Historic flood reconstructions for a safer future: The use of three types of surrogate models

A. Bomers; S.J.M.H.  Hulscher; B.  van der Meulen; R.M.J. Schielen

Historic flood reconstructions for a safer future: The use of three types of surrogate models

A. Bomers, S.J.M.H. Hulscher, B. van der Meulen, R.M.J. Schielen

Rivers, Ports, Waterways and Dredging Engineering

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

Abstract

Design discharges corresponding to large return periods are generally highly uncertain because of the relatively short data set of measured discharges. The uncertainty in these discharge predictions can be decreased by extending the data set of annual maximum discharges with historic flood events. However, efficient model approaches, in terms of computational times and model accuracy, are required to reconstruct historic flood events. In this study, the suitability of three data-driven surrogate models are evaluated, namely: a linear regression model, an artificial neural network and a support vector machine. The Rhine river flood event of 1809 is used as a case study. Although all types of surrogate models are capable of reproducing the maximum discharge during the 1809 flood event, the use of an artificial neural network resulted in the smallest 95% uncertainty interval.

Original language	English
Title of host publication	River Flow 2020
Subtitle of host publication	Proceedings of the 10th Conference on Fluvial Hydraulics
Editors	W. Uijttewaal, M.J. Franca, D. Valero, V. Chavarrias, C.Y. Arbos, R. Schielen, A. Crosato
Place of Publication	London
Publisher	CRC Press / Balkema - Taylor & Francis Group
Pages	1873-1880
Number of pages	6
Edition	1st
ISBN (Electronic)	978-1-003-11095-8
ISBN (Print)	978-0-367-62773-7
Publication status	Published - 2020
Event	River Flow 2020: The 10th Conference on Fluvial Hydraulics - Delft, Netherlands Duration: 7 Jul 2020 → 10 Jul 2020

Conference

Conference	River Flow 2020
Country/Territory	Netherlands
City	Delft
Period	7/07/20 → 10/07/20

Keywords

Surrogate modelling
historic flood reconstruction
discharge prediction
uncertainty interval

Access to Document

https://www.taylorfrancis.com/books/9781003110958

Cite this

Bomers, A., Hulscher, S. J. M. H., van der Meulen, B., & Schielen, R. M. J. (2020). Historic flood reconstructions for a safer future: The use of three types of surrogate models. In W. Uijttewaal, M. J. Franca, D. Valero, V. Chavarrias, C. Y. Arbos, R. Schielen, & A. Crosato (Eds.), River Flow 2020: Proceedings of the 10th Conference on Fluvial Hydraulics (1st ed., pp. 1873-1880). CRC Press / Balkema - Taylor & Francis Group. https://www.taylorfrancis.com/books/9781003110958

Bomers, A. ; Hulscher, S.J.M.H. ; van der Meulen, B. et al. / Historic flood reconstructions for a safer future: The use of three types of surrogate models. River Flow 2020: Proceedings of the 10th Conference on Fluvial Hydraulics. editor / W. Uijttewaal ; M.J. Franca ; D. Valero ; V. Chavarrias ; C.Y. Arbos ; R. Schielen ; A. Crosato. 1st. ed. London : CRC Press / Balkema - Taylor & Francis Group, 2020. pp. 1873-1880

@inproceedings{9eb66aea3cee4dd4b3095c7a2ad62a54,

title = "Historic flood reconstructions for a safer future: The use of three types of surrogate models",

abstract = "Design discharges corresponding to large return periods are generally highly uncertain because of the relatively short data set of measured discharges. The uncertainty in these discharge predictions can be decreased by extending the data set of annual maximum discharges with historic flood events. However, efficient model approaches, in terms of computational times and model accuracy, are required to reconstruct historic flood events. In this study, the suitability of three data-driven surrogate models are evaluated, namely: a linear regression model, an artificial neural network and a support vector machine. The Rhine river flood event of 1809 is used as a case study. Although all types of surrogate models are capable of reproducing the maximum discharge during the 1809 flood event, the use of an artificial neural network resulted in the smallest 95% uncertainty interval.",

keywords = "Surrogate modelling, historic flood reconstruction, discharge prediction, uncertainty interval",

author = "A. Bomers and S.J.M.H. Hulscher and {van der Meulen}, B. and R.M.J. Schielen",

year = "2020",

language = "English",

isbn = "978-0-367-62773-7",

pages = "1873--1880",

editor = "W. Uijttewaal and M.J. Franca and D. Valero and V. Chavarrias and C.Y. Arbos and R. Schielen and A. Crosato",

booktitle = "River Flow 2020",

publisher = "CRC Press / Balkema - Taylor & Francis Group",

edition = "1st",

note = "River Flow 2020 : The 10th Conference on Fluvial Hydraulics ; Conference date: 07-07-2020 Through 10-07-2020",

}

Bomers, A, Hulscher, SJMH, van der Meulen, B & Schielen, RMJ 2020, Historic flood reconstructions for a safer future: The use of three types of surrogate models. in W Uijttewaal, MJ Franca, D Valero, V Chavarrias, CY Arbos, R Schielen & A Crosato (eds), River Flow 2020: Proceedings of the 10th Conference on Fluvial Hydraulics. 1st edn, CRC Press / Balkema - Taylor & Francis Group, London, pp. 1873-1880, River Flow 2020, Delft, Netherlands, 7/07/20. <https://www.taylorfrancis.com/books/9781003110958>

Historic flood reconstructions for a safer future: The use of three types of surrogate models. / Bomers, A.; Hulscher, S.J.M.H. ; van der Meulen, B. et al.
River Flow 2020: Proceedings of the 10th Conference on Fluvial Hydraulics. ed. / W. Uijttewaal; M.J. Franca; D. Valero; V. Chavarrias; C.Y. Arbos; R. Schielen; A. Crosato. 1st. ed. London: CRC Press / Balkema - Taylor & Francis Group, 2020. p. 1873-1880.

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Historic flood reconstructions for a safer future: The use of three types of surrogate models

AU - Bomers, A.

AU - Hulscher, S.J.M.H.

AU - van der Meulen, B.

AU - Schielen, R.M.J.

PY - 2020

Y1 - 2020

N2 - Design discharges corresponding to large return periods are generally highly uncertain because of the relatively short data set of measured discharges. The uncertainty in these discharge predictions can be decreased by extending the data set of annual maximum discharges with historic flood events. However, efficient model approaches, in terms of computational times and model accuracy, are required to reconstruct historic flood events. In this study, the suitability of three data-driven surrogate models are evaluated, namely: a linear regression model, an artificial neural network and a support vector machine. The Rhine river flood event of 1809 is used as a case study. Although all types of surrogate models are capable of reproducing the maximum discharge during the 1809 flood event, the use of an artificial neural network resulted in the smallest 95% uncertainty interval.

AB - Design discharges corresponding to large return periods are generally highly uncertain because of the relatively short data set of measured discharges. The uncertainty in these discharge predictions can be decreased by extending the data set of annual maximum discharges with historic flood events. However, efficient model approaches, in terms of computational times and model accuracy, are required to reconstruct historic flood events. In this study, the suitability of three data-driven surrogate models are evaluated, namely: a linear regression model, an artificial neural network and a support vector machine. The Rhine river flood event of 1809 is used as a case study. Although all types of surrogate models are capable of reproducing the maximum discharge during the 1809 flood event, the use of an artificial neural network resulted in the smallest 95% uncertainty interval.

KW - Surrogate modelling

KW - historic flood reconstruction

KW - discharge prediction

KW - uncertainty interval

M3 - Conference contribution

SN - 978-0-367-62773-7

SP - 1873

EP - 1880

BT - River Flow 2020

A2 - Uijttewaal, W.

A2 - Franca, M.J.

A2 - Valero, D.

A2 - Chavarrias, V.

A2 - Arbos, C.Y.

A2 - Schielen, R.

A2 - Crosato, A.

PB - CRC Press / Balkema - Taylor & Francis Group

CY - London

T2 - River Flow 2020

Y2 - 7 July 2020 through 10 July 2020

ER -

Bomers A, Hulscher SJMH, van der Meulen B, Schielen RMJ. Historic flood reconstructions for a safer future: The use of three types of surrogate models. In Uijttewaal W, Franca MJ, Valero D, Chavarrias V, Arbos CY, Schielen R, Crosato A, editors, River Flow 2020: Proceedings of the 10th Conference on Fluvial Hydraulics. 1st ed. London: CRC Press / Balkema - Taylor & Francis Group. 2020. p. 1873-1880

Historic flood reconstructions for a safer future: The use of three types of surrogate models

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this