Comparación de metodologías estadísticas para el análisis de las variables registradas por la red de calidad hídrica de Bogotá en los ríos torca y salitre de la ciudad de Bogotá D.C.
Fecha
2017Registro en:
Balaguera, P. A. & Hernández, J. N. (2017). Comparación de metodologías estadísticas para el análisis de las variables registradas por la red de calidad hídrica de Bogotá en los ríos torca y salitre de la ciudad de Bogotá D.C. (Trabajo de grado de pregrado de Ingeniería Ambiental). Universidad Santo Tomás. Bogotá, Colombia.
reponame:Repositorio Institucional Universidad Santo Tomás
instname:Universidad Santo Tomás
Autor
Balaguera Vega, Paula Andrea
Hernández Chavarro, Julieth Nathalia
Institución
Resumen
Nowadays, there are diverse tools which can be used to realize deep analysis, that are more accurate and makes easier to read results of all what is related with water resources. Bogotá (as in many others cities with huge populations) has a problematic that has been growing in the last years, the persistent contamination in water bodies which are under their jurisdiction.
Nevertheless, as it has been said previously, there are tools that allow constantly monitoring of those water bodies, to get information in real time, like the water quality networks; in this case, the Bogotá’s water quality network, which is a system of monitoring points along the main rivers of the city (Salitre, Torca, Fucha and Tunjuelito rivers)
The main objective of this project is to create an improvement approach for the water quality network, focused to the two the rivers named previously (Salitre and Torca) through the application of an artificial neural network or connectionist system, that basically is a model of nodes interconnected, which allow to establish tasks that will be learned automatically by the network. The idea is to train the network to develop a specific action, which will improve the system by itself.
By now, in this paper will be shown the results of some statistics tests applied to the data obtained from the water quality network (linear discriminant analysis, hierarchical analysis, Kriging) taking into account that before this project, some students have realized an analysis for Tunjuelito and Fucha rivers respectively. So now, the idea is to contrast the methodologies applied for those rivers and see what is the difference between the information obtained for every river and how the methods work for every water body. Likewise, it is important to determinate another factors that affects directly in the information like population dynamics, domestic water spills, industrial water spills, anthropic activities, among others).