Exhumation of the southern transpressive Bucaramanga fault, eastern Cordillera of Colombia: Insights from detrital, quantitative thermochronology and geomorphology

International audienceNew detrital apatite and zircon fission-track thermochronology ages were determined in four rivers (Onzaga, Los Micos, Mogoticos, and Chaguacá) located along the southern termination of the Bucaramanga Fault System in the Eastern Cordillera of Colombia. Younger populations of apatite fission-track (AFT) ages were found in the Onzaga and Los Micos rivers, both draining the central zone of a transpressive system. The detrital thermochronological data are used to estimate erosion rates, assuming topographic steady-state and spatially variable conditions. We also tested the link between long-to short-term erosion patterns employing Hotspot and Cluster Analysis (HCA) of Hack's stream length index (SL), which enables us to detect transient signals related to landsliding, lithologic variability, or tectonic processes. Based on this classification, we identified 175 SL-HCA anomalies distributed across the study area but mainly concentrated on the western side of the Chicamocha River where gigantic landslides are present. This topographic analysis reveals knickzones where higher erosion rates are expected. Consequently, those hotspots are the source of much of the sediment being evacuated through the Chicamocha River system. This favours the hypothesis that the area has spatially variable topographic conditions over time. To test this, we compare observed and predicted detrital age distributions derived from 3D thermal-kinematic inversions of existing in-situ thermochronological data assuming the two different topography conditions. We show that exhumation rates calculated with the bootstrapping method are closer to the predictions obtained from our 3D thermal-kinematic models with a spatially variable topography assumption. Furthermore, for the Onzaga River, the comparison of cumulative density functions between observed and predicted ages are statistically similar, which validates our hypothesis. Our results demonstrate that the topography has been variable over time, mainly from 40 Ma to the present. Exhumation rates in the study area evolved from ~2 km/m.y. between ~40 and 30 Ma, decreasing to values below 0.1 km/m.y. from 30 to 25 Ma, before increasing to 0.5 km/m.y. from ~20 Ma to the present. The results also indicate that a paleotopography representing 10 to 50% of the current topography was present from 80 to 40 Ma. During the interval between 40 Ma to the present, the model suggests that more than 50% of the present-day topography was created