No abstract

No abstract

No abstract

Abstract:Many factors influence water quality within a reservoir. Deforestation, excessive erosion, introduction of new species, domestic and industrial waste disposal and agricultural runoff are only a few examples.It is well known by specialists in water resources management that water levels in a reservoir may also affect its quality. But how these processes occur and how appropriate water levels can be maintained are very hard questions to answer. This is because of the physical and biological processes occurring inside the water body, and also due to the various demands from society concerning water uses.Nowadays, through the use of models, knowledge of some of the conditions can enable us to predict future conditions. In many cases, reservoir models, such as physical models for water quality, may predict the future water quality situation. These models have been used successfully to enhance knowledge about the interactions among the different parts inherent to the water systems.Through the combination of water quality and optimization models, this study proposes a suitable methodology for the assessment of planning operations of a storage reservoir. The purpose of this paper is to consider a multipurpose reservoir, under different water demands and uses from societies, concerning reservoir water quality.The proposed optimization is realized through the use of dynamic programming combined with stochastic techniques that can handle the probabilistic characteristics of inflow quantity and quality. For the water quality assessment, the UNEP/ILEC one-dimensional model with two layers called PAMOLARE is applied. Finally, sensitivity analysis is carried out using a genetic algorithm model.

The present work focus was developing a system for early automatic detection of smoke plumes in visible-light images. The system used a realistic dataset gathered in 274 different days from a total of nine real surveillance cameras, with most smoke plumes being viewed from afar and 85% of them occupying less than 5% of the image area. We employed the innovative strategy of using the whole image for classification but "asking" the neural networks to indicate, in a multidimensional output, which image regions contained a smoke plume. The multidimensional output helped to focus the detector on the smoke regions. At the same time, the use of the whole image prevented wrong image classification caused by a constrained view of the landscape under analysis. Another strategy used was to rectify the detection results using a visual explanation algorithm, Gradient-weighted Class Activation Mapping (Grad-CAM), to ensure that detections corresponded to the smoke regions in an image. The detection algorithms tested were residual neural networks (ResNet) and EfficientNet of various sizes because these two types have given good results in the past in multiple domains. The training was done using transfer learning. Our dataset contained a total of 14125 and 21203 images with and without smoke, respectively, making it, to the best of the author's knowledge, one of the largest or even the largest reported dataset in the scientific literature in terms of the number of images with smoke collected from large distances of various kilometers. This dataset was fundamental to achieve realistic results concerning smoke detection efficiency. Our best result in the test set was an Area Under Receiver Operating Characteristic curve (AUROC) of 0.949 obtained with an EfficientNet-B0.