DEVELOPMENT OF ADVANCED TOOLS FOR MODELLING, PREDICTION AND MAPPING OF ENVIRONMENTAL NOISE AND SOUND PERCEPTION

ENVIRONMENTAL NOISE, PARTICULARLY FROM ROAD TRAFFIC, REPRESENTS A PERVASIVE GLOBAL HEALTH HAZARD THAT PROFOUNDLY IMPACTS HUMAN WELL-BEING, COGNITIVE FUNCTION, AND QUALITY OF LIFE. WHILE TRADITIONAL NOISE ASSESSMENT METHODOLOGIES RELY HEAVILY ON PHYSICAL ENERGETIC INDICATORS SUCH AS THE A-WEIGHTED EQUIVALENT SOUND PRESSURE LEVEL LA,EQ, THESE METRICS OFTEN FAIL TO CAPTURE THE QUALITATIVE COMPLEXITY OF HUMAN AUDITORY PERCEPTION. THIS DISSERTATION ADVANCES THE FIELD OF ENVIRONMENTAL ACOUSTICS BY ADOPTING A DUAL-LENS PERSPECTIVE THAT ADDRESSES THE TECHNICAL CHALLENGES OF PHYSICAL NOISE SOURCE MODELLING AND THE SUBJECTIVE COMPLEXITY OF SOUNDSCAPE APPRAISAL. CRUCIALLY, BOTH APPROACHES ARE SYSTEMATICALLY INVESTIGATED ACROSS THEIR SPATIAL AND TEMPORAL DIMENSIONS, ACCOUNTING FOR THE SPATIAL CONFIGURATION OF NOISE SOURCES AND THE MAPPING OF PERCEPTIONS, AS WELL AS THEIR DYNAMIC EVOLUTION OVER TIME. THE FIRST PILLAR OF THIS RESEARCH INTRODUCES A NOVEL STOCHASTIC AND MICROSCOPIC ROAD TRAFFIC NOISE MODEL. UNLIKE CONVENTIONAL MACROSCOPIC MODELS THAT TREAT TRAFFIC AS A HOMOGENEOUS LINE SOURCE BY MEANS OF AGGREGATED TRAFFIC INPUTS, THE PROPOSED FRAMEWORK SIMULATES THE KINEMATICS OF INDIVIDUAL VEHICLES AS DISCRETE POINT SOURCES. SPECIFICALLY, BY RANDOMLY ASSIGNING VEHICLE SPEEDS FROM PROBABILITY DISTRIBUTIONS, THE MODEL SUCCESSFULLY PRESERVES THE INHERENT VARIABILITY AND TRANSIENT FLUCTUATIONS OF REAL-WORLD TRAFFIC. METHODOLOGICAL REFINEMENTS EXPLICITLY TARGET THE SPATIO-TEMPORAL DIMENSION: THE SPATIAL SCALE IS REFINED THROUGH LANE-SPECIFIC DISAGGREGATION TO ENHANCE NEAR-FIELD ACCURACY, WHILE THE TEMPORAL EVOLUTION IS CAPTURED VIA THE INTEGRATION OF SEASONAL AUTOREGRESSIVE INTEGRATED MOVING AVERAGE (SARIMA) TIME-SERIES FORECASTING TO PREDICT NOISE LEVELS IN DATA-SCARCE SCENARIOS. RIGOROUS VALIDATION AGAINST A LONG-TERM MONITORING DATASET CONFIRMS THAT EACH OF THESE COMPONENTS, I.E., THE STOCHASTIC CORE, THE SPATIAL DISAGGREGATION, AND THE TEMPORAL FORECASTING, SIGNIFICANTLY IMPROVES PREDICTIVE ACCURACY, CONSISTENTLY REDUCING SYSTEMATIC ERRORS COMPARED TO ESTABLISHED MACROSCOPIC BENCHMARKS. THE SECOND PILLAR SHIFTS THE ANALYTICAL LENS FROM A PHYSICAL APPROACH TO HUMAN PERCEPTION BY ADOPTING THE SOUNDSCAPE APPROACH. IN ACCORDANCE WITH THE ISO 12913 STANDARD SERIES, THE STUDY EMPLOYS A MIXED-METHODS METHODOLOGY, INTEGRATING STRUCTURED SOUNDWALKS AND DATA FROM LARGE-SCALE CROWDSOURCING CAMPAIGNS, TO CAPTURE THE MULTIDIMENSIONAL NATURE OF SUBJECTIVE EXPERIENCE WITHIN UNIVERSITY CAMPUS AREAS. USING THE FISCIANO CAMPUS OF THE UNIVERSITY OF SALERNO AS A CASE STUDY, THE RESEARCH INTRODUCES A GEOSTATISTICAL METHOD USING INVERSE DISTANCE WEIGHTING (IDW) INTERPOLATION TO GENERATE HIGH-RESOLUTION PLEASANTNESS MAPS. THIS APPROACH EXPLICITLY ADDRESSES BOTH SPATIAL AND TEMPORAL DIMENSIONS: THE SPATIAL ANALYSIS IS GROUNDED IN THE CORRELATION BETWEEN PERCEPTUAL DATA AND A DETAILED MAP OF FUNCTIONAL ZONES, WHILE THE TEMPORAL DIMENSION IS INVESTIGATED THROUGH THE COMPARATIVE ANALYSIS OF DATASETS COLLECTED ACROSS DIFFERENT TIMEFRAMES. THESE MAPS, ALONGSIDE THE TEMPORAL FINDINGS, OFFER AN INTUITIVE TOOL FOR URBAN PLANNERS, REVEALING HOW SPATIAL MORPHOLOGY, FUNCTIONAL LAND USE, AND TEMPORAL VARIATIONS MODULATE THE SUBJECTIVE EXPERIENCE OF THE ACOUSTIC ENVIRONMENT. ULTIMATELY, THIS DISSERTATION CONTRIBUTES A MORE NUANCED FRAMEWORK FOR ENVIRONMENTAL NOISE ASSESSMENT BY PROVIDING PARALLEL ADVANCEMENTS IN BOTH PHYSICAL MODELLING AND PERCEPTUAL APPRAISAL. RATHER THAN SEEKING AN IMMEDIATE SYNTHESIS, THE DEVELOPED TOOLS REPRESENT ESSENTIAL BASES THAT FACILITATE THE FUTURE CONVERGENCE OF OBJECTIVE MEASUREMENT AND SUBJECTIVE EXPERIENCE. BY STRENGTHENING THESE DUAL PERSPECTIVES, THIS RESEARCH PROVIDES A ROBUST FOUNDATION FOR PROACTIVE URBAN PLANNING AND ACOUSTIC DESIGN, FOSTERING HEALTHIER AND SONICALLY RICHER URBAN ECOSYSTEMS.