10081 research outputs found
Sort by
Embracing uncertainties in climate change education: design and implementation of a future-oriented science education approach to develop sustainability competences
Get PDFClimate change is a complex phenomenon that demands an interdisciplinary approach to understanding its scientific, social, economic, and political dimensions. The inherent complexity of climate systems, characterised by nonlinear causality and uncertainties, complicates efforts to predict, educate, and act on climate-related issues. This doctoral research explores how climate change education can address these challenges by fostering sustainability competences, emphasising sustainability, uncertainty, and future thinking. Guided by the European framework for sustainability competences, the GreenComp, the research is structured into two parts. The first examines the context and background of climate change education and includes three case studies. The first case study analyses the role of uncertainties in climate science using risk-based and storyline approaches, assessing their potential to develop sustainability competences. The second investigates sustainability education within a six-month collaboration with the National Consortium of Packaging (CONAI), highlighting the complexity of stakeholder engagement in sustainability. The third focuses on interdisciplinary approaches to uncertainty through an activity at a European summer school for pre-service teachers in STEM disciplines. Building on these insights, the second part presents a novel educational approach integrating sustainability, uncertainty, and futures studies to address climate change. This approach culminated in the course “Towards New Future Scenarios: The Role of Physics in Dealing with Climate Change Challenges,” involving high school students. The course was evaluated using qualitative methods, focusing on learning outcomes related to climate concepts, sustainability competences, and decision-making in uncertain scenarios. Further investigations explored students' responses to personal climate-related experiences and their influence on learning, integrating insights from disaster education during a research visit to the University of Southampton. The research findings, disseminated through publications and collaborations, contribute to advancing climate change education as a tool for fostering critical competences and future-oriented action. The dissertation concludes with recommendations for further research and implementation in educational settings
Vulnerability as a tool to protect environmental migrants
Get PDFThe phenomenon of environmental migration is a pressing and complex issue, exacerbated by climate change and environmental degradation. Despite the increasing prevalence of climate-driven displacement, there is a significant gap in international and EU law regarding the protection of environmental migrants, as neither the 1951 Geneva Convention nor the United Nations Framework Convention on Climate Change (UNFCCC) directly addresses these individuals. This legal void results in environmental migrants being trapped in a state of uncertainty, with no clear status or legal protection. The Research investigates the intersection between environmental migration and the legal frameworks designed to protect vulnerable populations, with a particular focus on the European Union’s response to climate-induced displacement. In particular, the Research explore whether the concept of vulnerability, a dynamic and context-dependent legal tool, can offer a framework for providing protection to environmental migrants under current EU law.
The Research is structured around several key themes. First, it addresses the challenges of conceptualizing environmentally driven migration and its legal implications. It emphasizes the multi-causal nature of migration, where environmental factors intersect with social, economic, and political drivers, complicating the efforts to classify and respond to climate- induced displacement. The second major focus of the Research is on the potential of vulnerability as a tool to enhance legal protection for environmental migrants. Drawing on various theoretical perspectives, including feminist theories, disaster studies, and bioethics, vulnerability is presented as a dynamic and relational concept that can inform legal frameworks in a more inclusive and adaptive manner. A final assessment is made in ascertain whether the analysis presented throughout the Research can effectively close the protection gap for environmental migrants
A machine learning approach in coastal ecology and theoretical advancements in kernel based random forests
Get PDFThis PhD thesis is part of a PON scholarship DOT1303154-3 Dottorati PON - Bando 2021 - Cycle 37 (XXXVII) - Action IV.5 - Doctorates on Green topics. In the first part of the thesis, an application is provided of machine learning algorithms in the ecological coastal coasts. In the second part we examine thoroughly and in depth the mathematical properties of some of the machinery used in the first part, providing theoretical improvements of the models
Development of a CFD tool for turbulent natural convection and heat transfer simulations of liquid metals
Get PDFThis dissertation investigates numerical techniques for studying turbulent natural convection and turbulent heat transfer systems involving liquid metals. Specifically, two strategies are explored. The development of stand-alone solvers to account for all the occurring phenomena, and a code coupling strategy, where two or more numerical codes are integrated to exploit the different code peculiarities. In this work, the latter approach is realized by coupling the in-house finite element code FEMuS with the finite volume code OpenFOAM, using the open-source MED library for data exchange. Turbulent natural convection is studied in a Differentially Heated Cavity configuration. An anisotropic four-parameter turbulence model is implemented in the FEMuS code and validated against the DNS benchmark for liquid metal-filled cavities. To extend this analysis, the coupling application is first validated in the laminar natural convection regime and then applied to the turbulent case. The volume data transfer algorithm is used to leverage the more accurate thermal turbulence model of the FEMuS code with the extensively validated dynamic solver of OpenFOAM. Turbulent heat transfer is investigated in a liquid metal heat exchanger configuration. A boundary data transfer algorithm is validated using a Conjugate Heat Transfer problem, where thermal coupling occurs between the fluid and solid domains. This technique is then applied to a finned pipe heat exchanger, with FEMuS simulating the turbulent liquid metal flow and OpenFOAM modeling heat conduction in the solid structure
Green food purchase: exploring the gap between intention and behavior
No full textAddressing the gap between green attitudes and behaviors is crucial for addressing environmental challenges, especially within the context of green food consumption. Despite increasing environmental awareness, consumers often struggle to translate their intentions into actual purchasing behavior. Bridging this gap can contribute to sustainable consumption, public health improvements, economic and cultural changes, and stronger environmental protection. This doctoral dissertation explores the factors influencing the integration of environmentally sustainable dietary practices and develops a targeted advertising strategy to help companies promote sustainable products effectively and encourage informed consumer choices. The dissertation consists of four studies. Study 1 investigates cognitive predictors of green food purchase intentions, focusing on perceived environmental knowledge, trust in green claims, and skepticism towards green advertising. It finds that trust in green claims mediates the relationship between environmental knowledge and purchase intentions. Consumers with higher trust in credible green advertising are more likely to make sustainable food purchases. Study 2 examines implicit and explicit attitudes towards green food using the Implicit Association Test (IAT) and focus groups. While the IAT results were inconclusive, post-IAT survey data analysis indicates that perceived benefits of green food positively influence purchasing behavior. Focus group discussions highlight barriers to green food consumption, such as tradition versus innovation and preferences for eco-labeling versus descriptive dish names. Study 3 categorizes green food consumers into "Coherent-Buyers," "Non-Buyers with Favorable Intentions," and "Coherent Non-Buyers." The study identifies sociodemographic and emotional factors, such as pride in purchasing sustainable food, that differentiate these groups, emphasizing the need for tailored marketing strategies. Study 4 tests a green food nudge strategy in a local canteen, finding that while overall sales did not significantly change, targeted interventions were effective for specific dish categories. This research contributes to the understanding of green food consumption behavior by integrating cognitive, emotional, and attitudinal factors
Performance evaluation of low cost sensor to monitor urban air pollution
No full textThis study evaluates the performance of Smart Citizen Kits (SCKs), a network of low-cost sensors, in monitoring urban air pollution, with a focus on PM2.5, PM10, carbon dioxide (CO2), ozone (O3), nitrogen dioxide (NO2), temperature, and humidity. SCK measurements were compared to reference-grade instruments (ARPAE) during collocation campaigns in Bologna, Italy, during both summer and winter. The results indicate that while SCKs effectively capture the temporal trends of pollutants and environmental parameters, they show discrepancies in absolute concentrations, particularly for PM10 and NO2, due to cross-sensitivity, environmental influences, and sensor-specific biases. Strong correlations were observed between temperature (R² = 0.89) and humidity, confirming the reliability of SCKs for environmental monitoring. However, moderate correlations for PM2.5 (R² = 0.59) and low correlations for PM10 (R² = 0.19) highlight the need for post-data correction methods to improve quantitative accuracy. The study also demonstrated the influence of meteorological factors, such as relative humidity and wind speed, on sensor performance and pollutant dispersion. This study underscores the potential of low-cost sensor networks for tracking urban air quality variability, provided that calibration and correction techniques are applied to enhance their accuracy and reliability for regulatory and scientific applications
Developments in machine learning downscaling for storm surge in the northern Adriatic sea
No full textAccurate storm surge prediction is essential for coastal resilience and hazard mitigation, particularly as climate change increases the frequency and intensity of extreme events. While Machine Learning (ML) models have shown promise for downscaling storm surge predictions, they are often under-evaluated against high-resolution dynamical models and rarely tested on extreme conditions. This study addresses these limitations by integrating advanced dynamical modeling with ML approaches to evaluate storm surge prediction in the Northern Adriatic Sea. High-resolution simulations were developed using the SHYFEM hydrodynamic model with optimized configurations and high-quality forcing datasets. This provided a robust benchmark for assessing the performance of ML models ranging from simple Multivariate Linear Regression (MLR) to more complex Long Short-Term Memory (LSTM) networks. To improve model evaluation, a novel corrected mean absolute deviation (MADc) metric and a custom loss function (MADc2) were introduced, targeting improved performance on extreme events. Results show that while MLR offers computational efficiency, it lacks the ability to capture non-linear dynamics and extremes. In contrast, LSTM networks performed significantly better, particularly when trained using the MADc2 loss function. Training ML models on the output of the dynamical model led to strong consistency with observations, while direct training on tide gauge data at key locations (e.g., Punta della Salute and Trieste) revealed that some ML models could outperform the dynamical model in critical metrics. These findings highlight the potential of ML models, especially LSTM networks, as efficient and accurate alternatives to traditional numerical approaches. Given their lower computational demands and strong performance, ML techniques hold significant promise for operational storm surge forecasting, particularly in data-rich contexts or where computational resources are limited
Technological advances in external beam radiotherapy. Study of new treatment strategies.
No full textThis research covers the use of advanced technologies in external beam radiotherapy (EBRT), such as proton therapy for thoracic tumors, a central issue with significant implications for radiation oncology. By combining motion management, adaptive planning, and radiobiological modeling, my Ph.D. work aims to better understand the issue and prospect of proton therapy optimization. The primary objectives of this research are to evaluate the image-guided proton therapy (IGPT) procedures, determine the efficacy of online adaptive proton therapy (oAPT), and estimate secondary cancer risk using mechanistic models. To reach these objectives, my study uses a mixed-method approach involving systematic literature review, dosimetric simulation, and collaborative clinical research by leveraging multicentric scientific collaborations. Through the use of probabilistic robustness evaluation tools, computational-based secondary malignancy risk models, and comparative case studies, the study explores dose conformity, plan adaptation, and long-term biological outcomes. Key findings include the identification of persistent uncertainties in thoracic IGPT and evidence supporting the superiority of pencil-beam scanning proton therapy in mitigating secondary cancer induction in ultra-fractionated radiation therapy. These findings suggest that enhanced motion management and radiobiological modeling enhance the therapeutic index of proton therapy, indicating broader utility to treatment personalization. Findings from this research are expected to advance medical physics and radiation oncology, offering new knowledge on IGPT application particularly relevant for hybrid proton and photon therapy approaches.The work investigated how to develop hybrid treatments with radiobiological modeling for risk computation that identifies a solid basis for future studies in next-generation personalized Radio-oncolog
The Facetus “Cum nihil utilius”. Study and edition
No full textLa presente ricerca si concentra sullo studio e sull’edizione del Facetus «Cum nihil utilius», un testo mediolatino in versi risalente al XIII secolo. L’opera, tra i più fortunati testi scolastici del medioevo latino, si propone come complemento ai Disticha Catonis, offrendo una serie di precetti di buon comportamento. Con una tradizione manoscritta ampia e complessa (109 testimoni noti, più otto perduti), nonché una ricca storia di traduzioni e ricezione nelle letterature volgari europee, il Facetus ha suscitato l’interesse della critica soprattutto per le sue versioni volgari, lasciando in secondo piano lo studio del testo latino originale. Il mio progetto mirava a colmare questa lacuna attraverso un riesame sistematico della tradizione manoscritta, al fine di proporre un’edizione critica ricostruttiva che superasse i limiti delle due edizioni oggi disponibili. In considerazione della complessità della tradizione, si è optato per un’edizione provvisoria fondata sugli otto manoscritti più antichi (XIII secolo), accuratamente descritti nell’introduzione. Segue un censimento completo dei testimoni, molti dei quali ignoti agli studi precedenti. L’edizione è accompagnata da un apparato critico completo delle varianti (eccetto quelle di scarsa rilevanza e quelle ortografiche, comunque documentate in appendice), da una traduzione italiana a fronte e da una serie di appendici che integrano il lavoro filologico: le lezioni escluse dall’apparato, i versi interpolati, una tabella sinottica della forma del testo nei manoscritti alla base dell’edizione.This study examines and edits the Facetus «Cum nihil utilius», a 13th-century Latin verse text conceived as a complement to the Disticha Catonis and widely used in medieval schools. Despite its broad and complex manuscript tradition (109 known witnesses and eight lost) and its significant vernacular reception, modern scholarship has largely neglected the Latin original. The project offers a provisional critical edition based on the eight earliest manuscripts, accompanied by a full census of the tradition, many witnesses previously unrecorded. The edition includes a critical apparatus (with minor and orthographic variants listed in appendices), a facing Italian translation, and supplementary materials such as interpolated verses and a synoptic table of the textual forms preserved in the base manuscripts
Leveraging the power of data to provide better healthcare services
No full textUtilizing healthcare data can significantly enhance service delivery, improve patient outcomes, and optimize operational efficiency. This PhD project focuses on developing and applying advanced statistical and machine learning models to analyze complex datasets for accurate prediction and improved decision-making.
In Study 1 – Part I (Chapter 2), we predicted Length of Stay (LoS) and Prolonged LoS for emergency department inpatients at Sant’Orsola Malpighi University Hospital. Notably, Gradient Boosting excelled in PLoS prediction, while Ridge and XGBoost performed well in LoS prediction.
Study 1 – part II (Chapter 3), we specifically targeted the General Medicine department, characterized by high patient volume and heterogeneity. We compared nine ML regression models in predicting hospital LoS. Feature Importance plots and SHAP (SHapley Additive exPlanations) were employed to identify the top important features and enhance interpretability. The eXtreme Gradient Boosting Regression model had the lowest prediction error.
Study 2 (Chapter 4), conducted, when COVID-19 was at its peak, aimed to analyze the spatio-temporal patterns of the diffusion of SARS-CoV-2. The study aimed to derive a model for infection risk and identify place-specific factors, revealing varied impacts across city areas during the first three epidemic waves, with an estimated area-to-area influence within a 4.7 km radius.
Study 3 (Chapter 5) aimed to identify admission risk factors associated with Length of Stay using Poisson, negative binomial, and Hurdle regression models. Hurdle–NB provided the best fit model. The ICU setting and long-term hospitals significantly influenced LoS, and age, wave periods, and hospital types also played crucial roles.
Finally, Study 4 (Chapter 6) is an ongoing project targeting childhood cancer survivors, aiming to predict the risk of nonsurgical premature menopause using survival models and machine learning. The emphasis is on developing a landmark predictive model based on treatment-induced toxicities to improve long-term health outcomes for female childhood cancer survivors