Gamma-ray activities are usually measured by scintillometry in well-logging. A conventional paradigm is to link high γ-activity to three parameters: (1) potassium (K) and thorium (Th) contents in clays, (2) high abundances of U-rich detrital minerals contents of clays, (3) U-rich organic matter. Most previous studies concerning the origin of radioactivity in carbonates focused on basin sediments. Using a portable γ-spectrometer, Fiet and Gorin [5] evaluated the radioactivity of organic-rich pelagic sediments from the Aptian–Albian in the Fucoid Marl formation in the Marche–Ombrie Basin (Italy). They measured the total γ-emission and differentiated contributions of K, Th and U and showed that K and Th are good indicators of clay abundance in limestone. The U response was more difficult to interpret because it appeared to be linked not only to the abundance of organic matter in black shales, but also to the thickness of the black shale beds. Adams and Weaver [1] considered that carbonates are less radioactive than any detrital rock.

In shallow-water platform carbonates, high γ-activities are interpreted in the same way, except that these limestones are usually devoid of organic matter. However, in a preliminary γ-scintillometer survey of the ‘Gorges du Nan’ (Nan Canyon, Isère Valley) of the Vercors Massif (western Alps), we found inconsistencies between lithologies and the expected γ-responses. In order to understand the nature of the contradictions, we carried out additional γ-scintillometry field surveys and made detailed analyses of selected samples from sections of the ‘Gorges du Nan’ (Vercors) and the Gorges du Frou (Chartreuse) by low-level γ-spectrometry and ICP–MS.

Vercors and Chartreuse belong to the northern Subalpine Massif of the western Alps (Fig. 1). The sections of the Gorges du Nan and the Gorges du Frou are located in the North of the Vercors and in the centre of the Chartreuse, respectively. Both sections belong to the inner part of an ancient Barremian–Aptian carbonate platform [2–4]. These sections are well exposed and comprise four depositional sequences. Three (BA3, BA4 and BA5) are Upper Barremian in age, and one (AP1) is Lower Aptian in age. The present study focuses on the AP1 depositional sequence and especially on the transgressive systems tract locally called ‘Lower Orbitolina Marls’. This systems tract is characterized by many decimetre- to metre-thick marly levels, which thicken upward and they are interbedded with less argillaceous to pure limestone beds, which become thinner upward. The limestones are oolite- and rudist-rich at the bottom and evolve upwards into dasycladal and Orbitolina-rich limestones. The base of the section is dominated by shallow-water limestones and, in this part of the section, some parasequences end with an emergent surface. In contrast, the upper part of the section is richer in marls and displays mostly argillaceous limestones.

Field measurements were made with a NaI scintillometer (SPP2-NF, SAPHYMO-STEL). One-minute counting times were processed every 40 to 50 cm, except when the section was missing due to poor exposure. To characterize the radioactivity of each measurement station, we chose to use the highest count rate recorded during this time. The validity of this parameterization was checked (i) by comparing the results obtained from multiple 1-min counts at the same station, and (ii) by comparing two sets of results obtained independently at the same station along the Nan section with two different scintillometers. The measurements were made at 62 stations. Thirteen representative measurements of the stratigraphy and lithology samples were selected for low-level gamma analysis. In each case, a ∼250-g sample of bulk rock was powdered and homogenized before measurements were made with the ultra-pure Ge detector of a low-level gamma spectrometer at the Centro Brasileiro de Pesquisas Fisicas in Rio de Janeiro (Brazil). The abundances of U, Th and K were measured by ICP–MS and ICP–AES at the University of Grenoble (France), on aliquots of these samples. Other ICP–MS determinations were made on 45 samples at the University of Neuchâtel (Switzerland).

The portable scintillometer records the γ-ray peaks in a ∼4π steradians detection geometry without differentiating the emitting isotopes (Th, U and their daughter isotopes, K). Thus γ-ray counts obtained by scintillometry in the field cannot be directly compared to detailed analyses made in the laboratory, where the contribution of each γ-emitter can be distinguished. However, the variations of the total γ-activity along a stratigraphic section can be traced in greater detail than by ICP–MS, and by low-level gamma spectrometry, due to the much larger number of measurement stations.

The first concern was to establish the relationship between lithology and γ-counts in the two selected sections. Both lithologies are quite similar: the lower part consists essentially of limestones and the upper part is marl-dominated. In the Nan section, the γ-ray values decrease from bottom to top, whereas they are more or less constant in the Frou section (Fig. 2). These results show that there is no obvious correlation between abundance of clay contents in limestones and marls and high γ-ray activity. On the contrary, the maximum abundance of clays in the Nan section corresponds to the lowest gamma-ray recordings.

The second aim of this study was to establish the relative contributions of Th and U radioactive series and of K in the scintillometry counts. Laboratory low-level gamma spectrometry revealed that the ²³⁵U-, ²³⁸U- and ²³²Th-decay series were at equilibrium in all samples. The study also showed that the γ-emissions were always dominated by the ²³⁸U-decay series, even in the less radioactive limestones [6]. Uranium-series activity represents, in our experimental conditions, 75% of the total number (in dpm g⁻¹) of decay events recorded in the most argillaceous samples and up to ∼95% in the ‘purest’ limestones.

The ICP–MS results show that Th contents are higher (up to ∼2ppm) in the marly beds than in limestones. Similar behaviour is observed for K, with contents up to 100 ppm in the marly levels vs. <0.2 ppm in the limestones. It is clear that Th and K are linked directly to clay occurrence, as observed by many authors [5,7]. Also, the abundance curves of Th and K are very similar in both sections, with the same loci for maximum (marly level) and minimum (limestone) values. Uranium is definitely not linked to lithology. Instead, there is only a poor correlation between high U contents and clay-rich marly beds, and, some of the uranium peaks correspond to limestone levels. The highest peaks of γ-activity correspond to two key surfaces, the first one to a sequence boundary at the base of depositional sequence AP1 and the second one marking the maximum flooding surface indicating the maximum depth during this depositional sequence.

The AP1 sequence boundary corresponds to an erosional surface, locally marked by calcrete. The maximum flooding surface contains the deepest faunal association of the sequence and is enriched in echinoderm fragments. Usually such key surfaces are identified in well logs by their high radioactivity. These results are not unique. Indeed, similar data were obtained in sections of the same age and similar contexts in Spain and Tunisia [6]. The reasons of high uranium contents at sequence boundaries or maximum flooding surface is not yet fully understood, but we note a good correlation between the abundance of echinoderm fragments and high uranium contents in the vicinity of maximum flooding surfaces.

Three methods – portable scintillometer counting in the field, combined with low-level gamma spectrometry and ICP–MS analysis in laboratory – were used to determine the identity and concentration of γ-emitter isotopes in platform carbonate. The results confirm the common idea that K and Th are associated with clays and other mineral in detrital sediments. However, they also show that uranium does not follow the same behaviour and is not associated with clay or any detrital level. In each analysed sample, the isotopes of the uranium series are the dominant source of gamma-activity, the contribution of the K- and of the Th-decay series being subordinate. The idea that high U contents correspond to clay-rich horizons should be revised, at least for carbonates deposited on shallow water platforms.