Una aproximación a la gestión ambiental del Piedemonte Amazónico Colombiano (Florencia Caquetá) desde la computación bioinspirada de alto desempeño

Abstract

The goal of this work was to develop an approach for environmental management through a model to predict flood prone areas in the Amazonian Piedmont of Colombia, from the perspective of high-performance bio-inspired computing. To begin, we developed a diagnostic of the environmental management to understand how the flood management was, from the following case studies: the basin of the Perdiz and Sardina rivers in the municipality of Florencia in the Department of Caquetá, central área of the Colombian Amazon Piedmont. After the diagnostic, the significant variables for the model were determined to integrate both environmental management and bio-inspired computing aspects; in order to do so, three different methods were used: exploratory, projective and experimental. These three types of methods allowed us to create a valuable tool, in the field of hydrology, to forecast and contribute to the visualization of flood prone areas, and prevent the loss of human lives in the event of floods. Finally, the high performance computing model was verified by comparing historical data with the data generated by the proposed model. The bio-inspired computing approach for environmental to management allowed us to contribute with the definition of a new model called Geo-inspired Natural Vector Agents (AVNG - Agentes Vectores Naturales Geo-inspirados), capable of integrating environmental, geographical and control aspects, in order to generate information related to the floods in the Colombian Amazon Piedmont.