Content uploaded by Adina Moraru
Author content
All content in this area was uploaded by Adina Moraru on Feb 03, 2023
Content may be subject to copyright.
The role of soft data to understand flood dynamics in an ephemeral mountain river
Adina Moraru 1
1 Department of Civil and Environmental Engineering, Norwegian University of Science and Technology, Trondheim, Norway,
adina.moraru@ntnu.no
ABSTRACT
The maximum inundation area, critical locations, and flow paths in the wetted area of a flash flood in a
steep river were reconstructed using remote sensing combined with post-event field investigation and
soft data. The soft data contrasted with control points and used to assemble the flood extension
included newspaper articles, photographs and videos taken during the flood, and witness testimonies.
The soft data provided insights on the flood dynamics that were not available with remote sensing and
post-event field observations alone. The outcome of this investigation could be used to calibrate and/or
validate numerical models, as well as to improve the accuracy of flood early warning systems.
Keywords: Field observations; flash floods; remote sensing; soft data; steep rivers.
1 INTRODUCTION
The flash flood affecting Storelva river, an ephemeral mountain river located in Utvik in Stryn municipality
(western Norway), in 2017 was documented on-site, witnessed by the inhabitants, and covered extensively by
the media regionally and nationally. This flash flood devastated Utvik and was the trigger to initiate extensive
research on how to better flood risk preparedness in small steep rivers. The extreme hydrologic event was
analyzed in Bruland (2020) (Fig. 1 top-left). This article served as basis for a further analysis of the hydro-
morphodynamic processes during the flash flood (Moraru et al., 2021). The objective of this study is to highlight
the potential that soft data (i.e. newspaper articles, photographs and videos taken on-site during the flood,
witness testimonies, etc.) has to provide relevant information of the flood dynamics that can be used
complementary to field observations after the event and remote sensing.
2 METHODS
The flooded area at peak discharge (Fig. 1a), critical locations (Fig. 1b), main flow paths (Fig. 1c) and,
subsequently, maximum wetted and dry areas during the flood event (Fig. 1d) were reconstructed using on-site
and post-event i) visual documentation, such as photographs and videos (Fig. 1 top-center), and ii) aerial
surveying, such as orthophotographs and laser scanning (Fig. 1 top-right), by mapping the observations
contrasted with georeferenced points in an iterative process validated in GIS (Moraru et al., 2021).
Figure 1. Workflow followed to obtain the flooded area at peak discharge (a), critical locations (b), flow paths
(c) and maximum inundation area (d) during the 2017 flash flood (modified after Moraru et al., 2021).
3 RESULTS & DISCUSSION
The four main critical locations identified in the ca. 800m-long reach are shown in Figure 2 (CL1 to 4, white
circles). Two of these critical locations are on bridges (CL2 and CL3), and the other two are located near a bank
with very fine material, easy to erode by the high flow velocities of a flash flood. The preferential flow paths
Proceedings of the 3rd IAHR
Young Professionals Congress.
Online. 28 November - 2 December 2022
ISBN: 978-90-832612-8-7
211
(purple lines in Fig. 2) indicate that the erosion occurred mainly on the left floodplain, mobilizing the main river
channel and large volumes of sediment across the village. The combination of the flooded area at Qpeak with the
flow paths resulted in the refined maximum inundation area (blue area in Fig. 2). The maximum wetted area
during the entire flood event was 93,658 m2. The post-event aerial images alone provided a narrower flood area
and only partial information on the flow paths. This information could be refined and expanded with soft data.
The main contributors to uncertainty were i) partial spatial and/or temporal coverage of the observations due to
inaccessibility at Qpeak, ii) inadequate data resolution, iii) expertise and subjectivity of the user. The discrepancies
between observations provided by soft data and aerial surveying and the mapped outcome are expected to be
the lowest in the urban area, where the reference objects are dense. The documentation is scarce in the right
floodplain, where only post-event data was available. The devastation caused by the flood was quickly repaired
to make the village accessible and habitable again. This made GPS measurements unavailable, hence making
the information provided by soft data essential to map and understand the dynamics of the 2017 flash flood.
Figure 2. Maximum inundation area (blue area), critical locations (white circles) and the flow paths (purple lines)
during the 2017 flash flood (modified after Moraru et al., 2021).
4 CONCLUSIONS
The combined use of soft data with aerial surveying provided information on the critical locations, flow paths
and maximum inundation area during the 2017 flash flood in Utvik in western Norway. Most trace of this event
disappeared with the quick restoration work, making the outcome of this study essential to understand the flood
dynamics in an ephemeral mountain river in a Nordic region. The dataset presented in this study can be used
to calibrate and validate numerical models, as well as to improve the accuracy of flood early warning systems.
ACKNOWLEDGEMENTS
The author thanks an anonymous reviewer for the constructive feedback. This publication is part of the World
of Wild Waters (WoWW) project, under the umbrella of the NTNU’s Digital Transformation initiative.
REFERENCES
Bruland, O. (2020). How extreme can unit discharge become in steep Norwegian catchments? Hydrology
Research, 51(2), 290–307. https://doi.org/10.2166/nh.2020.055
Moraru, A., Pavlíček, M., Bruland, O. and Rüther, N. (2021). The story of a steep river: causes and effects of
the flash flood on 24 July 2017 in western Norway. Water (Switzerland), 13(12), 1688.
https://doi.org/10.3390/w13121688
Proceedings of the 3rd IAHR
Young Professionals Congress.
Online. 28 November - 2 December 2022
ISBN: 978-90-832612-8-7
212
