As key-components of the urban-drainage system, storm-drains and manholes are essential to the hydrological modeling of urban basins. Accurately mapping of these objects can help to improve the storm-drain systems for the prevention and mitigation of urban floods. Novel Deep Learning (DL) methods have been proposed to aid the mapping of these urban features. The main aim of this paper is to evaluate the state-of-the-art object detection method RetinaNet to identify storm-drain and manhole in urban areas in street-level RGB images. The experimental assessment was performed using 297 mobile mapping images captured in 2019 in the streets in six regions in Campo Grande city, located in Mato Grosso do Sul state, Brazil. Two configurations of training, validation, and test images were considered. ResNet-50 and ResNet-101 were adopted in the experimental assessment as the two distinct feature extractor networks (i.e., backbones) for the RetinaNet method. The results were compared with the Faster R-CNN method. The results showed a higher detection accuracy when using RetinaNet with ResNet-50. In conclusion, the assessed DL method is adequate to detect storm-drain and manhole from mobile mapping RGB images, outperforming the Faster R-CNN method. The labeled dataset used in this study is available for future research.
Several Sediment Delivery Ratio (SDR) models have been used to estimate Sediment Yield (SY), mainly in data-scarce and ungauged basins, such as in many regions of Brazil. However, it is difficult to choose the most suitable SDR model, mainly because of the lack of investigations of this approach using observed data. Here, we investigated the performance of five widely used SDR models (SDREST) to estimate sediment yield values (SYEST ) based on observed data in a tropical watershed. We used observed sediment yield values (SY OBS) during September 2011 to July 2017 in three sub-basins of the Guariroba Basin, Midwestern Brazil. To estimate the average annual soil loss, we used the Revised Universal Soil Loss Equation. The SDROBS and SYOBS ranged from 5.56 to 10.54% and 940.76 to 5,400.32 t yr-1, respectively. The Williams and Berndt (1972) method presented the best performance, with a percent bias ranging from -2.34 to 3.30% in SRD estimation. Therefore, this model provided suitable SDR and SY estimates, and may be useful to estimate SY in other tropical data-scarce and ungauged basins.
Entender o comportamento hidrológico de bacias é algo complexo, em contrapartida, os modelos hidrológicos são ferramentas cujo objetivo é de representar o ciclo hidrológico em grandes áreas. Uma ferramenta que vem auxiliando no desenvolvimento de modelos hidrológicos de larga escala é o uso de Sistemas de Informações Geográficas, em que auxilia na análise de solo, cobertura e relevo, de tal forma que se consiga um entendimento mais completo dos processos das bacias hidrográficas. Neste estudo foi aplicado o Modelo de Grandes Bacias do Instituto de Pesquisas Hidráulicas na bacia do Alto Taquari com objetivo calibrar o modelo hidrológico, e posteriormente, testar combinações de cenários futuros de mudança climática e uso e ocupação do solo, a fim de detectar mudanças em variáveis hidrológicas. Os resultados da calibração foram aceitáveis, com o valor de Nash-Sutcliffe de 0,64 no exutório da bacia e as simulações das projeções indicaram a diminuição da vazão média mensal para grande parte dos cenários, possibilitando traçar tendências para uma melhor tomada de decisão no que tange a gestão da bacia hidrográfica.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.