PAPER Coupled surface and subsurface flow modeling of natural hillslopes in the Aburrá Valley (Medellín, Colombia) Daniela Blessent 1 & Janet Barco 1,5 & André Guy Tranquille Temgoua 2,3 & Oscar Echeverrri-Ramirez 4 Received: 16 March 2016 /Accepted: 1 October 2016 # Springer-Verlag Berlin Heidelberg 2016 Abstract Numerical results are presented of surface- subsurface water modeling of a natural hillslope located in the Aburrá Valley, in the city of Medellín (Antioquia, Colombia). The integrated finite-element hydrogeological simulator HydroGeoSphere is used to conduct transient vari- ably saturated simulations. The objective is to analyze pore- water pressure and saturation variation at shallow depths, as well as volumes of water infiltrated in the porous medium. These aspects are important in the region of study, which is highly affected by soil movements, especially during the high- rain seasons that occur twice a year. The modeling exercise considers rainfall events that occurred between October and December 2014 and a hillslope that is currently monitored because of soil instability problems. Simulation results show that rainfall temporal variability, mesh resolution, coupling length, and the conceptual model chosen to represent the het- erogeneous soil, have a noticeable influence on results, partic- ularly for high rainfall intensities. Results also indicate that surface-subsurface coupled modeling is required to avoid unrealistic increase in hydraulic heads when high rainfall in- tensities cause top-down saturation of soil. This work is a first effort towards fostering hydrogeological modeling expertise that may support the development of monitoring systems and early landslide warning in a country where the rainy season is often the cause of hydrogeological tragedies associated with landslides, mud flow or debris flow. Keywords Colombia . Numerical modeling . Groundwater/ surface-water relations . Tropical climate . Natural hillslopes Introduction This paper focuses on hydrogeological modeling, which is required to better understand rainfall-induced landslides, an important class of landslides in Colombia, besides those caused by seismicity, volcanism, and human activities such as deforestation, land use management and mining (Ojeda and Donnelly 2006). The degree of saturation, and in particu- lar suction reduction, is one of the aspects that controls the triggering of failures along susceptible slopes (Crosta and Frattini 2003; Huat et al. 2006; Aristizábal et al. 2010b, 2011 ; Shanmugam and Wang 2015 ). Therefore, hydrogeological modeling of unsaturated soils takes priority over slope stability analysis, in order to identify where posi- tive pore pressure or suction reduction occur. A clear expla- nation of landslide triggering mechanisms related to hydroge- ology is provided by Brönnimann (2011). The Central and Western Andes of Colombia are charac- terized by a bimodal rainfall regime, with high-rain seasons in April–May and October–November (Poveda et al. 2011). The latter wet season is generally more severe in La Niña years, because this ocean–atmosphere coupled meteorological phe- nomenon brings even more rain than usual (Poveda et al. * Daniela Blessent dblessent@udem.edu.co 1 Facultad de Ingeniería, Universidad de Medellín, Carrera 87 N° 30-65, Medellín, Colombia 2 Procenvi Inc., 939 rue Chapman, G2E 3Z4 Quebec, QC, Canada 3 Department of Earth and Environmental Sciences, University of Ottawa, 120 University, K1N 6N5 Ottawa, ON, Canada 4 Departmento de Ingeniería Civil, Facultad de Minas, Universidad Nacional, MedellínCarrera 80 N° 65-223-Núcleo Robledo, Colombia 5 Present address: Agua y Medio Ambiente, Corporación Centro de Ciencia y Tecnología de Antioquia (CTA), Carrera 46 No. 56-11, piso 15, Edificio Tecnoparque, Medellín, Colombia Hydrogeol J DOI 10.1007/s10040-016-1482-z