TERRAIN SUSCEPTIBILITY TO TRIGGER SHALLOW PLANAR LANDSLIDE ON STEEP SLOPES: a geo-hydroecological approach at D'Antas creek basin, Nova Friburgo (RJ - Brazil).

Several methodologies are described in the literature for the classification of terrain susceptibility to landslides, including probabilistic analysis based on landslide inventories and diagnoses based on the crossing of thematic maps. The latter approach prevails in Brazil, where thematic information has privileged geological maps and some geotechnical and geomorphological information. Other relevant categories, such as vegetation and land use are scant. This study poses a functionally based methodology of pertinent geological-geotechnical, geomorphological, vegetation cover and land use categories to assess terrain susceptibility to shallow planar landslides by following the geo-hydroecological approach. The D’Antas Creek basin (53 km2) located in Nova Friburgo, Rio de Janeiro state (SE-Brazil), was chosen as a pilot-area, as it was severely affected by 327 landslides in January 2011, 85% of which were triggered as shallow planar landslides with the slip surface around 1,5m + 0.5 m/deep. The geo-hydroecological approach follows empirical-analytical and integrative procedures of morphological, functional and historical knowledge, in a systemic view of the geographic space and articulating different hierarchical levels of the geo-ecosystem, as summarized in Coelho Netto et al. (2020). This approach is based on relevant functional parameters, indexes and categories affecting shallow planar landslides, all of which were synthetized in three cartographic bases: hydro-geomorphological; geological-geotechnical and vegetation-land use. Human-made activities were used to ponder their interferences to increase or to reduce the terrain susceptibility. The use of GIS (Geographic Information System) and multi-criteria analysis (i.e. Analytic Hierarchy Process - AHP) were incorporated into the geo-hydroecological approach. Soil geotechnical parameters increased the ability of the susceptibility model to identify potential areas for landslides. Integrated hydro-geomorphological conditions presented a more significant impact with the susceptibility classes. Vegetation and land use cover were reclassified on the basis of their ecological, hydrological and mechanical functionality, allowing a simplification of its classes. The final landslide susceptibility map included four categories (very high; high; medium; and low); areas classified as very high and high susceptibility overlaps with > 75% of the 2011 landslide scars. These results highlight the relevance of the geo-hydroecological approach to the definition of terrain susceptibility. Its systemic foundation and the organization of the database in a GIS environment remains open for updates, as new functional parameters are identified as relevant, favouring updated diagnoses to guide territorial ordering aimed at prevention, mitigation and/or adaptation of human occupation. This approach reaffirms the idea that the assessment of terrain susceptibility is a continuous process that must be constantly updated, according to the dynamics of landscape changes over space and time.