The City of Ottawa is situated in an area known as the Champlain Sea, 17,000 years before present (BP) the entire area was covered with sea water. This area deposited marine clays which are known to be highly sensitive. The City of Ottawa needs to expand land use to allow for the expansion of infrastructure and housing to support its growth. This study is intended to assist the City of Ottawa’s geotechnical engineers in their decision-making by identifying future sensitive areas prone to landslides due to rainfall based on future climate model data. The project incorporates rainfall intensities from downscaled climate model data in the Transient Rainfall Infiltration and Grid-based Regional Slope-Stability (TRIGRS) model to investigate areas sensitive to landslides, then within a GIS platform, the future landslide susceptibility maps were created based on Factor of Safety (FS) values showing the areas prone to landslides. The data input for the model includes climate model data, topography, hydrogeology, geology and geophysical data obtained from a previous study. These data were prepared using ArcGIS software and converted into ascii format for TRIGRS model. The model was calibrated using historical rainfall intensities and validated by comparing to historical landslide areas. Sensitivity analysis were performed to ranges of geotechnical properties found within sensitive marine clays in the area to find the values best to create the ideal scenario, normal scenario and worst-case model scenario for the prediction. Rainfall intensities from projected climate data Intensities Duration Frequency (IDF) of 10 years and 50 years returning period and rainfall intensities of 12 hr, 24 hr, and 48 hr were selected for the model. Results from simulations find the projected climate rainfall intensity do not have impact or has minimal impact to slope stability in sensitive marine clay areas in Ottawa directly. However, higher rainfall runoff is expected from projected rainfall RCP8.5 than the RCP4.5. The infiltration rate remains constant throughout each simulation, which is the same value as the hydraulic conductivity. The time when the slope becomes unstable varies depending on initial water levels. Results from the ideal and normal scenario show no areas prone to slope failure after 48 hours of rainfall duration. However, the factor of safety decreases as the rainfall duration increases and is expected to decrease with longer rainfall durations. The worst-case scenario shows some areas prone to slope failure (FS < 1) with 2% probability of slope failure at 48 hours of rainfall duration. The distribution of these unstable areas are located along the Ottawa River, Rideau River, Carp River, Mississippi River and valleys along their tributaries, the majority of the area prone to slope instability from rainfall are in the east part of the City of Ottawa. While there are many uncertainties and limitations which contribute to the model results, this study is useful to engineers and planners in initial implementation of mitigation strategies to mitigate the damages and cost from landslides events. The susceptibility maps can also assist in decision making for planners in developing into these areas.

L'effet de la végétation locale, composée de Castanea sativa cultivé, sur la stabilité des pentes a été étudié sur un site d'essai au Mont Faito (Campanie, Italie). En Campanie, les sols pyroclastiques peu profonds sont sensibles aux glissements de terrain provoqués par les précipitations. Des périodes de pluies prolongées suivies de précipitations extrêmes à court terme déclenchent des glissements de terrain rapides et destructeurs au niveau des coupes routières et des escarpements pyroclastiques sur les falaises rocheuses dans les régions autour du volcan Vésuve.Des échantillons de sol pyroclastiques non perturbés contenant des racines de C. sativa matures ont été utilisés pour la caractérisation hydraulique par le biais d'un ensemble d'expériences en laboratoire. La perméabilité saturée, la réponse à l’évaporation et l’imbibition, la teneur en eau pour les fortes valeurs de succion et la biomasse sèche des racines ont été déterminées.La présence de racines a augmenté la perméabilité d'un ordre de grandeur dans les sols les plus superficiels (10-7 à 10-6 m s-1) et diminué la valeur d'entrée d'air des courbes de rétention (6 à 4 kPa). La variabilité de la perméabilité entre les couches de sol a été identifiée comme conditionnant l'écoulement de l'eau souterraine par rapport à la vitesse du mouvement du front de mouillage et à la génération de pressions positives de l'eau interstitielle dans le profil. L'étalonnage du modèle hystérétique pour caractériser les sols pyroclastiques naturels a fourni une méthode plus approximative de modélisation des réponses hydrauliques. Une bonne concordance entre le modèle et les observations a été obtenue.L’étude sur le terrain a permis de montrer que la distribution des racines de C. sativa est associée au régime des eaux souterraines. Les distributions spatiales et verticales de la densité et des traits des racines ont été quantifiées pour les racines de C. sativa prélevées dans des forages réalisés au Mont Faito. La succion minimale, la teneur minimale en eau et la pente minimale (indiquant un débit d'eau descendant) ont été surveillées tout au long de l'année et confrontées avec la distribution des racines et à la distribution spatiale des arbres. Une densité racinaire croissante était associée à des valeurs de succion plus faibles et à des gradients d'infiltration plus élevés, ce qui peut avoir une influence négative sur la stabilité de la pente.La modélisation du renforcement mécanique du sol par les racines des arbres a permis de comprendre l'importance des composantes hydrauliques et mécaniques sur la stabilité d'une pente. Les racines augmentent la résistance au cisaillement (jusqu'à 25,8 kPa) grâce à un renforcement mécanique et donc le facteur de sécurité de la pente augmente. L'examen du renforcement dû aux racines dans l'estimation du facteur de sécurité des surfaces de rupture potentielles a montré que la surface de rupture la plus faible a été trouvée à 2,2 m, où le renforcement dû aux racines était de 1,3 kPa, au lieu de 0,9 m sans le renforcement de 13,8 kPa. La surface de rupture la plus faible correspond aux surfaces de rupture observées lors de glissements de terrain antérieurs. Le site d'essai ne présentait pas les caractéristiques d'une zone de déclenchement d'un glissement de terrain. L'angle de pente des zones de déclenchement des glissements de terrain (35° à 45°) peut dépasser l'angle de frottement du sol (36,5° à 38,5°) et l'effet hydraulique ne serait pas suffisant pour garantir la stabilité de la pente pendant la saison humide (0 à 10 kPa). On estime que le renforcement dû aux racines peut maintenir les pentes jusqu'à un angle de 42°.On a donc constaté que la présence de racines d'arbres affectait la stabilité hydraulique et mécanique des couvertures de sol pyroclastiques. Ces conclusions peuvent être étendues aux autres zones de plantations de C. sativa. L'effet hydraulique de la végétation a été largement compensé par le renforcement mécanique dû aux racines / The effect of the local vegetation, composed of cultivated Castanea sativa, on slope stability was investigated on a test site in Mount Faito (Campania, Southern Italy). In Campania, shallow pyroclastic soil covers are susceptible to landslides triggered by rainfall. Prolonged rainfall periods followed by extreme short-term rainfall events trigger fast moving and highly destructive landslides in road cuts and pyroclastic scarps on rocky cliffs in the areas surrounding the Vesuvius volcano.Undisturbed pyroclastic soil samples containing roots of mature C. sativa were used for hydraulic characterization through an extensive set of laboratory experiments. Saturated permeability, evaporation and imbibition response, water content for high suction ranges, and the root dry biomass were determined.The presence of roots increased the hydraulic permeability by one order of magnitude in the most surficial soil (10-7 to 10-6 m s-1) and decreased the air-entry value of the water retention curves (6 to 4 kPa). The variability of soil permeability among soil layers was identified as conditioning of the groundwater flow with regard to the speed of the wetting front movement and generation of positive pore-water pressures within the soil profile. The calibration of hysteretic model to characterize natural pyroclastic soil provided a more approximate manner of modelling in situ hydraulic responses. A good agreement between the model and the field observations was obtained.Field monitoring was performed with the intent of showing that the distribution of roots of C. sativa is associated to the groundwater regime. The spatial and vertical distribution of root density and traits were quantified for C. sativa roots collected from several boreholes performed in Mount Faito. Minimum suction, minimum water content and minimum gradient (indicative of downward water flow), were monitored throughout the year and related to root distribution and spatial distribution of trees. An increasing root density was found to be associated to lower values of suction and higher gradients of infiltration, which can potentially have a negative influence of the slope stability.A modelling investigation on the mechanical reinforcement of soil by tree roots allowed us to understand the importance of hydraulic and mechanic components on the stability of a slope. Roots increase greatly the shear strength of soil (up to 25.8 kPa) through mechanical reinforcement and consequently, the safety factor of the slope increased significantly. Considering the root reinforcement in the estimation of potential failure surfaces safety factor showed that the weakest failure surface was found at 2.2 m, where the root reinforcement was 1.3 kPa, instead of 0.9 m without the root reinforcement of 13.8 kPa. The weakest failure surface found was in agreement with the failure surfaces observed from previous landslides. The test site did not present the characteristics of a landslide triggering area. The slope angle of the landslide triggering areas (35° to 45°) can easily exceed the soil friction angle (36.5° to 38.5°) and the hydraulic effect would not be enough to guarantee the stability of the slope during the wet season (0 to 10 kPa). However, the root reinforcement was estimated to be able to sustain the slopes until an angle of 42°.Therefore, the presence of tree roots was found to affect hydraulically and mechanically stability of pyroclastic soil covers. Such conclusions may be extended to the areas of Campania where C. sativa plantations are present. The hydraulic effect of vegetation was greatly compensated by the mechanical reinforcement of roots.