Previsão de vazões para o açude Orós/CE a partir de dados hidrometeorológicos utilizando perceptrons
The modeling of seasonal and interannual streamflow forecasting at northeastern Brazil represents a great relevance problem to the use and management of water resources; which demands greater prediction ability models. This is still a difficult task to solve due to the seasonal and interannual clima...
| Autores: | , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2015 |
| País: | Brasil |
| Institución: | Universidade Federal do Ceará (UFC) |
| Repositorio: | Repositório Institucional da Universidade Federal do Ceará (UFC) |
| Idioma: | portugués |
| OAI Identifier: | oai:repositorio.ufc.br:riufc/64253 |
| Acceso en línea: | http://www.repositorio.ufc.br/handle/riufc/64253 |
| Access Level: | acceso abierto |
| Palabra clave: | Redes neurais artificiais Previsão de vazões Açude Orós |
| Sumario: | The modeling of seasonal and interannual streamflow forecasting at northeastern Brazil represents a great relevance problem to the use and management of water resources; which demands greater prediction ability models. This is still a difficult task to solve due to the seasonal and interannual climate variability at the semi-arid region. This work presents the artificial neural networks (ANN) as an alternative for modeling the seasonal to interannual climate prediction,. For the development of this task the hydropraphic Oros weir Basin was chosen due to its importance as water resources in the State of Ceara. According to recent studies, the temperatures of the North Atlantic, South Atlantic and equatorial Pacific can be satisfactorily as predictors for the Northeast climate. The proposed model predicts, in July, the next rainy season (January to June) river flow regime. This time frame is of great relevance for the allocation of water resources. Among the studied models, those using the average temperature anomalies of April, May and June preceding the predicted year as input data showed the highest Nash-Suttcliffe efficiency (0.80). |
|---|