A landslide risk index for municipal land use planning in Portugal

In Portugal landslides caused 237 fatalities and >1600 displaced people in the period 1865-2015. Spatial distribution and temporal patterns of slope instability can be related with a complex set of natural and human factors responsible for generating damages. It is essential to develop new methodologies to synthetize risk dimensions to contribute to the landslide risk management at the municipal level. This work proposed a municipal landslide risk index (LRI) considering three risk dimensions: hazard, exposure and physical vulnerability of buildings. The hazard dimension includes the landslide susceptibility performed at the national scale, the probability of weather types associated with landslides and an extreme precipitation susceptibility index. The exposure dimension considered the population density and the number of buildings, whereas the average physical vulnerability of the buildings was computed using four statistical variables from the official census: (i) construction technique and construction materials; (ii) reinforced structure; (iii) number of floors; and (iv) conservation status. Each variable includes different classes that were empirically weighted. After evaluating the three risk dimensions and the LRI, a cluster analysis was performed in order to identify the most important landslide risk drivers in each municipality. Exposure is the main driving force of LRI in the metropolitan areas of Lisbon and Porto, while the hazard is more relevant in the NW municipalities and the physical vulnerability is the major driving force in the south of the country. This methodological approach contributes to a comprehensive and synthetized knowledge about the landslide risk driving forces within the 278 Portuguese municipalities. In addition, it contributes to the diversification and context-oriented strategies of landslide risk management that still lacks in most of the national-level risk governance processes. Finally, this methodology can be generalized to other geographical contexts, improving the risk management, land use planning and the disaster risk reduction.

Defining evacuation travel times and safety areas in a debris flow hazard scenario

Debris flows are one of the most hazardous types of landslides in mountain regions. In the upper part of the Zêzere valley (Serra da Estrela, Portugal) several debris flows events occurred in the last 200 years, some of them causing loss of lives and material damages. In this work, a methodology for pedestrian evacuation modelling, in a debris flow hazard scenario, was implemented. A dynamic run-out model, developed in previous studies, was used to evaluate the debris flows velocities, thickness of the deposits and extent of the mobilized material. The buildings potentially affected by the impact of debris flows were identified and the potentially exposed population was estimated by applying a dasymetric distribution. The results lead to the conclusion that, in the study area, the elderly are those who are most exposed to debris flows. Furthermore, the time lapse between the debris flows initiation and the arrival at the buildings at risk was estimated, allowing to account for the overall number of buildings where the evacuation time takes longer than the debris flows arrival. Additionally, the safe areas within the study area were identified, and several safe public buildings with the capacity to gather a large number of persons were selected. Considering that the study area is located in a mountain region, characterized by steep slopes, the evacuation modelling was performed based on an anisotropic approach, in order to consider the influence of slope direction on travel costs. At the end, three pedestrian evacuation travel time scenarios, based on different walking speeds to accommodate residents with different ages in safer places, were compared and the results mapped. The implemented methodology is not local dependent, which allows its reproduction elsewhere.

Mapping landslide susceptibility using data-driven methods

Most epistemic uncertainty within data-driven landslide susceptibility assessment results from errors in landslide inventories, difficulty in identifying and mapping landslide causes and decisions related with the modelling procedure. In this work we evaluate and discuss differences observed on landslide susceptibility maps resulting from: (i) the selection of the statistical method; (ii) the selection of the terrain mapping unit; and (iii) the selection of the feature type to represent landslides in the model (polygon versus point). The work is performed in a single study area (Silveira Basin - 18.2km² - Lisbon Region, Portugal) using a unique database of geo-environmental landslide predisposing factors and an inventory of 82 shallow translational slides. The logistic regression, the discriminant analysis and two versions of the information value were used and we conclude that multivariate statistical methods perform better when computed over heterogeneous terrain units and should be selected to assess landslide susceptibility based on slope terrain units, geo-hydrological terrain units or census terrain units. However, evidence was found that the chosen terrain mapping unit can produce greater differences on final susceptibility results than those resulting from the chosen statistical method for modelling. The landslide susceptibility should be assessed over grid cell terrain units whenever the spatial accuracy of landslide inventory is good. In addition, a single point per landslide proved to be efficient to generate accurate landslide susceptibility maps, providing the landslides are of small size, thus minimizing the possible existence of heterogeneities of predisposing factors within the landslide boundary. Although during last years the ROC curves have been preferred to evaluate the susceptibility model's performance, evidence was found that the model with the highest AUC ROC is not necessarily the best landslide susceptibility model, namely when terrain mapping units are heterogeneous in size and reduced in number.