Use of Neural Networks to Estimate a Global Self-Purification Capacity Index for Mountain Rivers: A Case Study in Bogota River Basin

摘要

In water resource management, data on water quality are not always available, so knowing which sections of a river need more regulatory or intervention actions to prevent water pollution is not possible. Therefore, this research sought to establish the global capacity for self-purification of water quality in rivers based on basic information readily available in their network of hydrographic and drainage basins. As a case study, the Bogota River basin was used, where a relationship was found between the values of the time series of the river water quality monitors and the physical characteristics of the basin and drainage network, using multivariate statistical tools (Artificial Neural Network). As a correlation of the self-purification potential obtained from the water quality data, a global self-purification index was formulated and calculated for mountain rivers that allow comparatively quantifying the assimilation capacity of the pollutants discharged into water bodies. The established index considers the processes based on the hydrotopographic characteristics such as speed, flow, and Length of the study section, as well as the information available on water quality, water quality objectives and the reactive-diffusive processes that each one suffers from the parameters included. The results obtained show the importance of the hydrotopographic parameters in the assimilation capacity of a river such as slope, annual precipitation, Temperature, Melton index, and percentage of watershed area. These parameters are used for agriculture, urban development, pasture, forests and number of discharge points, since their relationship with the aeration and sedimentation processes could be evidenced by its torrential regime, thus having the most significant reduction in the parameters of suspended solids and dissolved oxygen. This study is thus in turn supportive to the environmental river water pollution self-purification where water quality measurements are unavailable. [GRAPHICS] .