Cuantificación del riesgo por fenómenos de remoción en masa en Colombia: uso de sensores remotos en la evaluación de la precipitación antecedente como detonante
Thesis
On this research, a methodology was designed and implemented for the evaluation of the risk of landslide occurrence based on antecedent precipitation information obtained by remote sensors (CHIRPS V2.0) and the use of the Mapa Nacional de Amenaza por Movimientos en Masa scale 1:100.000 (Hazards by landslides map) by Disasters Management Institution from Colombia and the Inventario Nacional de Movimientos en Masa (Landslides database) from Colombian Geological Service. This research was divided into five stages; i) the first stage consisted in the construction of the precipitation climatology with the values reported daily by CHIRPS V2.0 with information from 1980 to 2020, ii) selection of most important events from the landslides database, through multiple criteria iii) definition of representative regions of the landslide, iv) analysis of antecedent precipitation for 15, 30, 60, 90 and 120 days from the selected events and finally v) definition the risk model for the definition of areas with excess antecedent precipitation susceptible to landslide. The final result of the model is presented as a 5 km x 5 km resolution raster file with a general performance of 53% capacity to identify areas with landslide events, reaching values of, inclusive, 68% for the identification in the second half of the year
