Work Package 3
Work Package Name: Climate Data, CC Forcing, Multi-Hazard Modelling
Lead Beneficiary: FMI
Work Package 3 (WP3) of the RISKADAPT project focuses on understanding how climate change can affect critical infrastructure like bridges, hospitals energy systems and tall buildings. This involves collecting and analyzing climate data—such as rainfall, wind, and temperature—to identify extreme weather patterns both now and in the future. WP3 also looks at how these weather events could lead to floods or strong winds that might damage buildings and structures. By using advanced modelling and simulations, the project helps predict where and how such risks might occur. The results from WP3 are essential for helping engineers, planners and decision-makers take informed action to protect infrastructure and communities from the impacts of climate change.
Deliverables
D3.1 – Access to EO data
D3.2 – Extreme value distributions for Europe – present climate
D3.3 – Extreme value distributions-future climate
D3.4 – Hydrologic/hydraulic modelling regarding floods
D3.5 – Model of high wind loading on a high rise building
D3.6 – Algorithms for glass window damage
D3.1 – Access to EO data
This deliverable describes the Earth Observation (EO) data sources used in the RISKADAPT project. The most important data source is the Copernicus Climate Store. The data types include in-situ, remote sensed, and model-based data. Model-based data can be divided between historical reanalyses based on observations and climate model runs that simulate the climate system’s response to a scenario for future emissions. Local observational data from national data providers can be used to complement the Copernicus data. This deliverable will give an overview of the EO data directly available for the partners. The use of open data sources through API interfaces is encouraged and demonstrated.
Read the full deliverable here.
D3.2 – Extreme value distributions for Europe – present climate
RISKADAPT will provide, in close cooperation with the end-users/other stakeholders, a novel, integrated, modular, interoperable, public and free, customisable user-friendly platform (PRISKADAPT), to support systemic, risk-informed decisions regarding adaptation to Climate Change (CC) induced compound events at the asset level, focusing on the structural system. PRISKADAPT will explicitly model dependencies between infrastructures, which, inter alia, will provide a better understanding of the nexus between climate hazards and social vulnerabilities and resilience.
Moreover, this project will identify gaps in data and propose ways to overcome them and advance the state of the art of asset level modelling through advanced climate science to predict CC forcing on the structure of interest, structural analyses, customised to the specific structure of interest, that consider all major CC induced load effects in tandem with material deterioration, novel probabilistic environmental Life Cycle Assessment (LCA) and Life Cycle Cost (LCC) of structural adaptation measures and a new model to assess climate risk that will combine technical risk assessment with assessment of social risks. PRISKADAPT will provide values to a set of indicators for each asset of interest, quantifying primary parameters and impacts, in the form of a Model Information System (MIS) that will provide all required information for adaptation decisions. PRISKADAPT will be implemented in the case studies in the pilots that involve specific assets, however, it will permit customisation with local values of parameters and data, so it can be applicable throughout Europe for CC adaptation decisions involving assets of similar function, exposed to multiple climate hazards.
This report is one of the three deliverables of WP3 “Climate Data, CC Forcing, Multi-Hazard Modelling” and corresponds T3.1 “Climate data for hydrological analyses, wind and rain forcing and material degradation. Extreme Value Analyses” of the RISKADAPT project. To meet the aim of this Task, in this report the terminology, methods, and tools that can be used to perform statistical extreme value analysis on climate data are presented. Input data can mainly come from reanalysis data (like ERA5) or from in-situ time series measurements. In addition, and as the focus of this report is on the analysis of present climate, i.e., approximate for years 1970-2020, pointers to Copernicus Climate Data Store (CDS) data sets and tools are provided that can be used to evaluate risks caused by extreme events for infrastructure, especially those that are related to project’s pilots.
This report provides methodological guidance on the application of Bayesian hierarchical modelling techniques to improve the reliability of extreme value analysis by incorporating spatial and temporal dependencies. By quantifying return levels and return periods of extremes of key climate variables, these statistical tools provide essential methods in studying climate change induced risks for infrastructure. We evaluate climate data sources relevant to extreme weather and hydrological events, including precipitation, wind speed, and temperature extremes. The report demonstrates statistical techniques using real-world datasets, with a particular focus on case studies relevant to RISKADAPT pilot sites. This deliverable also includes practical examples of extreme value analysis using Python-based computational tools, enabling application in different infrastructure contexts. The extreme value analysis presented here is relevant for climate scientists, engineers, policymakers, and infrastructure managers involved in climate adaptation planning.
However, while reanalysis data sets presented here, such as ERA5 Land, provide valuable insights, limitations in spatial and temporal resolution necessitate careful methodological considerations, including downscaling approaches, which will be dealt in more detail in other deliverables of the RISKADAPT project.
Read the full deliverable here.
D3.3 – Extreme value distributions-future climate
This deliverable will complement the deliverable D3.2. Continuing explaining the methods and tools for statistical extreme value analysis for climate variables, this report focuses on analysis of the future climate, approximately for years 2040-2100. The input data are produced either by climate model simulations or by historical reanalyses. We provide pointers to Copernicus CDS data sets and tools that can be used to evaluate risks caused by extreme events for infrastructure, especially those that are related to RISKADAPT project’s pilots. The deliverable is public and the intended audience includes project partners and anyone interested in extreme value analysis.
Read the full deliverable here.
D3.4 – Hydrologic/hydraulic modelling regarding floods
The primary objective of Deliverable D3.4 is to outline the hydrological and hydraulic modelling processes required to estimate hydrodynamic loads on the piers and abutments of any bridge, with a special focus on the Polyfytos Bridge located at the Polyfytos Lake, Greece (namely the Pilot 1 of the RISKADAPT project). This includes obtaining all the necessary input data (i.e., topography, land use, hydrologic soil group, extreme precipitation data, and lake bathymetry), as well as setting up the hydrological model (to convert extreme precipitation into extreme discharges) and the hydraulic model (to propagate discharges along the river/reservoir and calculate hydrodynamic loads on bridge piers under present and future climate projections).
The Aliakmon River catchment runoff (feeding the Polyfytos Lake) was estimated using the 2D hydrological model based on extreme precipitation and catchment characteristics (i.e., topography, land use and soil type). Extreme events were investigated by considering three return periods of extreme rainfall (i.e., 50, 100 and 1.000 years). Simulations of the Aliakmon River catchment model for the present climate determined that a 72-hour rainfall duration resulted in the highest discharge peaks, with maximum discharges at the inflow to the Polyfytos Lake being 2.400, 3.100 and 6.350 m³/s for the respective return periods. Future climate scenarios showed varying impacts on peak discharges, with some predicting reductions and others increases.
Hydraulic modelling assessed hydrodynamic loads and scour risks on the piers of the Polyfytos road bridge, which spans over the Polyfytos Lake reservoir. The study incorporated geometric and hydrological data to simulate unsteady flow conditions and evaluate the impact on scouring processes around bridge piers using a full-2D numerical hydraulic model. The hydraulic modelling results provided essential time series data for further analysis, including water depths, water surface elevations, depth-averaged flow velocities, and discharges at specified locations. Despite the lack of detailed bathymetry and sediment composition data, the analysis indicated that water flow velocities at the bridge piers remain low during high-water events, suggesting a low risk of scour formation.
Thus, the results presented in this deliverable provide the necessary input information for Task 4.2, dealing with material degradation and structural vulnerability of structures exposed to extreme events. Additionally, the deliverable contributes to Milestone 6, entitled “Assessment of hydrodynamic loads on piers/abutments of bridges.”
Deliverable D3.4 is meant primarily for other partners of the RISKADAPT project, mainly to support the work of Task T4.2, but also for other practitioners who would like to estimate the impact of hydrodynamic loads and scour risks on the bridge piers to learn about the input data and the models needed to support this action, especially in the data-scarce catchments where global data (e.g., precipitation) has to be used.
Read the full deliverable here.
D3.5 – Model of high wind loading on a high rise building
This report showcases the development and application of a general methodology to estimate the atmospheric load on tall buildings in case of extreme weather events, that shall become more frequent and intense in Europe due to climate change. In particular, the methodology herein presented enables the derivation of a semi-empirical function for an easy, fast and direct estimation of the atmospheric load on tall buildings. Such a methodology is applied to the case study of Cattinara Public Hospital of Trieste (North-East Italy) that is the Pilot 3 of the RISKADAPT project. Such a case study has been selected because of two key features: the region of Trieste is characterised by the Bora wind, which is a high-intensity wind that periodically impacts the city; and the Cattinara Hospital is a relatively tall building located on the top of a hill just outside the city, thus highly exposed to the Bora wind. For these reasons, Pilot 3 is a unique and ideal site for the purposes of the present study.
The strong wind and wind-driven rain are the two meteorological variables considered to estimate the atmospheric load on buildings which, in turn, is quantified as pressure load at the building facades. The facade pressures load is the key variable used to assess structural vulnerability of buildings and infrastructures. The present study is carried out through a chain of multiscale numerical simulations, which are strategically linked together to transport and convey information from the climatic timespatial scales, the meteorological mesoscale and the very small, local, building scale in the final result.
To this scope, a downscaling methodology has been adopted and applied for a detailed numerical study of the Cattinara Hospital, which was selected as representative of a tall building in Europe exposed to strong wind, i.e. the Bora wind, due to the peculiar local meteorology. First, the past and future climate in the Trieste region has been studied, scrutinising state-of-the art climatic databases (e.g., EURO-CORDEX and ERA5) which are the results of the most advanced climatic simulations for the European continent (scale 100 km). Second, to further ground the study in the real-world settings, a specific extreme event of strong Bora wind (registered in February 2012) has been selected and reproduced numerically utilising the Weather Forecast and Research (WRF) model, a state-of-the-art meteorological model. Such a simulations reproduce the weather conditions at the regional scale (spatial scale of 1 km) and allow to analyse in detail the meteorological processes and dynamics, providing accurate and realistic atmospheric variables in the neighbourhood of the Cattinara Hospital.
Third, the output from such meteorological simulation is used to setup very-highly resolved numerical simulations that reproduce the wind flow in the building area (spatial scale of 10 m) to gain insight of the local circulation around the buildings. These last simulations have been carried out using highly accurate techniques and turbulence models developed in the field of Computational Fluid Dynamics and using the open-source software OpenFOAM. The result of this numerical downscaling procedure made it possible to investigate the climatology and meteorology of the characterising the Trieste area at different scales, as well as to evaluate the atmospheric load on the Cattinara Hospital in the most realistic way for a typical extreme of wind.
Based on this study, a semi-empirical function for estimating pressure load on buildings is derived as a practical tool for technical and non-technical stakeholders. An extensive parametric study has been performed running several very-high resolved numerical simulations of the Cattinara Hospital case study by estimating the pressure load at the building facades under different wind intensities (from weak to exceptionally strong) and. The results have been synthetised in a set of functions estimating the atmospheric building load knowing the averaged mean wind velocity impacting on the structure.
The functions have been extended to include also the contribution to the load given by the wind-driven rain, by utilising well-established relations available in the literature. Overall, the estimating functions can be directly used in the configuration of single building on a gently slope terrain, as that one of Cattinara Hospital.
It is worth to notice that a universal function estimating the pressure load from the averaged wind speed can be hardly derived, due to the complex interaction between the terrain topography and themorphological elements (other tall buildings in the neighbourhood) that are specific features of each case study. However, the methodology developed and applied in the present study is a general and operational strategy that can be repeated and adapted to a wide variety of case studies, therefore representing a general result of applicative interest for infrastructure builders and for different types of technical and non-technical stakeholders.
Read the full deliverable here.
D3.6 – Algorithms for glass window damage
This deliverable presents a comprehensive numerical investigation into window glass damage on tall buildings subjected to extreme wind conditions, with a focus on developing a general algorithm to estimate and mitigate such damage in urban environments. The study is rooted in real-world data from Hong Kong, specifically analyzed under the RISKADAPT project’s Pilot 4, where the city’s high-rise structures frequently experience window glass shattering during typhoons, such as the notable Typhoon Mangkhut. By combining computational fluid dynamic, this work proposes a methodological framework to predict and prevent glass curtain wall failures.
The primary objective of this deliverable is to assess the vulnerability of window glass on tall buildings to strong winds and to devise a replicable procedure for estimating damage risks across urban settings. This is achieved by studying wind impacts on three high-rise neighborhoods in Hong Kong, where window shattering has been documented across multiple typhoon events. The scope encompasses both understanding the physical mechanisms behind glass failure—particularly the role of local wind acceleration—and providing a predictive tool to inform building design and urban planning.
This deliverable benefits the RISKADAPT project by providing a wind breakage probability model to enhance climate resilience in urban infrastructure. It directly supports the project’s goal of adapting buildings to extreme weather. The model can help to understand and mitigate the risks posed by extreme weather events, particularly typhoons and high-wind storms, which are becoming more frequent and intense due to climate change. The target audience for the wind breakage probability model encompasses a diverse group of professionals, including structural engineers, urban planners, building designers, and policymakers, with a particular emphasis on those working in typhoon-prone regions. For example, urban planners working on developing new residential or commercial districts in coastal areas—where typhoons or hurricanes occur frequently—can leverage the model to make informed decisions about site selection, building placement, and infrastructure design. By understanding the probability of window and glass curtain wall failures under specific wind pressures, planners can minimize risks to human safety, reduce economic losses, and ensure long-term sustainability.
Key findings highlight the critical role of local wind acceleration in amplifying pressure loads on windows, often exceeding design thresholds during typhoons. The LES results confirm the reasons of window damages. The proposed algorithm offers a practical guideline for estimating glass damage risk, incorporating wind pressure and breakage probability into a cohesive framework. Lessons learned emphasize the need for site-specific wind analysis in urban settings and the potential of computational modelling to bridge gaps between theoretical design and real-world performance.
Read the full deliverable here.