A new ground level neutron monitor for space weather assessment.

We report on a new ground-level neutron monitor design for studying cosmic rays and fluxes of solar energetic particles at the Earth's surface. The first-of-its-kind instrument, named the NM-2023 after the year it was standardised and following convention, will be installed at a United Kingdom Meteorological Office observatory (expected completion mid 2024) and will reintroduce such monitoring in the UK for the first time since ca. 1984. Monte Carlo radiation transport code is used for the development and application of parameterised models to investigate alternative neutron detectors, their location and bulk material geometry in a realistic cosmic ray neutron field. Benchmarked against a model of the current and most widespread design standardised in 1964 (the NM-64), two main parameterisation studies are conducted; a simplified standard model and a concept slab parameterisation. We show that the NM-64 standard is well optimised for the intended large-diameter boron trifluoride (BF 3 ) proportional counters but not for multiple smaller diameter counters. The new design (based on a novel slab arrangement) produces comparable counting efficiencies to an NM-64 with six BF 3 counters and has the added advantage of being more compact, lower cost and avoids the use of highly toxic BF 3 .

Convolutional Neural Networks for Challenges in Automated Nuclide Identification.

Improvements in Radio-Isotope IDentification (RIID) algorithms have seen a resurgence in interest with the increased accessibility of machine learning models. Convolutional Neural Network (CNN)-based models have been developed to identify arbitrary mixtures of unstable nuclides from gamma spectra. In service of this, methods for the simulation and pre-processing of training data were also developed. The implementation of 1D multi-class, multi-label CNNs demonstrated good generalisation to real spectra with poor statistics and significant gain shifts. It is also shown that even basic CNN architectures prove reliable for RIID under the challenging conditions of heavy shielding and close source geometries, and may be extended to generalised solutions for pragmatic RIID.