http://repositorio.lnec.pt:8080/jspui/handle/123456789/67 CICTI 2026-04-04T20:59:29Z http://repositorio.lnec.pt:8080/jspui/handle/123456789/1018799 Data analytics to advance the inference of origin–destination in public transport systems: tracing network vulnerabilities and age-sensitive trip purposes. Cerqueira, S.; Arsénio, E.; Barateiro, J.; Henriques, R. Knowing the passengers’ final destinations, underlying motifs, and commuting habits is critical to optimise public transportation systems, guide urban planning and contribute to a more sustainable urban mobility. In entry-only Automated Fare Collection systems, the body of literature has focused on the spatial dimension by estimating alighting stops, overlooking the inference of robust alighting times. Moreover, discriminating between transfers and activities is pivotal for determining their ultimate destinations. However, current methods often struggle to adapt to the stochastic nature of passenger behaviour, further disregarding the multiplicity of routes and stops to access specific facilities and individual motivations. Further research is required to address an effective spatio-temporal and contextual inference in both challenges. With the above concerns in mind, this research uses data analytics to propose an enhanced methodology for the inference of OD matrices, with the final goal of providing a comprehensive view of OD mobility patterns across distinct age-sensitive profiles—youth, adults, and older adults. Our methodological framework integrates the following approaches: (i) alighting stop-and-time inference, (ii) ensembled model for transfer classification, (iii) indicators retrieved from statistical analysis of network vulnerabilities (e.g., number of transfers, walkability needs), frequent destinations and their underlying putative motifs against the city amenities and others points-of-interest. The reliability of alighting data (timestamp and location) inference is improved by integrating OpenStreetMap data and the past boarding data from bus and railway systems. Considering Lisbon as the target study case, we apply the methodology over smart card data collected both from metro and bus systems. A comparative analysis with state-of-the-art methods revealed that the enhanced framework for alighting and OD inference led to longer journey times for trips. Furthermore, throughout the day, the older adult group experiences longer transfer times on average compared to both the children and young adult segment and the adult segment. 2025-05-22T00:00:00Z http://repositorio.lnec.pt:8080/jspui/handle/123456789/1018732 CONNECT – Local coastal monitoring service for Portugal Rodrigues, M.; Fortunato, A. B.; Martins, R.; Jesus, G.; Brito, A.; Oliveira, A.; Nahon, A.; Costa, J. L.; Alves, E.; Korani, Z.; Azevedo, A. LNEC. National and international Meetings CONNECT provides a local, high-resolution, coastal monitoring service. It integrates model-based forecasts and observations to provide physical and biogeochemical data on Portuguese estuaries. The service is demonstrated in two use cases, the Tagus estuary and the Mondego estuary. Coastal areas provide multiple ecosystem services and are key for the blue economy. These systems harbor ecologically important habitats for fish, shellfish and birds, act as buffers for nutrient and contaminant loads entering the ocean, and support diverse human activities. The increasing human activities in coastal areas, coupled with the impacts of climate change, are rising the hazards within these systems. As a result, quantifying and anticipating how human-induced or climate-driven factors influence coastal systems has become essential to support their sustainable management and to meet the objectives of environmental directives and policies. CONNECT offers a shelf-to-estuary-river, high-resolution, coastal monitoring service that integrates model-based forecasts and observations. It provides physical and biogeochemical data on Portuguese estuaries to the Copernicus Marine Service (CMEMS), enhancing the management, monitoring and forecasting of water quality and coastal inundation. This service integrates two operational coastal data infrastructures: • the UBEST coastal observatory (Rodrigues et al., 2021; https://ubest.lnec.pt) that focuses primarily on model forecasts partly forced by CMEMS regional models. • the CoastNet monitoring infrastructure (Castellanos et al. 2021; França et al. 2021; https://coastnet.pt) that provides near real-time observations from in-situ sensors, and remote sensing data mainly from CMEMS. By combining complementary sources of information, the integrated service: i) enhances confidence in both the model results and the observations by allowing automatic cross-comparison of data sources; ii) fosters the early detection of numerical and measurement anomalies (e.g., drifting time series due to biofouling); iii) trengthens the robustness by creating redundancy in the data access; iv) promotes the adoption of common formats and a seamless integration with CMEMS; v) simplifies the extension to other coastal systems, given the adaptability and transferability of the underlying service. The new integrated service strengthens the support of the implementation of the Water Framework Directive (WFD), the Marine Strategy Framework Directive (MSFD), the Floods Directive (FD), and other EU policies, such as the Green Deal, by providing further and integrated historical and real-time information on physics and biogeochemistry. This paper describes the main characteristics of the CONNECT coastal service (section 2) and its use cases, in particular the demonstration of the service in the Tagus and Mondego estuaries (section 3). Final remarks are presented in section 4. 2024-06-01T00:00:00Z http://repositorio.lnec.pt:8080/jspui/handle/123456789/1018502 Deep Learning-Based River Flow Forecasting with MLPs: Comparative Exploratory Analysis Applied to the Tejo and the Mondego Rivers Jesus, G.; Korani, Z.; Alves, E.; Oliveira, A. Yuh-Shyan Chen and Wei Yi Abstract: This paper presents an innovative service for river flow forecasting and its demonstration in two dam-controlled rivers in Portugal, Tejo, and Mondego rivers, based on using Multilayer Perceptron (MLP) models to predict and forecast river flow. The main goal is to create and improve AI models that operate as remote services, providing precise and timely river flow predictions for the next 3 days. This paper examines the use of MLP architectures to predict river discharge using comprehensive hydrological data from Portugal’s National Water Resources Information System (Sistema Nacional de Informação de Recursos Hídricos, SNIRH), demonstrated for the Tejo and Mondego river basins. The methodology is described in detail, including data preparation, model training, and forecasting processes, and provides a comparative study of the MLP model’s performance in both case studies. The analysis shows that MLP models attain acceptable accuracy in short-term river flow forecasts for the selected scenarios and datasets, adeptly reflecting discharge patterns and peak occurrences. These models seek to enhance water resources management and decision-making by amalgamating modern data-driven methodologies with established hydrological and meteorological data sources, facilitating better flood mitigation and sustainable water resource planning as well as accurate boundary conditions for downstream forecast systems. 2025-03-28T00:00:00Z http://repositorio.lnec.pt:8080/jspui/handle/123456789/1018416 CONNECT - Local coastal monitoring service for Portugal Oliveira, A.; Rodrigues, M.; Fortunato, A. B.; Martins, R.; Jesus, G. This report proposes two cross-cutting methodologies for joint exploitation of information for monitoring and modeling in coastal regions taking advantage of the products developed and enhanced in the scope of the CONNECT project, namely the CONNECT coastal service (Rodrigues et al., 2024). The workflow for their implementation is demonstrated in the Tagus estuary taking advantage of the results of this project. The proposed cases are: 1) a methodology for establishing forecast system grid limits based on the information provided by the in-situ and remote networks; and 2) a methodology to support monitoring infrastructures. The analyses are conducted in a generic way, to be applied anywhere, and are then illustrated using the model forecasts and the in-situ and remote sensing data available at the CoastNet monitoring network (Castellanos et al. 2021; França et al. 2021). The demonstration sites, initially selected to be the Mondego estuary and Ria Formosa, were switched to the Tagus estuary. Indeed, while both the Tagus and the Mondego estuaries have a similar in-situ network, their spatial scales are quite different (the Tagus estuary is much larger), and the application of the methodology using the available remote sensing data would be less effective in the Mondego. In the case of the Ria Formosa, in-situ data is not currently available for the relevant variables. Therefore, both methodologies will be demonstrated in the Tagus estuary. 2024-12-01T00:00:00Z