Methodologies to estimate economic flood damages\ud are increasingly important for flood risk assessment and\ud management. In this work, we present a new synthetic flood\ud damage model based on a component-by-component analysis\ud of physical damage to buildings. The damage functions\ud are designed using an expert-based approach with the support\ud of existing scientific and technical literature, loss adjustment\ud studies, and damage surveys carried out for past flood events\ud in Italy. The model structure is designed to be transparent\ud and flexible, and therefore it can be applied in different geographical\ud contexts and adapted to the actual knowledge of\ud hazard and vulnerability variables.\ud The model has been tested in a recent flood event in northern\ud Italy. Validation results provided good estimates of postevent\ud damages, with similar or superior performances when\ud compared with other damage models available in the literature.\ud In addition, a local sensitivity analysis was performed\ud in order to identify the hazard variables that have more influence\ud on damage assessment results

Abstract. Flood loss modelling is a crucial part of risk assessments. However, it is subject to large uncertainty that is often neglected. Most models available in the literature are deterministic, providing only single point estimates of flood loss, and large disparities tend to exist among them. Adopting any one such model in a risk assessment context is likely to lead to inaccurate loss estimates and sub-optimal decision-making. In this paper, we propose the use of multi-model ensembles to address these issues. This approach, which has been applied successfully in other scientific fields, is based on the combination of different model outputs with the aim of improving the skill and usefulness of predictions. We first propose a model rating framework to support ensemble construction, based on a probability tree of model properties, which establishes relative degrees of belief between candidate models. Using 20 flood loss models in two test cases, we then construct numerous multi-model ensembles, based both on the rating framework and on a stochastic method, differing in terms of participating members, ensemble size and model weights. We evaluate the performance of ensemble means, as well as their probabilistic skill and reliability. Our results demonstrate that well-designed multi-model ensembles represent a pragmatic approach to consistently obtain more accurate flood loss estimates and reliable probability distributions of model uncertainty.

Abstract. Methodologies to estimate economic flood damages are increasingly important for flood risk assessment and management. In this work, we present a new synthetic flood damage model based on a component-by-component analysis of physical damage to buildings. The damage functions are designed using an expert-based approach with the support of existing scientific and technical literature, and have been calibrated with loss adjustment studies and damage surveys carried out for past flood events in Italy. The model structure is designed to be transparent and flexible, and therefore it can be applied in different geographical contexts and adapted to the actual knowledge of hazard and vulnerability variables. The model has been tested in a recent flood event in Northern Italy. Validation results provided good estimates of post-event damages, with better performances than most damage models available in the literature. In addition, a local sensitivity analysis has been performed, in order to identify the hazard variables that have more influence on damage assessment results.

Abstract.One of the necessary components to perform catastrophe risk modelling is information on the buildings at risk, such as their spatial location, geometry, height, occupancy type and other characteristics. This is commonly referred to as the exposure model or data set. When modelling large areas, developing exposure data sets with the relevant information about every individual building is not practicable. Thus, census data at coarse spatial resolutions are often used as the starting point for the creation of such data sets, after which disaggregation to finer resolutions is carried out using different methods, based on proxies such as the population distribution. While these methods can produce acceptable results, they cannot be considered ideal. Nowadays, the availability of open data is increasing and it is possible to obtain information about buildings for some regions. Although this type of information is usually limited and, therefore, insufficient to generate an exposure data set, it can still be very useful in its elaboration. In this paper, we focus on how open building data can be used to develop a gridded exposure model by disaggregating existing census data at coarser resolutions. Furthermore, we analyse how the selection of the level of spatial resolution can impact the accuracy and precision of the model, and compare the results in terms of affected residential building areas, due to a flood event, between different models.