U and Th content in the Central Apennines continental crust: a contribution to the determination of the geo-neutrinos flux at LNGS

The regional contribution to the geo-neutrino signal at Gran Sasso National Laboratory (LNGS) was determined based on a detailed geological, geochemical and geophysical study of the region. U and Th abundances of more than 50 samples representative of the main lithotypes belonging to the Mesozoic and Cenozoic sedimentary cover were analyzed. Sedimentary rocks were grouped into four main “Reservoirs” based on similar paleogeographic conditions and mineralogy. Basement rocks do not outcrop in the area. Thus U and Th in the Upper and Lower Crust of Valsugana and Ivrea-Verbano areas were analyzed. Based on geological and geophysical properties, relative abundances of the various reservoirs were calculated and used to obtain the weighted U and Th abundances for each of the three geological layers (Sedimentary Cover, Upper and Lower Crust). Using the available seismic profile as well as the stratigraphic records from a number of exploration wells, a 3D modelling was developed over an area of 2^{\circ}x2^{\circ} down to the Moho depth, for a total volume of about 1.2x10^6 km^3. This model allowed us to determine the volume of the various geological layers and eventually integrate the Th and U contents of the whole crust beneath LNGS. On this base the local contribution to the geo-neutrino flux (S) was calculated and added to the contribution given by the rest of the world, yielding a Refined Reference Model prediction for the geo-neutrino signal in the Borexino detector at LNGS: S(U) = (28.7 \pm 3.9) TNU and S(Th) = (7.5 \pm 1.0) TNU. An excess over the total flux of about 4 TNU was previously obtained by Mantovani et al. (2004) who calculated, based on general worldwide assumptions, a signal of 40.5 TNU. The considerable thickness of the sedimentary rocks, almost predominantly represented by U- and Th- poor carbonatic rocks in the area near LNGS, is responsible for this difference.

💡 Research Summary

The paper presents a detailed regional study aimed at refining the prediction of the geo‑neutrino signal measured at the Gran Sasso National Laboratory (LNGS) in Italy. Geo‑neutrinos, electron antineutrinos emitted in the decay chains of ^238U, ^232Th and ^40K, carry information about the distribution of heat‑producing elements inside the Earth. While the global Reference Model (RM) of Mantovani et al. (2004) provides a first‑order estimate of the expected signal based on worldwide average concentrations of U and Th and a 2° × 2° crustal tile map, it does not capture local geological heterogeneities that can significantly affect the signal because the contribution from the crust within a few hundred kilometres of the detector dominates the total flux.

To address this limitation, the authors constructed a Refined Reference Model (RRM) for the Central Apennines, the region underlying LNGS. The work involved (i) systematic sampling of more than 50 rock specimens, including 28 sedimentary samples (both within 20 km and up to 200 km of the laboratory) and 22 samples from the upper and lower crust (collected ex‑situ from the Ivrea‑Verbano and Valsugana areas), (ii) chemical analysis of U and Th using inductively coupled plasma mass spectrometry (ICP‑MS) with detection limits of 0.01 ppm and an estimated accuracy of 10 %, complemented by gamma‑spectrometry with a NaI(Tl) detector for cross‑validation, and (iii) construction of a three‑dimensional geological model covering a 2° × 2° area (≈ 220 km × 220 km) down to the Moho, with a total volume of ~1.2 × 10⁶ km³. The 3‑D model incorporates seismic profiles from the CROP project, stratigraphic data from wells, and surface geological maps, allowing the authors to assign precise thicknesses and volumes to three main layers: the sedimentary cover (SC), the upper crust (UC) and the lower crust (LC). Within each layer, rocks were grouped into four “reservoirs” based on paleogeographic setting and mineralogy (e.g., carbonate‑rich low‑U/Th sediments, siliciclastic sediments, mafic and felsic crustal rocks). Weighted average concentrations of U and Th for each reservoir were derived from the analytical data.

The regional contribution to the geo‑neutrino flux, S_reg, was calculated by integrating the product of element concentration, rock density, and the 1/r² geometric factor over the entire 3‑D volume. Uncertainties from analytical errors, reservoir volume estimates, and density variations were propagated using Monte‑Carlo simulations. The resulting refined signal for the Borexino detector is S(U) = 28.7 ± 3.9 TNU and S(Th) = 7.5 ± 1.0 TNU, where 1 TNU corresponds to one event per 10³² target protons per year. This is about 4 TNU lower than the 40.5 TNU predicted by the original RM. The discrepancy is primarily attributed to the thick sedimentary cover around LNGS, which is dominated by carbonate rocks that are intrinsically poor in U and Th, whereas the RM assumed higher average concentrations for sediments.

The study demonstrates that (1) local geological and geochemical characterization can substantially modify geo‑neutrino flux predictions, (2) the dominant contribution to the signal indeed originates from the near‑field crust (≈ 400 km radius), and (3) precise regional modeling is essential for interpreting geo‑neutrino measurements in terms of global mantle composition and radiogenic heat production. The authors conclude that similar high‑resolution approaches should be applied to other detector sites (e.g., KamLAND, SNO+) to reduce systematic uncertainties and to enable geo‑neutrinos to become a robust probe of Earth’s interior.