Open access to publications, including peer-reviewed articles, conference papers, working papers, reports, and other scholarly communications by University of Galway researchers.
University of Galway Research Repository
Recent Submissions
Publication Open Access Spatially heterogeneous controls of soil organic carbon in a karst mountainous area of southern China: Insights from interpretable machine learning(Elsevier, 2026-03-29)Understanding the spatial variability and environmental drivers of soil organic carbon (SOC) is critical for improving carbon management in fragile karst landscapes. This study collected 110 topsoil samples across county Yangshan, southern China, and applied an interpretable machine learning framework combining Random Forest (RF) and SHapley Additive exPlanations (SHAP) to explore the spatial heterogeneity and key environmental controls of SOC. The measured contents ranged from 3.33 to 44.20 g/kg, with a coefficient of variation of 43.5%, indicating moderate variability of SOC in the study area. The RF-based spatial predictions revealed that higher SOC levels were mainly concentrated in the northern and southern subregions associated with clastic rocks, while lower SOC values clustered in central areas dominated by carbonate bedrocks. SHAP analysis indicated that soil physicochemical properties contributed over 53% to SOC, with total nitrogen and cation exchange capacity exerting the strongest influences, particularly in karst zones. Hydrological, vegetation, and terrain-related factors showed moderate importance, especially in high-elevation areas with natural vegetation and complex topography that promoted SOC accumulation. In contrast, climatic variables had relatively weak impacts, with their influences clustered in lowlands dominated by anthropogenic land uses. These findings revealed spatially heterogeneous controls on SOC between karst and non-karst landscapes, emphasizing the dominant role of soil properties under shallow, erosion-prone conditions and highlighting the role of topography and vegetation in enhancing SOC stocks in mountainous areas. The integrated use of interpretable machine learning approaches improves the understanding of localized SOC dynamics and provides a valuable reference for precision carbon management and ecological restoration in environmentally sensitive regions elsewhere.Publication Embargo Spectral data augmentation and classification: Methods, resources, and applications for raman and near-infrared spectroscopy(University of Galway, 2026-04-17)Spectral analysis methods have become increasingly interdisciplinary and ubiquitous across fields including pharmaceutical development, agriculture, and food safety, among others. Recent literature has focused on generative artificial intelligence (AI) and complex deep learning models to address the persistent challenge of limited data. This challenge arises from the niche nature and expensive processes associated with generating large spectral datasets. While deep learning and generative AI methods offer the potential to expand spectral datasets, an understanding of the fundamental spectral data characteristics, as well as the intricacies of the state-of-the-art, is essential to avoid over-engineered models that fail to generalise. Consequently, many traditional methods that respect the inherent properties of spectral data are often overlooked, with current literature rarely providing guidance on their implementations and implications. Our contributions aim to address these challenges surrounding resource accessibility and computational analysis methods for two vibrational spectroscopy techniques, namely Raman and near-infrared (NIR) spectroscopy. Specifically, this thesis provides five key contributions in computational spectroscopic analysis. First, we present a systematic review and tutorial on state-of-the-art spectral preprocessing, data augmentation, and generative AI methods. This provides essential guidance and clarity on the methods for spectroscopic applications while addressing gaps in foundational techniques, reproducible implementations, and the ethical considerations involved in managing data integrity. Second, we investigate the impact of synthetic data augmentation on deep neural networks (NN) trained on limited Raman spectral datasets, establishing upper bounds on synthetic data requirements and evaluating the cost-benefit considerations in terms of computational resources and implementation effort. Third, we demonstrate that simpler modelling approaches can achieve competitive performance when deep learning is impractical, specifically showing how one-vsrest (OVR) classification strategies outperform traditional multi-class approaches for NIR spectra. Fourth, we introduce a comprehensive open-source Raman spectral dataset comprising 3,510 spectra of thirty-two pure solvents and reagents commonly used in active pharmaceutical ingredient (API) development. We outline the protocols for acquiring, annotating, and releasing this dataset, and present our analysis on improving data quality alongside benchmark evaluations using machine learning methods. Finally, we demonstrate how transfer learning and data augmentation can significantly improve both model robustness and state-of-the-art performance when working with limited data. These contributions provide practical advice for selecting appropriate computational methods in spectroscopic analysis. They also demonstrate the conditions under which synthetic data augmentation provides genuine benefits and establish alternative classification strategies that can outperform complex deep learning approaches. Additionally, they make valuable opensource resources available to address data accessibility challenges for pharmaceutical-based tasks and utilise these resources in practical settings to improve performance and generalisability.Publication Embargo Advanced humanised bone model recapitulating molecular and functional signatures of osteoporotic bone for therapeutic research(University of Galway, 2026-04-16)Postmenopausal osteoporosis is a debilitating bone disease that is associated with bone loss and increased risk of fractures of hip, wrist and vertebrae, which impacts 200 million women worldwide with significant medical and economic burden and is expected to grow with the aging population. Although several anabolic and anti-catabolic therapeutic interventions have been developed to treat osteoporosis, 50-70% of treated individuals still experience an osteoporotic fracture depending on the therapeutic intervention. Decline in circulating estrogen levels after the menopause in women alters the cellular function of osteoblasts and osteoclasts, damages bone vasculature, reduces osteocyte mechanosensitivity, increases sclerostin production (WNT inhibitor) and impacts survival of bone cells. Overall, this leads to an increase in bone turnover, with bone resorption outpacing formation, and bone loss, but also changes in bone tissue mineral composition and distribution. The mechanisms driving these changes have been explored; estrogen deficiency alters the mechanobiological responses of osteocytes, enhances osteocyte-mediated osteoclast activity and causes heterogeneity in the mineral distribution but the underlying mechanisms are not fully understood. Romosozumab, a humanized monoclonal antibody against sclerostin, offers therapeutic potential for postmenopausal osteoporosis. It binds to sclerostin and prevents the inhibition of the WNT/β-catenin pathway and thereby promotes osteogenesis. However, the mechanisms of the therapeutic efficacy remain poorly understood. Most of our understanding about postmenopausal osteoporosis and therapeutics comes from ovariectomized animal models, which do not accurately reproduce human bone dynamics or predict clinical therapeutic responses due to interspecies differences. Although recent in vitro bone models have been developed, these simplified models cannot reproduce human multicellular responses, limiting their translational potential. Thus, there is a distinct need for advanced 3D models of human bone, which can account for paracrine regulation by multiple cells (osteocytes, osteoblasts, osteoclasts, vascular cells) and biophysical conditions. Furthermore, the mechanistic effects of estrogen deficiency on bone vasculature are not fully understood. In the first study of this thesis, (1) an advanced 3D vascularized, mineralized and humanized bone model was developed by following an endochondral ossification priming approach, and (2) this model was applied to mimic postmenopausal osteoporosis and provide a mechanistic understanding of changes in vascularization and bone mineralization in estrogen deficiency. The study confirmed the successful development of a humanised multicellular bone model, which induced formation of vasculature, associated with hypertrophy (collagen X), and promoted mineralization. When the model was applied to study estrogen deficiency, the development of distinct vessel-like structures (CD31+) in the postmenopausal 3D constructs was observed. Moreover, during estrogen withdrawal vascularized bone demonstrated a significant increase in mineral deposition and apoptosis, which did not occur in non vascularized bone. These findings reveal a potential mechanism for bone mineral heterogeneity in osteoporotic bone; whereby vascularized bone becomes highly mineralized whereas in non vascularized regions this effect is not observed. In the second study of this thesis, in vitro vascularized bone models were advanced by incorporating cyclic mechanical stimulation using a commercial bioreactor, to account for the influence of biophysical stimuli that exist in vivo. The results emphasize the need to incorporate biophysical stimulation in studies of skeletal biology and pathogenesis by demonstrating that mechanical loading with estrogen (healthy condition) enhanced mineral production, hypertrophy, apoptosis and vascularization. In contrast, mechanical loading with estrogen withdrawal (disease condition) increased collagen 1 and accelerated the transition of osteoblasts to osteocytes, which was associated with pathological hypertrophy and apoptosis. The findings revealed an interplay between estrogen signalling and mechanical loading in a 3D multicellular, vascularized and humanized bone microenvironment. In the third study, human osteoclasts were included in the bone mimetic models to study estrogen deficiency, sclerostin inhibition and osteocyte driven osteoclastogenesis. Using this advanced 3D bone mimetic model, it was demonstrated that estrogen withdrawal disrupted WNT/β-catenin signalling by enhancing sclerostin production leading to increased osteoclastogenic signalling. Importantly, treatment with an anti-sclerostin monoclonal antibody (romosozumab) partially restored WNT signalling and attenuated osteocyte induced osteoclastogenesis. Furthermore, estrogen withdrawal enhanced fibronectin production in the models, which was consistent with the transcriptomic analysis of human osteoporotic bone. This increase in fibronectin may represent a compensatory ECM response to associated with heterogeneity in the mineralization observed postmenopausal osteoporosis. Interestingly, sclerostin inhibition increased fibronectin network formation indicating a previously unexplored link between sclerostin and ECM organization. Together, this research thesis realized the generation of an advanced humanized multicellular bone mimetic platform and established its translational relevance to study postmenopausal osteoporosis therapeutics. The model elucidated the effect of estrogen withdrawal on vascularization and mineralization dynamics under physiological mechanical loading. Furthermore, role of estrogen deficiency in inhibiting WNT/β-catenin pathway and driving osteocyte-mediated osteoclastogenic signalling was explored. The therapeutic anti-sclerostin antibody (romosozumab) was tested on the models to explore anti-resorptive and anabolic potential. The translational relevance of the model was investigated by comparing its transcriptional profiles of osteoporotic human bone specimens. Together, this research provides a valuable platform for studying osteocyte-ECM-osteoclast interactions and for evaluating next generation therapies in a physiologically relevant niche.Publication Embargo Decolonising law in the postcolonial nation-state in Sub-Saharan Africa: The case of Sierra Leone(University of Galway, 2026-04-16)This thesis examines the enduring legacies of colonial law in Sierra Leone and explores pathways for decolonising the nation’s legal order. It introduces the concept of the colonial deficit as a unifying analytic to explain how inherited legal structures—spanning land tenure, criminal justice, plural legal systems, and constitutional governance—continue to undermine sovereignty, equity, and reform. Drawing on postcolonial theory, TWAIL, and Olúfẹ́mi Táíwò’s framework of African agency, the study situates Sierra Leone within broader sub-Saharan patterns of continuity, hybridity, and rupture in legal reform. Through a critical historical and comparative analysis, the thesis demonstrates how colonial legal architectures persist in shaping social, political, and epistemic hierarchies, producing exclusionary practices in both statutory and customary domains. It engages with the findings of the Sierra Leone Truth and Reconciliation Commission and comparative experiences from Ghana, South Africa, Rwanda, and Uganda to illuminate possibilities for harmonising customary and statutory systems, promoting gender inclusion, and advancing participatory constitutionalism. The study proposes an integrated reform roadmap centred on four interdependent domains: land reform, criminal justice reform, constitutional redesign, and epistemic renewal. By linking structural reform with epistemic transformation, the roadmap demonstrates how law can be reimagined not merely as an inherited scaffold but as a tool for sovereignty, justice, and African-led norm creation. Ultimately, the thesis contributes to postcolonial legal scholarship by offering a contextually grounded analysis of Sierra Leone, advancing theoretical frameworks for decolonisation, and providing actionable strategies to reclaim legal futures. It affirms that the decolonisation of law is not simply a process of critique or reversal but a forward-looking project of creation—one that recognises African agency, pluralism, and the potential for emancipatory justice across generations.Publication Embargo A climate aware, mechanics consistent framework for predicting leading edge erosion in wind turbine blades(University of Galway, 2026-04-15)This thesis investigates leading edge erosion in wind turbine blades and develops a framework linking environmental conditions, material behaviour, and coating durability. The study combines meteorological analysis, numerical modelling, and analytical formulation to predict how erosion evolves under current and future climate conditions in Ireland. The research begins by integrating finite element modelling with meteorological data from multiple Irish sites to generate an erosion index for comparing regional erosion potential. The results reveal clear spatial variability, showing higher predicted damage than inland regions due to stronger rainfall intensity and droplet impact speeds. The second stage examines the influence of climate change using projected rainfall characteristics from global and regional models. The analysis indicates that future scenarios generally increase erosion severity, driven mainly by changes in rainfall intensity rather than total precipitation, highlighting the importance of short-duration high-intensity events. A semi-analytical model is then developed to describe the transient elastodynamic response of layered coatings under axisymmetric droplet impact. The model captures stress-wave propagation in multilayer systems and shows that surface-guided motion dominates the short-time response, helping explain why erosion damage initiates near the coating surface. Finally, finite element simulations assess the performance of different adhesive systems in blade structures. The results demonstrate that adhesives with higher energy absorption capacity reduce interfacial stresses and delay debonding, improving coating durability under repeated impacts. Together, these findings provide a unified framework linking environmental exposure, wave-driven mechanical response, and material behaviour, offering guidance for designing erosion-resistant coatings and adhesive systems suited to Ireland’s evolving climatic conditions.