Articles | Volume 25, issue 5
Research article 18 May 2021
Research article | 18 May 2021
Resampling and ensemble techniques for improving ANN-based high-flow forecast accuracy
Everett Snieder et al.
No articles found.
Usman T. Khan and Caterina Valeo
Hydrol. Earth Syst. Sci., 20, 2267–2293,Short summary
This paper contains a new two-step method to construct fuzzy numbers using observational data. In addition an existing fuzzy neural network is modified to account for fuzzy number inputs. This is combined with possibility-theory based intervals to train the network. Furthermore, model output and a defuzzification technique is used to estimate the risk of low Dissolved Oxygen so that water resource managers can implement strategies to prevent the occurrence of low Dissolved Oxygen.
Related subject area
Subject: Urban Hydrology | Techniques and Approaches: Modelling approachesUrban surface water flood modelling – a comprehensive review of current models and future challengesSpace variability of hydrological responses of Nature-Based Solutions and the resulting uncertaintyEvent selection and two-stage approach for calibrating models of green urban drainage systemsModeling the high-resolution dynamic exposure to flooding in a city regionDrainage area characterization for evaluating green infrastructure using the Storm Water Management ModelCritical scales to explain urban hydrological response: an application in Cranbrook, LondonIncrease in flood risk resulting from climate change in a developed urban watershed – the role of storm temporal patternsPatterns and comparisons of human-induced changes in river flood impacts in citiesScale effect challenges in urban hydrology highlighted with a distributed hydrological modelComparison of the impacts of urban development and climate change on exposing European cities to pluvial floodingSpatial and temporal variability of rainfall and their effects on hydrological response in urban areas – a reviewHydrodynamics of pedestrians' instability in floodwatersFormulating and testing a method for perturbing precipitation time series to reflect anticipated climatic changesUsing rainfall thresholds and ensemble precipitation forecasts to issue and improve urban inundation alertsEnhancing the T-shaped learning profile when teaching hydrology using data, modeling, and visualization activitiesOn the sensitivity of urban hydrodynamic modelling to rainfall spatial and temporal resolutionPrecipitation variability within an urban monitoring network via microcanonical cascade generatorsEstimation of peak discharges of historical floodsIndirect downscaling of hourly precipitation based on atmospheric circulation and temperatureAssessing the hydrologic restoration of an urbanized area via an integrated distributed hydrological modelUsing the Storm Water Management Model to predict urban headwater stream hydrological response to climate and land cover changeEvaluating scale and roughness effects in urban flood modelling using terrestrial LIDAR dataContribution of directly connected and isolated impervious areas to urban drainage network hydrographsThermal management of an unconsolidated shallow urban groundwater bodyOnline multistep-ahead inundation depth forecasts by recurrent NARX networksA statistical analysis of insurance damage claims related to rainfall extremesJoint impact of rainfall and tidal level on flood risk in a coastal city with a complex river network: a case study of Fuzhou City, ChinaUrbanization and climate change impacts on future urban flooding in Can Tho city, VietnamMulti-objective optimization for combined quality–quantity urban runoff controlDevelopment of flood probability charts for urban drainage network in coastal areas through a simplified joint assessment approachAuto-control of pumping operations in sewerage systems by rule-based fuzzy neural networksCoupling urban event-based and catchment continuous modelling for combined sewer overflow river impact assessmentDynamic neural networks for real-time water level predictions of sewerage systems-covering gauged and ungauged sites
Kaihua Guo, Mingfu Guan, and Dapeng Yu
Hydrol. Earth Syst. Sci. Discuss.,
Revised manuscript accepted for HESS
Yangzi Qiu, Igor da Silva Rocha Paz, Feihu Chen, Pierre-Antoine Versini, Daniel Schertzer, and Ioulia Tchiguirinskaia
Hydrol. Earth Syst. Sci. Discuss.,
Revised manuscript accepted for HESSShort summary
Our research objective and originality is to investigates the uncertainties of the hydrological responses of Nature-based Solutions (NBS) that result from the multiscale space variability of both the rainfall and the NBS distribution. Results show that the intersection effects of spatial variability of rainfall and spatial arrangement of NBS can generate uncertainties on peak flow and total runoff volume estimations on NBS scenarios.
Ico Broekhuizen, Günther Leonhardt, Jiri Marsalek, and Maria Viklander
Hydrol. Earth Syst. Sci., 24, 869–885,Short summary
Urban drainage models are usually calibrated using a few events so that they accurately represent a real-world site. This paper compares 14 single- and two-stage strategies for selecting these events and found significant variation between them in terms of model performance and the obtained values of model parameters. Calibrating parameters for green and impermeable areas in two separate stages improved model performance in the validation period while making calibration easier and faster.
Xuehong Zhu, Qiang Dai, Dawei Han, Lu Zhuo, Shaonan Zhu, and Shuliang Zhang
Hydrol. Earth Syst. Sci., 23, 3353–3372,Short summary
Urban flooding exposure is generally investigated with the assumption of stationary disasters and disaster-hit bodies during an event, and thus it cannot satisfy the increasingly elaborate modeling and management of urban floods. In this study, a comprehensive method was proposed to simulate dynamic exposure to urban flooding considering human mobility. Several scenarios, including diverse flooding types and various responses of residents to flooding, were considered.
Joong Gwang Lee, Christopher T. Nietch, and Srinivas Panguluri
Hydrol. Earth Syst. Sci., 22, 2615–2635,Short summary
This paper demonstrates an approach to spatial discretization for analyzing green infrastructure (GI) using SWMM. Besides DCIA, pervious buffers should be identified for GI modeling. Runoff contributions from different spatial components and flow pathways would impact GI performance. The presented approach can reduce the number of calibration parameters and apply scale–independently to a watershed scale. Hydrograph separation can add insights for developing GI scenarios.
Elena Cristiano, Marie-Claire ten Veldhuis, Santiago Gaitan, Susana Ochoa Rodriguez, and Nick van de Giesen
Hydrol. Earth Syst. Sci., 22, 2425–2447,Short summary
In this work we investigate the influence rainfall and catchment scales have on hydrological response. This problem is quite relevant in urban areas, where the response is fast due to the high degree of imperviousness. We presented a new approach to classify rainfall variability in space and time and use this classification to investigate rainfall aggregation effects on urban hydrological response. This classification allows the spatial extension of the main core of the storm to be identified.
Suresh Hettiarachchi, Conrad Wasko, and Ashish Sharma
Hydrol. Earth Syst. Sci., 22, 2041–2056,Short summary
The study examines the impact of higher temperatures expected in a future climate on how rainfall varies with time during severe storm events. The results show that these impacts increase future flood risk in urban environments and that current design guidelines need to be adjusted so that effective adaptation measures can be implemented.
Stephanie Clark, Ashish Sharma, and Scott A. Sisson
Hydrol. Earth Syst. Sci., 22, 1793–1810,Short summary
This study investigates global patterns relating urban river flood impacts to socioeconomic development and changing hydrologic conditions, and comparisons are provided between 98 individual cities. This paper condenses and communicates large amounts of information to accelerate the understanding of relationships between local urban conditions and global processes, and to potentially motivate knowledge transfer between decision-makers facing similar circumstances.
Abdellah Ichiba, Auguste Gires, Ioulia Tchiguirinskaia, Daniel Schertzer, Philippe Bompard, and Marie-Claire Ten Veldhuis
Hydrol. Earth Syst. Sci., 22, 331–350,Short summary
This paper proposes a two-step investigation to illustrate the extent of scale effects in urban hydrology. First, fractal tools are used to highlight the scale dependency observed within GIS data inputted in urban hydrological models. Then an intensive multi-scale modelling work was carried out to confirm effects on model performances. The model was implemented at 17 spatial resolutions ranging from 100 to 5 m. Results allow the understanding of scale challenges in hydrology modelling.
Per Skougaard Kaspersen, Nanna Høegh Ravn, Karsten Arnbjerg-Nielsen, Henrik Madsen, and Martin Drews
Hydrol. Earth Syst. Sci., 21, 4131–4147,
Elena Cristiano, Marie-Claire ten Veldhuis, and Nick van de Giesen
Hydrol. Earth Syst. Sci., 21, 3859–3878,Short summary
In the last decades, new instruments were developed to measure rainfall and hydrological processes at high resolution. Weather radars are used, for example, to measure how rainfall varies in space and time. At the same time, new models were proposed to reproduce and predict hydrological response, in order to prevent flooding in urban areas. This paper presents a review of our current knowledge of rainfall and hydrological processes in urban areas, focusing on their variability in time and space.
Chiara Arrighi, Hocine Oumeraci, and Fabio Castelli
Hydrol. Earth Syst. Sci., 21, 515–531,Short summary
In developed countries, the majority of fatalities during floods occurs as a consequence of inappropriate high-risk behaviour such as walking or driving in floodwaters. This work addresses pedestrians' instability in floodwaters. It analyses both the contribution of flood and human physical characteristics in the loss of stability highlighting the key role of subject height (submergence) and flow regime. The method consists of a re-analysis of experiments and numerical modelling.
Hjalte Jomo Danielsen Sørup, Stylianos Georgiadis, Ida Bülow Gregersen, and Karsten Arnbjerg-Nielsen
Hydrol. Earth Syst. Sci., 21, 345–355,Short summary
In this study we propose a methodology changing present-day precipitation time series to reflect future changed climate. Present-day time series have a much finer resolution than what is provided by climate models and thus have a much broader application range. The proposed methodology is able to replicate most expectations of climate change precipitation. These time series can be used to run fine-scale hydrological and hydraulic models and thereby assess the influence of climate change on them.
Tsun-Hua Yang, Gong-Do Hwang, Chin-Cheng Tsai, and Jui-Yi Ho
Hydrol. Earth Syst. Sci., 20, 4731–4745,Short summary
Taiwan continues to suffer from floods. This study proposes the integration of rainfall thresholds and ensemble precipitation forecasts to provide probabilistic urban inundation forecasts. Utilization of ensemble precipitation forecasts can extend forecast lead times to 72 h, preceding peak flows and allowing response agencies to take necessary preparatory measures. This study also develops a hybrid of real-time observation and rainfall forecasts to improve the first 24 h inundation forecasts.
Christopher A. Sanchez, Benjamin L. Ruddell, Roy Schiesser, and Venkatesh Merwade
Hydrol. Earth Syst. Sci., 20, 1289–1299,Short summary
The use of authentic learning activities is especially important for place-based geosciences like hydrology, where professional breadth and technical depth are critical for practicing hydrologists. The current study found that integrating computerized learning content into the learning experience, using only a simple spreadsheet tool and readily available hydrological data, can effectively bring the "real world" into the classroom and provide an enriching educational experience.
G. Bruni, R. Reinoso, N. C. van de Giesen, F. H. L. R. Clemens, and J. A. E. ten Veldhuis
Hydrol. Earth Syst. Sci., 19, 691–709,
P. Licznar, C. De Michele, and W. Adamowski
Hydrol. Earth Syst. Sci., 19, 485–506,
J. Herget, T. Roggenkamp, and M. Krell
Hydrol. Earth Syst. Sci., 18, 4029–4037,
F. Beck and A. Bárdossy
Hydrol. Earth Syst. Sci., 17, 4851–4863,
D. H. Trinh and T. F. M. Chui
Hydrol. Earth Syst. Sci., 17, 4789–4801,
J. Y. Wu, J. R. Thompson, R. K. Kolka, K. J. Franz, and T. W. Stewart
Hydrol. Earth Syst. Sci., 17, 4743–4758,
H. Ozdemir, C. C. Sampson, G. A. M. de Almeida, and P. D. Bates
Hydrol. Earth Syst. Sci., 17, 4015–4030,
Y. Seo, N.-J. Choi, and A. R. Schmidt
Hydrol. Earth Syst. Sci., 17, 3473–3483,
J. Epting, F. Händel, and P. Huggenberger
Hydrol. Earth Syst. Sci., 17, 1851–1869,
H.-Y. Shen and L.-C. Chang
Hydrol. Earth Syst. Sci., 17, 935–945,
M. H. Spekkers, M. Kok, F. H. L. R. Clemens, and J. A. E. ten Veldhuis
Hydrol. Earth Syst. Sci., 17, 913–922,
J. J. Lian, K. Xu, and C. Ma
Hydrol. Earth Syst. Sci., 17, 679–689,
H. T. L. Huong and A. Pathirana
Hydrol. Earth Syst. Sci., 17, 379–394,
S. Oraei Zare, B. Saghafian, and A. Shamsai
Hydrol. Earth Syst. Sci., 16, 4531–4542,
R. Archetti, A. Bolognesi, A. Casadio, and M. Maglionico
Hydrol. Earth Syst. Sci., 15, 3115–3122,
Y.-M. Chiang, L.-C. Chang, M.-J. Tsai, Y.-F. Wang, and F.-J. Chang
Hydrol. Earth Syst. Sci., 15, 185–196,
I. Andrés-Doménech, J. C. Múnera, F. Francés, and J. B. Marco
Hydrol. Earth Syst. Sci., 14, 2057–2072,
Yen-Ming Chiang, Li-Chiu Chang, Meng-Jung Tsai, Yi-Fung Wang, and Fi-John Chang
Hydrol. Earth Syst. Sci., 14, 1309–1319,
Abrahart, R. J., Anctil, F., Coulibaly, P., Dawson, C. W., Mount, N. J., See, L. M., Shamseldin, A. Y., Solomatine, D. P., Toth, E., and Wilby, R. L.: Two decades of anarchy? Emerging themes and outstanding challenges for neural network river forecasting, Prog. Phys. Geog., 36, 480–513, https://doi.org/10.1177/0309133312444943, 2012. a, b, c
Bennett, N. D., Croke, B. F., Guariso, G., Guillaume, J. H., Hamilton, S. H., Jakeman, A. J., Marsili-Libelli, S., Newham, L. T., Norton, J. P., Perrin, C., Pierce, S. A., Robson, B., Seppelt, R., Voinov, A. A., Fath, B. D., and Andreassian, V.: Characterising performance of environmental models, Environ. Modell. Softw., 40, 1–20, https://doi.org/10.1016/j.envsoft.2012.09.011, 2013. a
Duncan, A.: The analysis and application of Artificial Neural Networks for early warning systems in hydrology and the environment, PhD thesis, College of Engineering, Mathematics and Physical Sciences, University of Exeter, Exeter, UK, 2014. a
Fleming, S. W., Bourdin, D. R., Campbell, D., Stull, R. B., and Gardner, T.: Development and operational testing of a super-ensemble artificial intelligence flood-forecast model for a pacific northwest river, J. Am. Water Resour. As., 51, 502–512, https://doi.org/10.1111/jawr.12259, 2015. a, b, c, d
Freund, Y. and Schapire, R. E.: Experiments with a New Boosting Algorithm, in: ICML'96: Proceedings of the 13th International Conference on Machine Learning, Bari, Italy, 3–6 July 1996, 148–156, 1996. a
Govindaraju, R. S.: Artificial Neural Networks in Hydrology. II: Hydrologic Applications, J. Hydrol. Eng., 5, 124–137, https://doi.org/10.1061/(ASCE)1084-0699(2000)5:2(124), 2000. a
Maier, H. R., Jain, A., Dandy, G. C., and Sudheer, K.: Methods used for the development of neural networks for the prediction of water resource variables in river systems: Current status and future directions, Environ. Modell. Softw., 25, 891–909, https://doi.org/10.1016/j.envsoft.2010.02.003, 2010. a, b
Papacharalampous, G., Tyralis, H., Langousis, A., Jayawardena, A. W., Sivakumar, B., Mamassis, N., Montanari, A., and Koutsoyiannis, D.: Probabilistic hydrological post-processing at scale: Why and how to apply machine-learning quantile regression algorithms, Water, 11, 2126, https://doi.org/10.3390/w11102126, 2019. a
Pisa, I., Santín, I., Vicario, J. L., Morell, A., and Vilanova, R.: Data preprocessing for ANN-based industrial time-series forecasting with imbalanced data, in: European Signal Processing Conference, 2019, European Signal Processing Conference, EUSIPCO, https://doi.org/10.23919/EUSIPCO.2019.8902682, 2019. a
Saffarpour, S., Erechtchoukova, M. G., Khaiter, P. A., Chen, S. Y., and Heralall, M.: Short-term prediction of flood events in a small urbanized watershed using multi-year hydrological records, in: Proceedings of the 21st International Congress on Modelling and Simulation (MODSIM2015), Broadbeach, Australia, 29 November–4 December 2014, 2234–2240, https://doi.org/10.36334/MODSIM.2015.L7.saffarpour, 2015. a
Toronto and Region Conservation Authority: Lower Don River West Remedial Flood Protection Project, available at: https://trca.ca/conservation/green-infrastructure/lower-don-river-west-remedial-flood-protection-project/ (last access: 12 May 2021), 2020a. a
Zhang, H., Yang, Q., Shao, J., and Wang, G.: Dynamic Streamflow Simulation via Online Gradient-Boosted Regression Tree, J. Hydrol. Eng., 24, 04019041, https://doi.org/10.1061/(ASCE)HE.1943-5584.0001822, 2019. a
Flow distributions are highly skewed, resulting in low prediction accuracy of high flows when using artificial neural networks for flood forecasting. We investigate the use of resampling and ensemble techniques to address the problem of skewed datasets to improve high flow prediction. The methods are implemented both independently and in combined, hybrid techniques. This research presents the first analysis of the effects of combining these methods on high flow prediction accuracy.
Flow distributions are highly skewed, resulting in low prediction accuracy of high flows when...