santo (ES). The method employs thorium, uranium and potassium as tracers of the mineralogical properties of beach sand minerals, which reflect the geological history of transport and sorting processes. A cross plot of eTh/eU and eTh/K shows a considerable positive correlation with the geological evolution of the Quaternary coastal deposits from 35 areas of the southeastern Brazilian coast. Additionally, the distribution of heavy minerals along this coastal zone was mapped. © 2006 Elsevier B.V. All rights reserved. Keywords: Quaternary deposits; γ-ray spectrometry; heavy minerals; sediment transport

1. Introduction The Quaternary period is characterized by sea level fluctuations, which are related to the formation of sedimentary strata. Systematical geological mapping and radiocarbon dating of Quaternary coastal deposits have allowed distinguishing several generations of terraces built after the maximum levels associated with various transgressive episodes of the Quaternary. Those results are used in geologic–geomorphologic evolutionary models, which contribute to the understanding of the structure and functioning of the Brazilian coastal systems (Martin et al., 1987, 1997; Andrade et al., 2003). The history of the geological evolution of the ⁎ Corresponding author. Tel.: +55 21 26295770; fax: +55 21 26295887. E-mail address: [email protected] (R.M. Anjos). 0025-3227/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.margeo.2006.03.001

Quaternary deposits is also directly connected with mineral occurrences of economical interest. Sandy deposits, which represent a large portion of the Brazilian coastal plains, can be used for glass production or as building materials. Usually, heavy minerals are found scattered on marine deposits, but in some favorable circumstances, it can be found in high concentrations, suitable for exploitation. Sea level changes can result in the re-working of alluvial and colluvial deposits, by current and wave actions, and so this mechanism can promote the re-concentration of dense minerals such as ilmenite, zircon, tin, gold and diamonds. In Brazil, there are sparse and discontinuous exploitation of deposits of ilmenite, zircon, rutile and monazite. Heavy mineral deposits are explored only in beaches and sandy laces in the coasts of North Rio de Janeiro, Espírito Santo and South Bahia States (Silva, 2000). With the imminent exhaustion of mineral

30

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

resources in land and coastal zones, it would be strategic the appraisal of marine resources, both in coastal waters and in the continental shelf, in order to fill the demand of these resources. However, this would require understanding the geological mechanisms for the origin and transport of sediments in the Brazilian coastal zone. According to Kogan et al. (1969), the formation of sedimentary strata is associated with the breakdown of source rocks, followed by the transport of the terrigenous matter, its re-deposition and sedimentation. Therefore, beach sands are weathering-resistant remainders of geological formations, transported by wind, rivers, and glaciers to the coast, and deposited on the beaches by actions of waves and currents. Consequently, mineralogical properties of beach sand reflect the geological history of the original rock formations. In the parent materials for siliciclastic sediments, most of uranium and thorium atoms are bound in accessory and dark-colored minerals (red, purple or black), typically as small highly brittle particles. When source rocks disintegrate through weathering and erosion, they are released to accumulate as dark-colored population in sands and other fine-grained deposits. These dark-colored minerals possess higher (by 20–100%) specific weight than the rockforming minerals (such as quartz and feldspars) and are known as heavy minerals. Consequently, the transport of these heavy minerals is affected by gravitational separation (de Meijer, 1998; Asadov et al., 2001). Common heavy mineral suites in beaches are composed by zircon, ilmenite, magnetite, garnet, monazite, rutile, and other accessory minerals, some of these enriched in uranium and thorium. Potassium, on the other hand, is abundant in the potash-rich rock-forming minerals, mainly in the orthoclase, also known as the Kfeldspar, as well as in accessory heavy minerals. Therefore it is not expected any significant enrichment or depletion of potassium during the transport process. γ-ray spectrometers have been used for a wide range of geological applications including the lithological mapping of rocks and unconsolidated sediments, mineral exploration (mainly for heavy minerals and phosphorite), sediment transport studies and investigation on discharged and dumped nuclear wastes at nuclear weapon test sites (Jones, 2001). The basis of the radiometric techniques is the measurement of γ-ray emissions from the primordial radionuclides. For practical proposes, the natural gamma emitter series comprise 40K and decay products of the 238U and 232 Th series, most notably 214Bi and 208Tl, respectively. Following Kogan et al. (1969), potassium γ-ray activity of sands having the same origin (parent rock) should be equal, regardless of its mineralogical compo-

sition. For this reason, potassium provides a kind of radioactive label of the source, and the distribution of the primordial radionuclides in beach sands can represent an important tool in geological studies on the provenance and transport of sediments. Statistical analysis of potassium, uranium and thorium contents performed by Kogan et al. (1969) for various soil types showed that variations in concentration of these elements in soils of each genetic type, originated from several places, follow an approximate normal distribution. However, this result indicates that the content of each of these elements in soils of a particular genetic type is determined by several not very correlated parameters, such as variations of the radioactive elements in the initial soil-forming rocks, the humidity and hydrological parameters, differences in landscape and climatic conditions, in the mechanical composition of soils etc. Some works (de Meijer, 1998; Asadov et al., 2001; de Meijer et al., 2001; van Wijngaarden et al., 2002) have stated that the sediment transport is a complex problem and therefore, in order to be appropriately described by γray spectrometry, it would be essential to use additional parameters such as grain size distributions. As in beaches heavy minerals are often concentrated in rather localized spots, usually in the swash zone or at eroding cliffs heavy mineral concentration (fraction of heavy minerals per kg of beach sand) may vary rapidly both in time and in location. Therefore, the technique of measuring activity concentrations and determining lines of equal activity would not be suitable for determining the origin and transport of sediments. Similarly, measuring the total activity on the beach surface is not a valid method, due to the rapidly varying ratio between light and heavy minerals (de Meijer et al., 2001). The γ-ray technique can, however, be broadly adaptable and can envisage several practical applications, such as applying the correlations between Th, U, K, Th/U and Th/K. This paper presents a method in which natural radionuclides correlations are used as tracers of the mineralogical proprieties of beach sand minerals with the aim of associating the sandy sediment composition of the coastal zones with its geological provenance. This technique was employed in the analysis of beach sand samples from the southeastern Brazilian coast and can be considered easier to measure than the classical heavy mineral analysis method. 2. Geological setting The region studied is the coastline between the proximities of the division of Bahia and Espírito Santo States (18°20′S and 39°40′W), and the Southern coast

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

of São Paulo (24°10′S and 46°30′W). Fig. 1 shows a schematic geological map illustrating the main geological features of the southeastern Brazilian coast (CPRM, 2005; DRM-RJ, 2005). The Quaternary sedimentary deposits that compose the beach sands are originated from crystalline basement rocks of several lithologies, mostly of Pre-Cambrian age that forms the landward limit of marginal sedimentary basins (from south to north, these basins are Santos, Campos and Espírito Santo). The crystalline basement rocks are mostly gneisses, which are metamorphic rocks of two dominant origins: para-gneisses and ortho-gneisses. Para-gneisses are derived from the metamorphism of sedimentary rocks, giving out aluminum rich minerals such as garnets, kyenite, sillimanite, staurolite and K-feldspars. Ortho-gneisses are derived from the metamorphism of igneous rocks, generating minerals such as K-feldspars, biotite and anfiboles. Paleozoic homogeneous granites are also common basement rocks, containing mostly quartz, K-feldspars, biotite and accessory minerals such as titanite, ilmenite and magnetite.

31

Some of the coastal sediments were re-worked from the continental shelf, as a result of the geological and oceanographic evolution of the continental shelf and coastal plains during the Quaternary, in association with the sea level changes. At Campos and Espírito Santo Basins (from sites 26 to 35 in Fig. 1), part of the coastal sediments was also originated from subaerial and wave erosion of Tertiary deposits, known regionally as the Barreiras Group. Only locally, intrusive volcanic, such as basic dikes (Mesozoic diabases) and alkaline rocks (Mostly Paleocene to Eocene sienites), are regarded as sediment sources. The best-known occurrences of these are the coastal Mesozoic dikes, ubiquitous along the Rio de Janeiro and São Paulo States and the alkaline intrusions present on the magmatic alignment extending westwards from Arraial do Cabo and Morro de São João (sites 22 to 25 in Fig. 1). Quaternary sediment deposits, which occur at the southeastern Brazilian coast, can reach wide extensions as in the outfall of the Paraíba do Sul and Doce Rivers (located around sites 26 and 27, and between sites 34

Fig. 1. Geological map of the Brazilian southeast (source: CPRM, 2005; DRM-RJ, 2005). In this illustration the sampling locations indicated by numbers are also included.

32

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

and 35, respectively), while in other areas the coastal plain almost disappears. Usually, sedimentary deposits of the southeastern Brazilian coast are from Holocene to Pleistocene age, and sandy surfaces can be from shallow marine and/or lagoonal origin, covered by eolian sands. Thus, beach surfaces are made by quartzose sands from medium size to very fine grains, from light yellow color to brown. Sediments of eolian origin, in the top layer, are composed by sands of fine quartz grains from yellow– brown color to reddish, often enriched with clay and ferrous oxides. 3. Materials and methods 3.1. Sample collection and preparation Superficial beach sand samples of 35 coastal sites were collected, covering 1500 km along three Brazilian States: São Paulo (SP), Rio de Janeiro (RJ) and Espírito Santo (ES). Fig. 1 shows the collection sites. Some coastal segments were composed by more than one beach, some of them with very difficult access. The sampling sites selection was made on the basis of accessibility and in order to cover equally well the whole coastal range. The coastal zones were subdivided in samples sets representing each particular beach. Each sample set comprised subsets collected from lower to upper beach. From the analysis of these sample subsets, it was possible to evaluate how the natural radionuclides concentrations vary as a function of the distance from the shoreline for each beach. The sample sets analysis allowed to evaluate the mean behavior of the natural radionuclides distribution for each beach of a specific coastal zone and then, from the sum of all these samples, the mean behavior along each coastal zone could be evaluated. In addition, the sample procedure was performed in two seasons of the year, since heavy mineral concentrations in beaches may vary in time and location (de Meijer et al., 2001). Visual inspection (color) and radiation intensity (hand-held Geiger monitor) were used to select the exact sampling spots, in the search for concentrated heavy mineral sand deposits. For each beach, from six to eighteen samples were collected from the upper few millimeters to a few centimeters below the surface layer, depending on the thickness of the concentrated layer. Sand profile samples from the Areia Preta beach (Guarapari–ES) and Buena (RJ) were also measured. Finally, some types of mineral sand samples (such as monazite, zircon, ilmenite, rutile, clay mineral and quartz) from Buena were also analyzed.

Sample preparation was performed at the Laboratory of Radioecology (LARA) of the Physics Institute of the Federal Fluminense University. The samples were oven dried at 110 °C for 24 h, homogenized, ground and screened. Different sieves were used to evaluate the effects of grain size separation on the eTh/eU ratio, which were found to be negligible. Then, only the fraction larger than 2 mm was discarded for the remaining samples. The samples were packed into radon impermeable plastic containers, kept sealed during four weeks, in order to reach the equilibrium of the 238U and 232Th series and their respective progeny. About 200 g of sample material was used for individual sample measurements. 3.2. Calibration and measurements by γ-ray spectrometry Radioactivity concentration measurements were carried out at LARA, using the conventional γ-ray spectrometry technique. A NaI(Tl) detector (sodium iodide gamma scintillator that is thallium activated optically coupled to a photomultiplier tube) and a 10% efficiency hyper-pure Germanium γ-ray detector (HPGe) were used. One 8192-channel pulse height analyzer was connected to each detector. In order to minimize the background radiation, the detection system and the sand sample container were placed inside a shield, comprising copper and lead layers with thicknesses of 2.5 and 12.0 cm, respectively. Spectra from each sample were accumulated during 16 to 24 h. After each measurement, one empty or water filled plastic cylindrical container was placed in the detection system during a counting period of 24 h, in order to collect the background count rates. Background in the detector is partly due to gamma radiation penetrating the shield (originating from building materials and cosmic rays) and the sample material. The density of water is usually lower than that of the sample, but even so corrections for self-absorption are better than for an empty container. However, no significant effects on the extracted values of the activities were observed when empty containers were used in the present experiments. The systems calibration was performed using International Atomic Energy Agency reference materials for K, U and Th activity measurements: RGK-1, RGU-1 and RGTh-1, respectively (IAEA, 1987). The potassium calibration standard (RGK-1) is extra-pure potassium sulphate with 44.8% of K and uranium and thorium contents lower than 0.001 and 0.01 ppm (parts per million), respectively. The uranium standard (RGU-1)

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

is U-ore diluted with silica with 400 ppm of U, a negligible amount of K (less than 20 ppm) and some traces of Th (less than 1 ppm). The thorium standard (RGTh-1) is Th-ore diluted with silica with 800 ppm of Th, but containing some U (6.3 ppm) and K (0.02%). Although the NaI(Tl) γ-ray spectrometer has a poor energy resolution, it is suitable for the purposes of the present work, since its high efficiency allows fast and precise determinations of K, U and Th concentrations in rock, soil or sand samples. The accuracy of the energy calibration depends on possible interferences of each nuclide in each peak region (Grasty et al., 1991; Chiozzi et al., 2000; Anjos et al., 2004, 2005). The procedures for calibration and measurement described by Grasty et al. (1991) and Chiozzi et al. (2000) were adopted in the present work. Energy calibration was performed in the energy range from 0.5 to 3 MeV. The following photon emissions, observed as single or double peaks, were used: 22 Na (511.0 and 1274.5 keV), 137 Cs (661.7 keV), 40 K (1460.8 keV), 208 Tl + 214 Bi (596 keV), 228Ac + 214 Bi (967 keV), 214Bi (609.3, 1120.3 and 1764.5 keV), 214Pb (786.0 keV), 228 Ac (1459.2 keV) and 208Tl (583.0 and 2614.4 keV). The derivation of K, U and Th concentrations of sand samples was performed in two steps for the NaI detector and in one step for the HPGe detector. Initially, single peaks from K, U and Th were identified: 40K (1460.8 keV), 214 Bi (1764.5 keV) and 208 Tl (2614.4 keV). The last two peaks provide the specific activity of parent 238U and 232Th, assuming that the samples are in radioactive equilibrium. For the HPGe detector, the 232Th absolute activity was determined by comparing the peak intensity with that of a 232 Th calibrated sample with the same geometry. This methodology has the advantage of including effects of coincident summing in the calibration procedure. The peak intensity is defined as the difference between the peak area for a sample spectrum and the background contribution. A similar procedure was adopted to determine the absolute activity of 238U and 40K. However, since 228 Ac, a daughter nuclide of 232Th, produces 1459.2 keV γ-rays, which interfere with 40 K (1460.8 keV), the derivation of the 40K activities had a special attention during the spectral analysis. In most situations, when the beach sand samples were composed by low concentrations of heavy minerals, the potassium and thorium contents were such that the 40K activities could be properly corrected for the 232Th admixture. In a few situations when the sample showed an enhancement of the Th content, the subtraction of the 1459 keV line lead to an increase in the uncertainties in the derivation of the activities values

33

of 40K. Thus, when the 232Th activity concentration was ten to twenty times higher than the 40 K activity concentration, it was difficult if not impossible to determine the 40K content accurately, since the subtraction of the 232Th contribution leads to non-significant values of the 40K activities. This problem was present in the analysis of monazite-rich samples. The activities were calculated from the intensity of each line taking into account the branching ratios of the γ-decay, the counting time and the detector efficiency. So, using the sample dry weights, their 40K, 232Th and 238 U concentrations could be expressed in activity per unit mass (Bq kg− 1). Since the correlations between Th, U and K are usually given in equivalent ground concentrations, the concentrations of 238U and 232 Th were expressed in parts per million and 40K in percents, so that the specific parent activity of a sample containing 1 ppm of 232Th and 1 ppm of natural U is 4.08 and 12.3 Bq kg− 1, respectively. For natural potassium, a concentration of 1% by weight of sample corresponds to a 40K specific activity of 317 Bq kg− 1. Systematic uncertainties stem from several sources: the assumption that the samples are in secular equilibrium, the mass and volume of the samples, efficiency calibrations, peak area determination, activity to mass conversion and branching ratios. Random uncertainties are associated with background and sample counts. All these errors were estimated to be of the order of 5% for the Th, U and K contents. For the samples that showed an enhancement of the Th content, the error for 40K was larger, more than 20%. In a few cases in which the activity concentration of 232Th was 20 or more times the 40K concentration, the latter cannot be extracted reliably any more, as its uncertainties were higher in one order of magnitude than the K values themselves. Due to the poor energy resolution of the NaI detector, there were some overlaps between the γ-ray peaks from potassium, uranium and thorium. Therefore, a stripping technique, described by Chiozzi et al. (2000), was applied as a second step to measure the natural radioactivity levels of the samples, making all the required corrections. By this technique, the counting rates in each of the energy spectral windows are related to the concentration Ckj of a given element k (k = 1, 2 and 3 for K, U and Th, respectively) in a given standard sample j by the equation: X Rij ¼ ðAik Ckj Þ þ Bi ð1Þ where Rij is the counting rate for the ith region of the jth standard, i = 1, 2 and 3 for 40K, 238U and 232 Th, respectively, Aik are the calibration coefficients and Bi is

34

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

the background counting rate. Eq. (1) can be expressed in matrix form as ½R ¼ ½A½C þ ½B

ð2Þ

where [R] is a 3 × 1 matrix of the recorded counts, [A] is a 3 × 3 matrix of the calibration coefficients, [C] is a 3 × 1 matrix of the radionuclide concentration and [B] is a 3 × 1 matrix of background counting rates. The known matrix [A] is related to the efficiency of the detection system and it is obtained from the known specific activity of the standard samples and the net counting rates of the corresponding spectral windows. In order to derive the concentration of potassium, uranium and thorium in the unknown samples, the net counting rate of each spectral window (Ri − Bi) must be inserted in the following matrix equation: CK ¼ A−1 fRK −BK g CU ¼ A−1 fRU −BU g CTh ¼ A−1 fRTh −BTh g

ð3Þ

The precision of NaI measurements was corroborated by the HPGe measurements performed on the same samples. The derived uncertainties of natural radionuclide contents of sand samples had the same range for both detectors, as describedpreviously.Thelowerlimitsofdetection(LLD)for 238 U, 232Th and 40K were determined from background radiation level for the above counting time. For the NaI detector, these values were estimated to be 0.4 ppm for uranium, 1.8 ppm for thorium and 0.04% for potassium. For theHPGedetector,0.2ppmforuranium,1.0ppmforthorium and 0.02% for potassium were estimated. 4. Results Sands are defined as mineral grains with diameters between 63 μm and 2 mm and may be divided into light

and heavy mineral fractions. The compositions of heavy mineral suites have been used, during several decades, as provenance indicators for coastal sand deposits and selective transport process, known to be dependent on grain size and density. The current knowledge on the average particle size of the minerals shows that the grain diameter decreases with increasing specific density. For instance, monazite exhibits the highest value of specific density among the heavy minerals group (de Meijer, 1998). Heavy minerals tend to incorporate high concentrations of the naturally occurring radionuclides of the uranium and thorium decay series in their crystalline structure. Light minerals such as quartz and feldspar are low in uranium and thorium, but especially feldspar may contain relatively high concentrations of 40K. Mean values of potassium, uranium and thorium contents in minerals groups (monazite, zircon, ilmenite, rutile, clay mineral, quartz and heavy mineral concentrate — HMC) collected in the Buena deposit (site 27 in Fig. 1) are shown in Table 1. The total heavy mineral content (or HMC) of sands was obtained gravimetrically by floating-off the light minerals in bromoform. One can see the difference in concentrations, with high K and low U and Th values in light minerals and an opposite trend in heavy minerals. Additionally, the eTh/K and eTh/eU ratios also emphasize the very high concentration of thorium in monazite, so that small monazite admixtures will produce a strong impact on the Th concentration of a group of minerals. Therefore, these results indicate that there is a clear correlation between natural radionuclides concentrations and the specific mineral. In order to evaluate the effect of grain size separation on the eTh/eU ratio, and to verify whether the radiometric technique can be used as an alternative method to provide provenance information of the sand deposits, a few samples were separated in grain size

Table 1 Mean values of potassium, uranium and thorium concentrations in minerals and mineral associations of heavy mineral deposits Mineral

K (%)

eU (ppm)

eTh (ppm)

eTh/K

eTh/eU

Quartz Claymineral Zircon Rutile Ilmenite HMC Monazite

0.4 (0.2) 0.9 (0.3) 0.6 (0.3) 0.05 (0.03) 0.03 (0.02) b0.1 a –a

0.4 (0.2) 3.8 (0.5) 270 (40) 36 (5) 13 (2) 100 (10) 2500 (400)

1.2 (0.5) 30 (10) 140 (40) 26 (4) 180 (40) 1100 (100) 45,000 (3000)

3±2 40 ± 20 200 ± 100 500 ± 300 6000 ± 5000 N10,000 –

3±2 8±4 0.6 ± 0.3 0.7 ± 0.2 14 ± 4 11 ± 2 18 ± 4

For each sample five replicates were measured. Values in parentheses represent the standard deviation from the mean. HMC = heavy mineral concentrate. a K content cannot be extracted reliably due to large Th content.

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

35

Table 2 Uranium and thorium concentrations as a function of grain diameter for sand samples taken on the beach of Buena (site 27 in Fig. 1) Grain size ϕ (μm)

K (%)

eU (ppm)

eTh (ppm)

eTh/ eU

2000 b ϕ b 500 500 b ϕ b 250 250 b ϕ b 150 150 b ϕ b 125 125 b ϕ b 106 106 b ϕ b 63 ϕ b 63

0.18 ± 0.02 0.16 ± 0.03 0.2 ± 0.1 b0.2 a b0.3 a b0.3 a 0.2 ± 0.1

b1.0 6.5 ± 0.9 30 ± 3 105 ± 9 270 ± 20 480 ± 40 40 ± 3

b10 59 ± 5 300 ± 20 1000 ± 70 3300 ± 200 5400 ± 300 480 ± 30

– 9±2 10 ± 2 10 ± 2 12 ± 2 11 ± 2 12 ± 2

a

K content cannot be extracted reliably due to large Th content.

fractions where heavy minerals would be isolated in the fine fraction. Table 2 shows the natural radionuclides distribution as a function of grain size for sand samples collected on the Buena beach. From this table one notices that, contrary to the K content, which varies only slightly with grain size, both U and Th contents increase exponentially with decreasing grain size. On the other hand, grain size does not produce any meaningful effect on the eTh/eUBuena ratio, since this ratio presents a mean value of 11 ± 1, similar to the ones obtained from the HMC sample (see Table 1). Similar results were found for the natural radionuclides distributions as a function of grain size for sand samples collected on the Guarapari beach (site 33 in Fig. 1). The grain size for these samples did not produce any significant effect on the eTh/eUGuarapari ratios, which had a mean value of 43 ± 6. The above results suggest that the samples from these two beach groups are different not only from their mineral suite composition, but also from their provenance. Actually, the coastal sediments between the southeast region of Espírito Santo State and the northern region of Rio de Janeiro State are originated from subaerial and wave erosion of Tertiary deposits, known as the Barreiras Group. Additionally, the Quaternary sediment deposits around Buena region also reflect the mineralogy of the hinterland modified by the effects of river transport and ocean currents, since this region is in the outfall of the Paraíba do Sul River. On the other hand, the Guarapari area does not exhibit rivers influence and the coastal plain almost disappears in this region. With the aim of studying the variation of heavy mineral concentrations as a function of time (season of the year) some analyses were performed. The results can be observed in Fig. 2a and b, which shows depth distributions in sand samples of the Areia Preta beach for the Th and U contents, respectively. These sampling were performed in two seasons: end of the Summer and

Fig. 2. Depth distributions of a) thorium content; b) uranium content and; c) eTh/eU ratio in sand samples from the Areia Preta beach (Guarapari, site 33 in Fig. 1). This sampling was performed in two seasons of the year: end of the Summer and beginning of the Spring of 2003.

36

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

beginning of the Spring of 2003. From these figures one notices that the vertical distributions of thorium and uranium contents vary with the season of the year and sampling depth in such way that Th and U concentrations in the surface are higher in the beginning of Spring than in the end of Summer. This behavior can be understood, since in the Winter and beginning of the Spring periods there occur the strongest effects of wind erosion and tide in Brazilian beaches, generating the removal of light minerals from the surface. In Summer the opposite happens and the deposition of light minerals decreases the concentration of heavy minerals in the surface. Additionally, in Guarapari the presence of rocks surrounding the beach makes the sediment transport more unstable in this area. As heavy minerals are transported by wave motions, any obstacles in their

Fig. 3. Depth distributions of a) thorium and uranium contents and; b) eTh/eU ratio in sand samples from the Buena beach (site 27 in Fig. 1).

Table 3 Natural radionuclides distributions as a function of the distance from the shoreline for sand samples from the Mambucaba Histórica beach Distance from shoreline

K (%)

− 15 m Shoreline 15 m 30 m 45 m 60 m 75 m

1.9 2.6 2.4 2.3 1.9 2.2 2.5

(0.2) (0.2) (0.2) (0.4) (0.2) (0.3) (0.2)

eU (ppm)

eTh (ppm)

eTh/ eU

1.9 (0.3) 2.5 (0.8) 1.8 (0.5) 20 (5) 80 (40) 30 (20) 8 (3)

14 (2) 17 (5) 13 (2) 140 (20) 600 (150) 200 (150) 58 (20)

7.2 6.8 7.2 6.8 7.3 7.1 7.2

Mean values of potassium, uranium and thorium concentrations are showed. Five replicates were measured for each distance from shoreline. Values in parentheses represent the standard deviation from the mean.

paths would reflect in the redistribution or re-work of this fraction. These sand profiles could also represent a selective dissolution of certain unstable heavy minerals with the burial (or mechanical abrasion on grains during transport), since this phenomenon is often observed in beach profiles (Morton and Hallsworth, 1999). Anyway, these facts could justify the small fluctuation observed on the eTh/eU ratio as function of depth showed in the Fig. 2c. This figure shows that the mean values of the eTh/eU ratios are almost independent of time and depth variations: (42 ± 8) in Spring and (45 ± 12) in Summer. Fig. 3 shows depth distributions of natural radionuclides in sandy samples collected in Buena, which is a large and stable beach. Concentrations of Th and U from Buena samples exhibit much lower concentrations than those for typical heavy mineral sands, suggesting that the profile is primarily from quartz and other light minerals. On the other hand, a more constant behavior of the eTh/eU ratio can be observed, with mean value of 10.6 ± 0.4, reflecting the stability of this area. Comparing Tables 1 and 2, and Figs. 2 and 3 one can notice that distinct values of eTh/eU ratio have been observed in sand samples from the Guarapari and Buena beaches: eTh/eUGuarapari ≈ 43 and eTh/eUBuena ≈ 11. Additionally, the observed values in each beach have not been significantly modified due to effects of season, sampling depth, grain size or density of the sediments. All these results also suggest that the concentration of light minerals present in the sand sample does not produce significant changes in the eTh/eU ratio. So, the present results indicate that each of the studied sites have in their eTh/eU ratios radiometric signatures, which can be used as provenance indicators for coastal sand deposits. In addition, the radiometry of heavy minerals can be considered easier to measure than the

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

37

Table 4 Mean values of potassium, uranium and thorium concentrations obtained from superficial beach sand samples of 35 coastal zones ID

Coastal zone

Number of beaches analyzed

State

K (%)

eU (ppm)

eTh (ppm)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

Santos Caraguatatuba Mambucaba Praia Brava Porto do Frade Angra dos Reis Mangaratiba Barra de Guaratiba Grumari Pontal Recreio Barra da Tijuca Ipanema Arpoador Copacabana Leme Piratininga Camboinhas Itaipu Itaipuaçu Maricá Arraial do Cabo Cabo Frio Búzios Rio das Ostras Gargaú-Guaxindiba Buena Marataízes Itapemirim Piuma Anchieta Meaipe Guarapari Serra São Mateus

4 3 3 1 2 4 2 1 2 1 3 4 1 1 1 1 2 1 2 2 2 5 3 8 3 3 4 1 1 1 3 2 1 2 1

SP SP RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ RJ ES ES ES ES ES ES ES ES

1.4 (0.1) 2.8 (0.3) 2.3 (0.3) 2.5 (0.3) 2.6 (0.3) 2.6 (0.3) 2.8 (0.3) 0.4 (0.2) 0.2 (0.1) 0.2 (0.1) 0.2 (0.1) 0.2 (0.1) 0.3 (0.1) 0.2 (0.1) 0.2 (0.1) 0.2 (0.1) 0.4 (0.2) 0.1 (0.1) 0.4 (0.1) 0.2 (0.1) 0.2 (0.1) 1.9 (0.3) 0.2 (0.1) 1.0 (0.2) 0.2 (0.1) 0.3 (0.1) 0.3 (0.2) 0.7 (0.1) 1.3 (0.2) 0.2 (0.1) b0.3 a b0.5 a –a b0.4 a b0.3 a

3 (1) 4 (3) 33 (14) 3.1 (0.9) 0.8 (0.5) 1.3 (1.3) 3.2 (1.4) 0.5 (0.5) 0.4 (0.3) 0.6 (0.2) 2.4 (0.6) 1.8 (0.2) 0.4 (0.2) 0.4 (0.2) 2.1 (1.0) 3.1 (0.9) 1.2 (1.1) 0.4 (0.3) 0.4 (0.4) 0.4 (0.4) 1.0 (0.4) 3 (1) 0.8 (0.3) 1.1 (0.6) 1.0 (0.3) 2 (1) 22 (7) 2 (2) 0.4 (0.3) 0.6 (0.4) 40 (20) 80 (20) 300 (100) 15 (5) 60 (30)

12 (4) 6 (6) 200 (100) 2 (1) 7 (1) 11 (3) 18 (11) 4 (2) 2 (20) 2 (1) 2 (1) 2 (1) 6 (1) 7 (2) 6 (4) 5 (3) 7 (4) 4 (1) 6 (1) 6 (2) 4 (1) 12 (2) 3 (2) 5 (3) 3 (3) 20 (8) 230 (90) 9 (9) 3 (1) 2 (1) 600 (300) 1600 (500) 14,000 (5000) 600 (300) 400 (300)

Samples are ordered according to location from southeastern Brazilian coast (see Fig. 1). Values in parentheses represent the standard deviation from the mean. a K content cannot be extracted reliably due to large Th content.

classical heavy mineral analysis method, since some effects (such as grain size and density) on the Th/U ratio can be neglected. Consequently, these results also indicate that γ-ray spectrometry is a suitable tool for the study of sedimentary source of the southeastern Brazilian beaches. Additional information can be obtained from the present results, since the sand samples for each coastal zone were collected from lower to upper beach. Therefore, it is also possible to evaluate how the U, Th and K activities vary with distance from the shoreline and how dependent the activity ratios are on selective transport of sandy sediments. Table 3 presents the persample concentrations for a typical area, where there is an enrichment of heavy minerals. The chosen site was

Mambucaba Histórica, one of three beaches that constitute the Mambucaba site (site 3 in Fig. 1). The results show the existence of a higher heavy mineral deposition in the middle than in the other areas of this beach, indicating that the sediment distribution in such area is primarily from quartz and other light minerals at the shoreline. In addition, it can be noticed that this coastal zone presents a radiometric signature with a stable eTh/eU mean value of 7.1 ± 0.2. Based on the assumption proposed by Kogan et al. (1969) that potassium γ-ray activities of sands having the same origin (parent rock) should be equal, regardless of their mineralogical composition, a comparison between K contents in sand samples of the southeastern Brazilian

38

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

coast and its geological features was performed, as can be observed in Fig. 1 and Table 4. Fig. 1 shows the coastal zone locations where several sand samples sets were collected and Table 4 associates the coastal sites to its respective numbers in the map. Since some coastal segments can be composed by more than one beach, this table also indicates the number of beaches composing each coastal zone. Mean values of K, U and Th contents obtained for each coastal zone are also presented. The analysis of the mean values of potassium content obtained from superficial beach sand samples of 35 coastal zones, shown in Table 4, leads to interesting results concerning the formation of sedimentary strata of these beach sands. The coastal plain involving the north of São Paulo State and south of Rio de Janeiro State (sites 2 to 7 in Fig. 1) presents enrichment of potassium content (with mean value of ≈ 2.6%), whereas there is an extensive zone along the southeast Brazilian coast where the potassium content is depleted by one order of magnitude. In these localities, K contents vary between 0.2% and 0.4%. On the other hand, there are some sites along this coastal zone where potassium contents show some enrichment: ≈ 1.9% in Arraial do Cabo (site 22 in Fig. 1); ≈ 1.0% in Búzios (site 24); ≈ 0.7% in Marataízes (site 28); and ≈ 1.3% in Itapemirim (site 29). Thorium, uranium and potassium concentrations of granitic rocks are intimately related to its mineral compositions and general petrologic features (Whitfield et al., 1959; Rogers and Ragland, 1961; Doveton and Prensky, 1992). Following these authors, thorium uranium and potassium contents tend to be high in felsic rocks and to increase with alkalinity or acidity. Potassium is usually found in potash feldspars, such as microcline and orthoclase, or in micas, like muscovite and biotite. Rocks that are free of these minerals have very low K-activity. The enrichment in the radioactive potassium 40K spread throughout the main rock-forming minerals, such as K-feldspar, present on the pre-Cambrian basement rocks (granites, ortho-gneisses and migmatites), can explain the high value of about 2.6% for the K-activity along the sites on the north São Paulo State and south Rio de Janeiro State. The same interpretation can be used to explain the K enrichment already mentioned in Arraial do Cabo, Búzios, Marataízes and Itapemirim. Fig. 4. Correlations between a) uranium and thorium contents; b) uranium and potassium contents and; c) thorium and potassium contents in 35 sites of the Brazilian southeast. The gray lines in Fig. 4c show that the data can be arranged in zones with distinct values of eTh/ K ratio. The 40K contents of the samples from Espírito Santo State (sites 31, 32, 34 and 35) are represented in b and c by its topmost values, since any lower value would fall within the same zone (see text for details).

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

Finally, the low K-activity (0.2 to 0.4%) in the other coastal zones shows that its sediments do not retain memory of pre-Cambrian formations. When natural gamma radiation measurements of coastal sand deposits are applied to determine the provenance of the sands, it is observed a strong relationship between the natural radionuclides frequency distributions and the geological evolution of the coastal zone. Thus, it is important to verify the correlations between potassium, thorium and uranium contents in the study of the provenance and transport of sediments and its mineral composition. Figs. 4 and 5 show these correlations for the several beaches analyzed in this work. Since some coastal segments were composed by more than one beach and the sampling was performed in several sites of the same beach, in Figs. 4 and 5 the mean values of Th, U and K contents for each beach are shown, in order to allow an overall interpretation for several segments of the Brazilian coast under investigation. Unless there is a complex mixture of radioactive minerals in the formation of the sedimentary strata of beach sands, these cross plots can be used to identify the most common differences between the coastal segments studied. A binary diagram showing thorium and uranium content is illustrated in Fig. 4a, where it is possible to see that eU and eTh have a direct relation. The sand samples

39

from North Rio de Janeiro and Espírito Santo States (see sites 27 and 31 to 35 in Fig. 1) constitute one cluster with the highest values of eU and eTh. The samples from São Paulo State, the remainder coastal segments of Rio de Janeiro State, and a few isolated coastal segments of Espírito Santo State (sites 28 until 30 in Fig. 1) represent another cluster, with the lowest values of eU and eTh. However, this figure shows that there is an anomalous coastal segment situated in Mambucaba (RJ) with high values of eU and eTh. This site is represented by gray squares. Similarly, the same anomalous behavior for Mambucaba can be observed in Fig. 4b and c. Among these figures, the eTh × K diagram presents a better arrangement of the data than the eU × K diagram, since in the latter there is some overlap between a few clusters of sediments with different geological formations. This behavior can be understood as due to the mobility of uranium during the formation of sandy sediments, since, contrary to the potassium and thorium, uranium can be easily oxidized to a water-soluble form and can be readily leached from pegmatites and granites and re-deposited in sediments at large distances from the source rock. On the other hand, in the eTh × K diagram of Fig. 4c, the data are distributed in several sand mineral groups, which can be classified in zones with distinct values of the eTh/K ratio, each of which representing different mineral associations. So, the lines which radiate

Fig. 5. Cross plot of thorium–potassium and thorium–uranium ratios in 35 sites of the Brazilian southeast. The vertical bars represent values of eTh/ eU equal to 2 or 7. The horizontal lines show that the data can be arranged in zones with distinct values of eTh/K ratio or mineral compositions. The eTh/K ratios of the samples from Espírito Santo State (sites 31, 32, 34 and 35) are represented in this figure by its lower values (corresponding to the topmost values of 40K content), since any higher value would fall within the same zone.

40

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

from the origin have gradients matched with eTh/K ratio values that can be used in the broad distinction of the radioactive minerals marked on the plot. Similar behaviors have been observed in several studies of granite rocks (Whitfield et al., 1959; Rogers and Ragland, 1961; Doveton and Prensky, 1992). The potassium–thorium cross plot is nowadays widely used for the recognition of clay mineral associations and the discrimination of micas and feldspars in granitic rocks (Schlumberger, 1986; Macfarlane et al., 1989). As both thorium (by adsorption) and potassium (chemical composition) are associated with clay minerals, the ratio eTh/K expresses relative potassium enrichment as an indicator of clay-mineral species, as well as might be diagnostic of other radioactive minerals (Doveton and Prensky, 1992). In this way, the different sediment types found in the Brazilian coast can be distinguished using their values of the eTh/K ratio obtained from Fig. 4c. The cluster of points from the South coast of Rio de Janeiro and North coast of São Paulo States (sites 2 to 7 in Fig. 1) is clearly separated from other Brazilian coastal segments and can be interpreted as quartz– feldspar systems. This result is corroborated by the measurement of eTh/K ratio from quartz samples (see Table 1). On the other hand, the cluster of points from North Rio de Janeiro (sites 26 and 27) and Espírito Santo States (sites 31 to 35) can be interpreted as heavy thorium-bearing minerals systems. Again, this result is corroborated by the measurements of the eTh/K ratio from heavy minerals samples presented in Table 1. In particular, for sites 31, 32, 34 and 35 for which the 40K content could not be extracted reliably, the results have been represented in Fig. 4b and c by its topmost values as any lower values would fall within the same range. Site 33 (Guarapari) is not shown in these figures, since the activity concentration of 232 Th was 20 times higher than the 40K concentration. Between the two extreme situations of quartz–feldspar systems and heavy thorium-bearing minerals systems, it is possible to notice clusters that present an enrichment of clay mineral associations. Thorium-to-uranium ratio has also proved to be useful in the recognition of “geochemical facies” (Macfarlane et al., 1989). Based on the analyses of numerous rock samples, Adams and Weaver (1958) demonstrated the usefulness of the thorium-to-uranium ratio as an indicator of relatively oxidizing or reducing conditions. Thorium is unaffected by redox conditions and remains insoluble as Th4+. Uranium, however, exists as insoluble U4+ under highly reducing conditions, which leads to U enrichment in sediments, whereas it exists as soluble U6+ under oxidizing conditions, leading

to U loss from sediments. The eTh/eU ratios, therefore, vary from 0 to 2 in anoxic environments and higher than 7 in a strongly oxidizing environment. Thus, the eTh/eU ratio can be used as a proxy for the redox conditions of the depositional environment, suggesting that a relationship with mean sea level may exist. Following Adams and Weaver (1958) when the ratio was computed to be less than two, the depositional environment had promoted uranium fixation under probable reducing conditions, and was most commonly marine. By contrast, ratio values greater than seven, implied uranium mobilization through weathering and/or leaching, and therefore indicated an oxidizing, possibly terrestrial environment. The cross plot of eTh/eU and eTh/K (Fig. 5) summarizes compositional changes related to potassium as well as implications concerning redox potential. The horizontal divisions were obtained from the eTh/K ratios presented in Table 1 for various mineral associations. The vertical bars represent the values of eTh/eU equal to 2 and 7. Based on the eTh/eU ratios and the diagnostic values suggested by Adams and Weaver (1958), an oxidizing environment is indicated for most beach sands samples from the North Rio de Janeiro and Espírito Santo States (sites 26 to 35 in Fig. 1 and Table 4). In this area, the geological and oceanographic evolution of the continental shelf and coastal plains during the Quaternary have created conditions for the formation of large economic heavy mineral deposits, represented by highest values of eTh/K in Fig. 5. These deposits are characterized by distinct eTh/eU values that can be associated with the differentiated transport of sediments along the coastline. The highest values are observed in Guarapari and Serra (both with eTh/eU ≈ 43), which would be consistent from the postulated origin as eolian sands derived from crystalline basement rocks or still from the subaerial and wave erosion of Tertiary deposits. High and medium values are observed in Meaipe (≈20), Anchieta (≈ 16), Buena and Guagaú-Guaxindiba (≈11), São Mateus and Itapemirim (≈ 7) and Marataízes (≈4). These stacked repetitions of the eTh/ eU ratios probably reflect high lateral variability in clastic facies and interplay between mostly brackish and fresh-water regimes of distributary channels, bays and marginal marine deposits, which would be expected to typify a delta complex. From the main geological features of North Rio de Janeiro and Espírito Santo States, one notices that its coastal plain is originally derived from Pre-Cambrian crystalline basement rocks that are considered as

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

primary sources of heavy mineral deposits. A secondary source could be the nearly continuous occurrence of the Barreiras Group Tertiary deposits, which form extensive tablelands with heights between 10 and 100 m, sinking slowly toward the ocean, dissected by local drainage basins. As a result, at least since the Tertiary, heavy minerals have been transported to the shelf due to marine erosion of the Barreiras Group and, during lowstands of the sea level, by the southeast Brazilian coast drainage systems. From the geological features of the coastal plain and the eTh/eU values obtained for North Rio de Janeiro State and from measurements performed between Gargaú and Buena beaches, it is possible to obtain information about the provenance and distribution of

41

heavy minerals on the beaches of Paraíba do Sul River deltaic complex. According to the eTh/eU ratios obtained along the sites 26 to 35 in Fig. 1, the highest values were found in the areas that have no association with a major river (Guarapari and Serra), and so the heavy minerals would be mainly from the breakdown of the active sea cliffs of the Barreiras Group. Two main heavy mineral sources can be defined to estimate the dominant direction of the sediment transport along the coast between Gargaú and Buena beaches: Paraíba do Sul River and the Barreiras Group. The values of the eTh/eU ratios ≈ 11 observed along this coast are consistent with an oxidizing terrestrial environment of alluvial sediments, in such a way the beach suite is a mixture of these two mineral

Fig. 6. Illustration of the geographic model for littoral transport of sediments in the coast of North Rio de Janeiro State (Cassar and Neves, 1993; Gonçalves, 2004).

42

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43

sources. This result is corroborated by the geographic model for littoral transport of sediments in the northern coast of Rio de Janeiro State proposed by Cassar and Neves (1993) and illustrated in Fig. 6, which shows that the Paraíba do Sul River is responsible for an important contribution of sediments to the strandplains at north and south of the river mouth. According to Gonçalves (2004), hornblende, amphiboles, sillimanite, garnets, tourmaline, zircon, biotite and opaques (mainly ilmenite) compose the river suite. Monazite, zircon, tourmaline, rutile, sillimanite, limonite and opaques (mainly ilmenite) compose the suite of the Barreiras Group. The absence of monazite in the river suite could justify the decrease of the eTh/K ratio in this coastal zone, which suggests that the river suit is primarily from quartz and other light minerals. Similar behavior is observed around the site 35 in Fig. 1, while the result of eU/eTh ≈ 7 suggests that the Doce River deltaic complex is also responsible for an important contribution of sediments to the strandplains at north of the river mouth. The same behavior is observed around the sites 28, 29 and 30, with eTh/eU values below 7, suggesting that in addition to the contribution of the Barreiras Group and Novo, Itapemirin and Itabapoana Rivers, some minerals could have shallow marine and/or lagoonal origins, since this area is near the Boa Vista, Mangue, das Pitas and d'Anta lagoons. Going northward of these areas, the eTh/eU ratio increases: Anchieta (≈16), Meaipe (≈20) and Guarapari (≈43), suggesting that the contribution from alluvial, shallow marine or lagoonal sediments tends to disappear. Medium eTh/eU ratios characterize the coastal line along Maricá to Rio das Ostras (sites 21 to 25 in Fig. 1). This result is consistent with this area, since its sandy facies can be explained as from shallow marine or lagoonal origins covered by eolian sands. Stacked repetitions of high, medium and low eTh/eU ratios are observed in the south of Rio de Janeiro State (sites 3 to 20). Again, this behavior suggests a high lateral variability in sandy facies and interplay between mostly brackish and fresh-water regimes of distribution channels, bays and a major marine transgression. In the Southern Rio de Janeiro, next to the division with São Paulo (sites 3 to 7), the strip of the Quaternary sediment almost disappears and the area is surrounded by granitic formations. Not only does the mean potassium value of 2.6% attest the origin of the sediment in such formations, but also the medium values of the eTh/eU ratio represent the fluvial transport which is the main source of heavy minerals. In fact, a new heavy mineral deposit was found around the Mambucaba River deltaic complex (site 3 in Fig. 1). Despite of the Th and U

contents in Mambucaba samples exhibit concentration with the same order of magnitude of Buena, their concentrations are much lower than those for typical heavy mineral sands. However, the data indicate that the Mambucaba deposit could be enriched in zircon or rutile. An analysis in this area is presently in progress, including measurements along the Mambucaba River. 5. Conclusions The present results have demonstrated the usefulness of the traditional and cheap NaI(Tl) detector as a suitable tool for γ-ray spectroscopy studies of natural radioactivity, as far as very careful experiments are performed. Radiometric analysis of Quaternary deposits from the southeastern Brazilian coast showed that coastal plains of North Rio de Janeiro and Espírito Santo States have heavy mineral deposits, each one with specific features for their heavy mineral concentrations. For instance, Guarapari deposits have monazite enrichment above Buena, Serra or São Mateus deposits. Particularly, it was discovered the existence of a deposit in Mambucaba, in which the source of heavy minerals is possibly from alluvial deposits. The results obtained in this work also confirm the main features of the geological mapping of Quaternary coastal deposits, showing that the γ-ray technique can be used as a tool in the understanding of the origin and transport of sediments. Finally, the eTh/ eU ratios proved to be useful when used as proxies for the redox conditions of the depositional environment. Acknowledgments The authors would like to thank the Brazilian financial agencies: CNPq, CAPES and FAPERJ. We are grateful to A.M.A. Santos, J.G. Aguiar and M. Baptista Filho who assisted with sample collection; INB–Buena unit for donating the heavy mineral samples and; Dr. C.G. Silva for his comments on earlier drafts of this manuscript. Constructive comments from the anonymous referee were highly appreciated.

References Adams, J.A.S., Weaver, C.E., 1958. Thorium to uranium ratios as indications of sedimentary processes: example of concept of geochemical facies. Am. Assoc. Pet. Geol. Bull. 42, 387–430. Andrade, A.C.S., Dominguez, J.M.L., Martin, L., Bittencourt, A.C.S.P., 2003. Quaternary evolution of the Caravelas strandplain – Southern Bahia State – Brazil. An. Acad. Bras. Ciênc. 75 (3), 357–382. Anjos, R.M., Okuno, E., Gomes, P.R.S., Veiga, R., Estellita, L., Mangia, L., Uzêda, D., Soares, T., Facure, A., Brage, J.A.P., Mosquera, B.,

R.M. Anjos et al. / Marine Geology 229 (2006) 29–43 Carvalho, C., Santos, A.M.A., 2004. Radioecology teaching: evaluation of the background radiation levels from areas with high concentrations of radionuclides in soil. Eur. J. Phys. 25 (2), 133–144. Anjos, R.M., Veiga, R., Soares, T., Santos, A.M.A., Aguiar, J.G., Frascá, M.H.B., Brage, J.A.P., Uzeda, D., Mangia, L., Facure, A., Mosquera, B., Carvalho, C., Gomes, P.R.S., 2005. Natural radionuclide distribution in the Brazilian commercial granites. Radiat. Meas. 39, 245–253. Asadov, A., Krofcheck, D., Gregory, M., 2001. Application of γ-ray spectrometry to the study of grain size distribution of beach and river sands. Mar. Geol. 179, 203–211. Cassar, J.C.M., Neves, C.F., 1993. Aplicação das Rosas de transporte litorâneo à costa norte-fluminense. Rev. Bras. Eng. 11, 81–103 (in Portuguese). Chiozzi, P., De Felice, P., Fazio, A., Pasquale, V., Verdoya, M., 2000. Laboratory application of NaI(Tl) γ-ray spectrometry to studies of natural radioactivity in geophysics. Appl. Radiat. Isotopes 53, 127–132. CPRM, 2005. Rio de Janeiro Project. Geological Service of Brazil, CPRM. http://www.cprm.gov.br/geo/rjinicio.html (in Portuguese). de Meijer, R.J., 1998. Heavy minerals: from Edelstein to Einstein. J. Geochem. Explor. 62, 81–103. de Meijer, R.J., James, I.R., Jennings, P.J., Koeyers, J.E., 2001. Cluster analysis of radionuclide concentrations in beach sand. Appl. Radiat. Isotopes 54, 535–542. Doveton, J.H., Prensky, S.E., 1992. Geological applications of wireline logs: a synopsis of developments and trends. Log Anal. 33 (3), 286–303. DRM-RJ, 2005. Simplified geological map of Rio de Janeiro State: the rocks say its histories. Departamento de Recursos Minerais do Estado do Rio de Janeiro — DRM-RJ. http://www.drm.rj.gov.br/ admin_fotos/mapa_estado_rj/Mapa_estado.jpg. Gonçalves, Z.C., 2004. Proveniência e distribuição de minerais pesados no complexo deltaico do rio Paraíba do Sul, Master Thesis, UFF. 196 pp. Grasty, R.L., Mellander, H., Parker, M., 1991. Airborne gamma-ray spectrometer surveying. IAEA Technical Reports Series, 323.

43

IAEA, 1987. Preparation and certification of IAEA gamma spectrometry reference materials, RGU-1, RGTh-1 and RGK-1. International Atomic Energy Agency. Report-IAEA/RL/148. Jones, D.G., 2001. Development and application of marine gamma-ray measurements: a review. J. Environ. Radioact. 53, 313–333. Kogan, R.M., Nazarov, L.M., Fridman, Sh.D., 1969. Gamma Spectrometry of Natural Environments and Formations. Moskva Atomizdat, Moscow. 337 pp. Macfarlane, P.A., Whittemore, D.O., Townsend, M.A., Doveton, J.H., Hamilton, V.J., Coyle III, W.G., Wade, A., Macpherson, G.L., Black, R.D., 1989. The Dakota aquifer program annual report, FY89. Appendix B. Kansas Geological Survey, Open-file rept. 90-27. http://www.kgs.ukans.edu/Dakota/vol3/fy89/index.htm. Martin, L., Suguio, K., Flexor, J., Dominguez, J.M.L., Bittencourt, A.C.S.P., 1987. Quaternary evolution of the central part of the Brazilian coast — the role of relative sea-level variations and longshore drift. Unesco Rep. Mar. Sci. 43, 97–115. Martin, L., Suguio, K., Dominguez, J.M.L., Flexor, J., 1997. Geologia do Quaternário Costeiro do Litoral Norte do Rio de Janeiro e do Espírito Santo. CPRM, Belo Horizonte. 104 pp., (in Portuguese). Morton, A.C., Hallsworth, C.R., 1999. Processes controlling the composition of heavy mineral assemblages in sandstones. Sediment. Geol. 124, 3–29. Rogers, J.J.W., Ragland, P.C., 1961. Variation of thorium and uranium in selected granitic rocks. Geochim. Cosmochim. Acta 25, 99–109. Schlumberger, 1986. Log Interpretation Charts. Schlumberger Well Services, Houston. 122 pp. Silva, C.G., 2000. Marine placers. Braz. J. Geophys. 18, 328–334. van Wijngaarden, M., Venema, L.B., de Meijer, R.J., Zwolsman, J.J.G., Van Os, B., Gieske, J.M.J., 2002. Geomorphology 43, 87–101. Whitfield, J.M., Rogers, J.J.W., Adams, J.A.S., 1959. The relationship between the petrology and the thorium and uranium contents of some granitic rocks. Geochim. Cosmochim. Acta 17, 248–271.