Akay, S. S., Et Al.

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Engineering Science and Technology, an International Journal 27 (2022) 101016

Contents lists available at ScienceDirect

Engineering Science and Technology,


an International Journal
journal homepage: www.elsevier.com/locate/jestch

Full Length Article

Quantification and visualization of flood-induced morphological changes


in meander structures by UAV-based monitoring
Semih Sami Akay a,⇑, Orkan Özcan b, Füsun Balık Sß anlı a
a
Geomatic Engineering, Faculty of Civil Engineering, Yıldız Technical University, Istanbul 34220, Turkey
b
Eurasia Institute of Earth Sciences Istanbul Technical University, Istanbul 34469, Turkey

a r t i c l e i n f o a b s t r a c t

Article history: River basins are one of the most dynamic terrains on Earth, therefore, river morphology should be mon-
Received 9 December 2020 itored continuously and its effects on the environment (i.e. natural hazards) should be mapped whenever
Revised 3 May 2021 possible. This study was carried out to quantify and visualize –flood- induced morphological changes in
Accepted 20 May 2021
meander structures by UAV-based monitoring. Two-year (2018 and 2019) field surveys were conducted
Available online 04 June 2021
in the Büyük Menderes River (BMR) Basin, Turkey. Multi-temporal orthomosaics and digital surface mod-
els (DSMs) were constructed using the Structure from Motion (SfM) technique to evaluate the change of
Keywords:
river morphology. Diachronic spatial and volumetric changes in the meander structures were assessed by
Natural hazard
Unmanned aerial vehicles (UAV)
the Digital Shoreline Analysis System (DSAS) and Digital Elevation Model (DEM) of Difference (DoD) tools.
Multi-temporal surveys This study shows the importance of mapping and monitoring river basins that have an impact on human
River morphology life in many ways with evaluating and analyzing the results of extreme events such as floods promptly.
Ó 2021 Karabuk University. Publishing services by Elsevier B.V. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction extreme events [5,6]. In cases where floods are unavoidable, the
damages can be minimized by taking necessary precautions [7].
Rivers, which are one of the most important natural water In particular, investigation of changes in bank line and sediment
sources on the earth surface and in the ecosystem, provide trans- quantities can provide information about possible future floods
portation, energy production, water supply to domestic, agricul- [8,9]. Understanding the changes in meander structures by means
tural and industrial activities while including animal and plant of either deposition or erosion through time is essential in order to
diversity as well. Therefore, some artificial and natural changes prevent the morphologic and river basin management problems
in rivers affect the ecosystem in river basins negatively such as and to develop strategies for the protection of river basins [10].
uncontrolled use of water, quarrying and pollution by wastes. In Morphological changes in river basins can be quantified by topo-
addition to the decline in water quality, the morphology of river graphical data obtained via various methods. Especially, the
beds changes with deposition and erosion on the river banks changes that occurred in riverbeds (i.e. channel changes etc.) can
[1,2]. Besides, the morphology of the river channel changes with be monitored with satellite and aerial imagery [11]. However,
high or low flow rates. Significant geomorphological changes changes over a short period of time, such as small-scale topo-
may occur in river beds as a result of changes in river water level graphic changes, may be topographically overlooked. UAVs are less
due to seasonal precipitation and temperatures, which causes affected by meteorological conditions and providing more flexible,
deposition and floods [3,4]. Therewithal, climate changes trigger spatial resolution and operational cost opportunities, when it is
extreme rainfall and flood events that negatively affect the devel- compared to satellite images [12,13]. Therefore, thanks to
opment of regions and cause deaths with environmental and eco- unmanned aerial vehicle (UAV) technology that can provide very
nomic damages. Climate change, which causes regional high-resolution data, morphological changes can now be examined
temperature differences, changes in wind type, precipitation in more detail [10]. Determining the changes that on surfaces by
regime and air quality, creates temperature changes and triggering natural events can be performed by producing fast and high accu-
racy data with UAVs [15-18]. The morphological changes in the
river beds and the failures by floods in the bankline can be deter-
⇑ Corresponding author. mined with multitemporal UAVs flights. Therefore, the amount of
E-mail addresses: semih.sami.akay@std.yildiz.edu.tr (S.S. Akay), ozcanork@itu. the erosion, deposition and failures can be determined [19,20].
edu.tr (O. Özcan), fbalik@yildiz.edu.tr (F. Balık Sßanlı).
The deposition and erosion rate of channel bed can be quantified
Peer review under responsibility of Karabuk University.

https://doi.org/10.1016/j.jestch.2021.05.020
2215-0986/Ó 2021 Karabuk University. Publishing services by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Semih Sami Akay, O. Özcan and Füsun Balık Sßanlı Engineering Science and Technology, an International Journal 27 (2022) 101016

by multi-temporal unmanned aerial vehicle (UAV)-based measure- Periodic field studies with UAV photogrammetric flights were car-
ments to estimate the sediment transport along the river [4,17,18]. ried out in January, June, and September in both 2018 and 2019 to
Nowadays, UAVs are widely used in various applications as an determine temporal changes in the meander structures.
alternative to satellite systems for monitoring and modeling the In the study, periodic field studies were carried out with two
earth surface. Although UAVs are affected by battery life limita- different types of multirotor UAVs having an integrated camera
tions and weather conditions such as wind, they can provide better and positioning systems. The payload weights and the camera res-
topographic data production of study areas in cloudy weather. Fur- olution of both UAVs (DJI Mavic Platinum Pro (MPP) and DJI Phan-
thermore, UAVs offer significant advantages in terms of time, cost, tom 3 Pro (P3P)), were 734 gr/1280 gr and 12.7 MP, respectively.
data acquisition, and image resolution as compared to terrestrial Flights were carried out with an overlap ratio of 85% in both frontal
photogrammetry and satellite remote sensing systems. UAVs are and lateral directions on all meander structures at 100-meter alti-
used to produce very high-resolution topography of riverbeds by tude. The same flight paths and parameters were used for every
Structure from Motion (SfM) method [14,21]. SfM, which brings periodic flight. In winter, when the UAV flights were affected by
automation for data analysis, plays an important role in the 3D the weather conditions, field studies were planned at the time of
reconstruction of complex surfaces over remote areas, and offers no precipitation. All flight parameters and the acquired data are
rapid solutions for change detection analysis [22]. It is seen that shown in Table 1.
the UAV-based SfM method offers significant advantages and Since there were neither permanent artificial nor natural
speed in natural hazards, disaster risk management, emergency objects existed that can be used as ground control points (GCPs),
and security operations, when time and scale parameters are taken portable GCPs were distributed in meander structures and the
into account. UAV-based SFM method is used in determining the coordinates of GCPs were measured with Real-Time Kinematic
effects in many areas such as floodplains, earthquakes, landslides, (RTK) GPS method (Fig. 2). For each meander structure, 4 to 6 GCPs
geomorphology and volcanic areas and planning land arrange- were arranged in a way that each GCP centered the tributary area
ments with high resolution and low cost data provided by mobile of the investigated field surface before each flight. In studies car-
operations [23-25]. Accordingly, digital surface models (DSMs) and ried out with UAVs, it is seen that as the number of GCPs homoge-
orthomosaics can be produced with high accuracy and spatial res- neously distributed over the field increases, the change in spatial
olution by UAVs, which are already used for morphodynamic mod- accuracy is low and this change is in millimeter degree [35,36].
eling of rivers, for tracking river ecosystem or evolution and for During coordinate measurements of the GCPs, the vertical and hor-
estimation of flood impacts on residential and non-residential izontal errors of coordinate were measured between 2.0 and
regions [17,18,26]. The surfaces where natural or artificial events 14.0 cm and 1.6–12.0 cm, respectively.
occur can be determined and morphological changes can be exam-
ined by mapping the surface with multi-temporal data. However,
UAVs also offer a quick solution to determine the damage or to take 3. Methods
timely precautions [4,13,26-28]. Besides, the amount of trans-
ported material due to either landslides or flood induced land use/- 3.1. Data generation and processing
cover changes in agricultural areas can be determined by
calculating the height differences between DSMs [29,30]. In order SfM method, which is a new, fast and low-cost photogrammet-
to monitor the temporal geomorphological changes, UAV-derived ric method, provides high accuracy three-dimensional data by
DSM data can be as an alternative to the traditional measurement matching the objects in the photo series with different overlap
methods [18,31,32]. Today, UAVs are preferred to determine the rates, taken at different locations [34]. Bundle adjustment, which
changes in river morphology, the amount of sediment erosion or provides camera positions and extraction of 3D positions of
deposition and the effects of flood events [4,28]. objects, is used in the process of determining parameters with high
In this study, it was aimed to monitor and quantify the sedi- accuracy and improve SFM method. In addition, bundle adjustment
ment changes over two years in three different meander struc- ensures that the cost function is minimized [37,38]. Point clouds,
tures, which were selected according to their features such as DSMs, and orthomosaics which can provide very precious data in
width, sinuosity, and level. Further, the effects of the flood, which many different large and small-scale studies, were produced by
occurred during the field study, were examined to monitor the SfM method in order to examine two-dimensional (2D) and
changes along the banklines and the amount of eroded/deposited three-dimensional (3D) morphological changes of the surfaces
sediment was quantified by the Digital Shoreline Analysis System periodically [2,17]. In the study, high-resolution data production
(DSAS) and DEM of difference (DoD) tools. processes were performed with SfM method by gathering the
entire data. Aerial photographs were acquired by UAVs during
the field studies and real-world coordinates of the GCPs were inte-
2. Study area and field surveys grated into the system for each meander structures (Fig. 2). Flight
plans were made with Pix4D capture software by drawing poly-
The Büyük Menderes River (BMR) Basin, which is located in gons to cover the study areas. GCP coordinate measurements were
south-western Anatolia, is Turkey’s seventh most populous basin carried out with the Leica GPS900 system. In order to increase the
that contributes to agriculture, industry, and natural ecosystems measurement accuracy, GCPs were assigned on aerial photographs,
with biological diversity and natural resources [33]. The BMR orig- and high-resolution spatiotemporal data were produced by using
inates from the plateaus near Dinar and Çivril in Central Western Pix4D software subsequently. Hereby, the DSM of the meander
Anatolia and it flows into the sea on the estuary of the Aegean structures were produced by a six-month period of orthomosaics.
Sea. There are ten provinces within the boundaries of the BMR All orthomosaics and DEMs were generated with a ground sam-
Basin, and that is located close to big city centers such as Aydın pling distance (GSD) of 5 cm. For the entire data, the average hor-
and Denizli. In the study, three meander structures were examined izontal and vertical of RMSE were calculated as 10.3 cm and 6.6 cm,
using long-term satellite images before the field studies. The study respectively. Changes in the banklines and in the sediment
locations were selected in accordance with their sinuosity. Further, amounts were mapped and quantified with the progress of the
the movement of the sediments was monitored to investigate the data production phase. 4
morphological changes in the selected meander structures along In January 2019, the heavy flood due to the extreme rainfall
with the BMR. Fig. 1 shows the study locations in the BMR Basin. affected the meander structures. The changes in the selected
2
Semih Sami Akay, O. Özcan and Füsun Balık Sßanlı Engineering Science and Technology, an International Journal 27 (2022) 101016

Fig. 1. Demonstration of study locations and the BMR Basin (The white box indicates the basin location on Turkey).”).

Table 1
Flight specifications of each UAV survey.

# of Flights Study Location Date # of Images UAV RMSE (pix)


Flight 1 Meander 1 28.01.2018 142 P3P 0.3
Meander 2 29.01.2018 146 P3P 0.3
Meander 3 29.01.2018 261 P3P 0.3
Flight 2 Meander 1 03.06.2018 143 P3P 0.3
Meander 2 04.06.2018 146 P3P 0.3
Meander 3 04.06.2018 259 P3P 0.3
Flight 3 Meander 1 26.09.2018 143 P3P 0.3
Meander 2 26.09.2018 143 P3P 0.2
Meander 3 26.09.2018 259 P3P 0.2
Flight 4 Meander 1 28.01.2019 314 P3P 0.3
Meander 2 27.01.2019 321 P3P 0.3
Meander 3 27.01.2019 405 P3P 0.3
Flight 5 Meander 1 10.06.2019 148 MPP 0.5
Meander 2 10.06.2019 415 MPP 0.3
Meander 3 10.06.2019 309 MPP 0.3
Flight 6 Meander 1 21.09.2019 142 P3P 0.3
Meander 2 21.09.2019 484 MPP 0.3
Meander 3 21.09.2019 315 MPP 0.3

meanders were investigated by comparing the orthomosaics that 3.2. Inter-comparison of morphological changes
were obtained before and after the flood event in one-year time
scale between January 2018 (Fig. 3a-3c for the Meander 1–3, The temporal changes that occur in land use/cover of the sur-
respectively) and January 2019 (Fig. 3d-3f for the Meander 1–3, face can be examined with the DSAS method. DSAS method pro-
respectively). Herein, it was observed that when the river water vides the calculation of the rates of intersection, change and
level increased, first the sediment areas were submerged, and then distance with the sections created by using time and location infor-
transported as expected. mation in linear areas such as river shoreline and glacier coasts
3
Semih Sami Akay, O. Özcan and Füsun Balık Sßanlı Engineering Science and Technology, an International Journal 27 (2022) 101016

[39]. In this study, the DSAS method was used to calculates the
changes in banklines by forming transecting lines between the
baseline and obtained banklines at different time intervals. First
of all, banklines of meander structures were determined on the
orthomosaics by digitization. After the process of banklines extrac-
tion, the DSAS method was applied to examine the changes in the
banklines from January 2018 to January 2019. Statistical analysis of
the DSAS method was done by determining the changes between
the first date and the last date (Net Shoreline Movement –
NSM) and the changes between the near and distant banklines
(Shoreline Change Envelope - SCE) [39]. The bankline that was
determined in January 2019 was used for constructing a reference
line as a baseline at a distance of 2 m to the land to minimize the
effect of vegetation on the height change. The changes in the
banklines were calculated by drawing sections between the bankli-
nes and the baselines at 1 m intervals. Fig. 4 shows the monthly
extracted banklines, reference line (baseline), and transects to
determine the dimensions of the changes on banklines between
2018 and 2019. As the changes along the banklines can be exam-
ined by visual interpretation from UAV-derived orthomosaics of
June and September for 2018 and 2019, it was quantified by
obtaining the average changes (NSM) between the first date and
the last date in the banklines that were between 1.8 m and
3.5 m respectively. Additionally, the maximum and the minimum
changes in the banklines (SCE) were observed to be in Meander 2
and Meander 1 with a change of 74.5 m and 0.9 m, respectively.
Furthermore, river centerlines were produced by using bankli-
nes and sinuosity index values (SI) shown in Fig. 5. Although they
were very close to each other in all other extractions, the centerli-
Fig. 2. Representation of (a) GCP distribution and aerial images locations on
nes extracted for January 2019 moved towards to the land due to
Meander 1 point cloud in Pix4D software, (b) + shaped landmarks used in UAV
surveys and (c) an example of a portable GCP on land. flooding.
The average SI values of Meander 1, 2 and 3 were calculated as
2.15, 1.15, and 2.25, respectively. Therefore, Meander 2 was

Fig. 3. UAV surveys of the meander structures belong to (a-c) January 2018 and (d-f) January 2019 (during the flood event).

4
Semih Sami Akay, O. Özcan and Füsun Balık Sßanlı Engineering Science and Technology, an International Journal 27 (2022) 101016

Fig. 4. Visualization of change detection along the banklines by the DSAS for each meander structures. Demonstration of the corresponding transects produced by DSAS and
banklines for (a) Meander 1, (b) Meander 2 and (c) Meander 3. The banklines extracted from orthomosaics for each month are shown in different colors.

Fig. 5. Representation of river centerlines produced with the banklines on the orthomosaics for (a) Meander 1 (b) Meander 2 (c) Meander 3.

defined as sinuous, while the remaining meanders were defined as 2018 (Fig. 6). In this study, the changes were carried out by calcu-
meandering according to Rosgen (1994) [40]. It is also observed lating areal and volumetric differences in order to examine them in
that there is direct relation between the sinuosity index, deforma- two and three dimensions. GCD can give the accuracy of volumet-
tions and failures in riverbeds. It can be inferred that riverbeds ric changes due to DEM resolution. Since there were no planimetric
with low sinuosity index were less prone to riverbank failures with measurements in the past, the interpretation of the areal accuracy
low sediment transport rate [41,42]. is depends on the acquired data accuracy.
The volumetric changes in sediment amounts were extracted
from UAV-derived DSMs. Geomorphic Change Detection (GCD),
which use DoD method, enables the determination of the volumet-
ric change by calculating the elevation differences in the DEM pro- 4. Quantitative results
duced by repeated measurements in river beds [43,44]. The DoD
method, which calculates volumetric changes between topo- Morphological changes in the meander structures were quanti-
graphic surfaces with a pixel-based basis in the difference of the fied by calculating sediment budgets in the regions where sedi-
DSMs, is widely used to monitor and interpret the geomorpholog- ment changes were higher rather than the entire meander
ical and topographical changes with multi-temporal DSMs [44-46]. structure which were determined by visual interpretation. It is
Likewise, changes on the surface as a result of morphological pro- clear that the area considered for volumetric changes is the chan-
cesses can be determined by calculating the difference of multi- ged area that is not submerged during the surveys, as it becomes
temporal DSMs with the DoD method [32]. In this study, the evident in C D and E from Fig. 6 and all the submerged area is
deposited/eroded regions and their changes over time were deter- not usable for calculations using only a photogrammetrical survey
mined by the DoD method in the period between 2018 and 2019. because of the limitations of this method in water covered areas, as
Following the DoD process, a change map was created showing well. Thus, a 5 m buffer-distance from the banklines of January
the change in pixel-by-pixel height. Totally six DSMs for each 2018 towards the land was created in order to provide reference
meander structures were used to calculate the volumes and to lines. The areas between the banklines and the reference lines were
quantify the topographic changes within the region of interests selected as the change regions according to which the change anal-
(ROIs) by computing a DoD change map. Fig. 6 shows an example yses of sediment budgets were performed. Hereby, the amount of
of DSMs for Meander 1 and the DoD change map produced with deposited/eroded sediment was calculated between January, June,
DSMs for each meander structure between January 2018 and June and September for 2018 and 2019. Fig. 7 shows the areal and
5
Semih Sami Akay, O. Özcan and Füsun Balık Sßanlı Engineering Science and Technology, an International Journal 27 (2022) 101016

Fig. 6. Example of the (a-c) DoD analysis for Meander 1, (d) Volumetric change analysis in the ROI of Meander 2 from January to June 2018, (e) Volumetric change analysis in
the ROI of Meander 3 from January to June 2018 and (f) Demonstration of the extracted ROIs with banklines for all surveys for Meander 1. The black dashed boxes on the
orthomosaics shows the areas where the analysis was carried out in the meander structures.

Fig. 7. Cumulative changes of areal and volumetric values between diachronic surveys in the associated meanders.

6
Semih Sami Akay, O. Özcan and Füsun Balık Sßanlı Engineering Science and Technology, an International Journal 27 (2022) 101016

volumetric values of the temporal sediment changes in the ROIs GCD tool and the sediment change accuracy was determined not
between 2018 and 2019. to exceed ± 15 m3 in the study locations. As a result of the study
The areal changes showed that the sediment amounts increased performed within two-year of monitoring, morphological changes
in these regions from January to September in 2018. Besides, the on the meander structures were calculated with high accuracy by
sediment deposition reached the highest value among others in UAV-based data.
September. In addition, Meander 3 has the highest deposition
value in both 2018 and 2019. Although the amount of sediment 4.1. Flood-induced morphological changes
deposition in 2019 approached the value in 2018, these values
did not exceed 2018. Alongside the temporal changes occurring in the stated ROIs,
The lowest spatial change was found in Meander 1 with 71.4 m2 some riverbank failures were observed (Fig. 8). The areal and vol-
between January and June for 2018 following by Meander 3 with umetric representation of the changes is expressed in Fig. 6 and
92.1 m2 between January and September for 2019. On the contrary, Fig. 8 with the results of the DoD and DSAS methods. Especially,
the highest change was in Meander 3 for both years with the failure with a dimension of 11.2 m in length and 5.4 m in width
5557.9 m2 and 11716.8 m2 between January and September, caused floodwaters to inundate the agricultural areas in Meander
respectively. 2.
Inter-comparison results showed that the flood had a high Flood-induced failures led to erosion with a dimension of
impact on sediment areas and the most spatial change occurred 93.8 m2 and 161.6 m2 in Meander 1 and Meander 2, respectively.
before and after the flood. On the other hand, it was observed that The change analysis before and after the flood showed that 134%
the spatial changes in June and September were similar for both of the sediment area in September disappeared as a result of the
years. Analysis refers to the increase and decrease in sediment val- flood in the erosion surface of Meander 1. After the flood, the
ues. The sediment areas were submerged due to the increase in the eroded area was reconstructed and the deposited area expanded
water level in January 2019. Therefore, the rising value referred to by 387% in June 2019. The spatial values of the 2018 and 2019 in
erosion values, which were calculated by the analyses carried out diachronic study of the spatial variation between January and June
with January 2019. In 2018, volumetric changes showed that the due to flooding were found to be 751% and 1196% for Meander 1
amount of sediment changes between June and September was and Meander 2, respectively. Table 2 shows the corresponding
higher than that of between January and June. In the two long- changes between 2018 and 2019.
term periods between January and September, the highest value In volumetric measurements, raising refers to the increase in
of the volumetric change was calculated as 10825.7 m3 in Meander surface height, lowering refers to the decrease. The total of sedi-
3 for 2018 and 39379.9 m3 in Meander 3 for 2019. The lowest vol- ment changes refers to the entire change determined as raising
umetric change was observed at Meander 1 for both years and and lowering. Therefore, the resulting erosion and failures are
these values were calculated as 435.9 m3 between January and expressed by lowering, and deposition and surface increase are
June and 723.1 m3 between June and September, respectively. expressed by raising. In addition to the total change, the display
The calculated accuracy in volumetric changes is revealed by the of the lowering and raising amounts is important in terms of

Fig. 8. Multi-temporal monitoring of flood-induced riverbank failures outside the ROIs (a-c) in Meander 1 and (d-f) in Meander 2.

7
Semih Sami Akay, O. Özcan and Füsun Balık Sßanlı Engineering Science and Technology, an International Journal 27 (2022) 101016

Table 2
Quantitative changes of riverbank failures in Meander 1 and Meander 2.

2018 Change Type Periods


January - June June - September January - September
Meander 1 (Erosion Surface) Area (m2) 42.7 46.7 89.3
Meander 2 (Erosion Surface) 18.5 34.8 53.3
Meander 1 (Erosion Surface) Volume (m3) Lowering Raising Total Lowering Raising Total Lowering Raising Total
14.9 15.3 30.2 35.5 6.5 42.0 89.9 30.9 120.8
Meander 2 (Erosion Surface) 0.3 45.2 45.5 0.1 26.9 27.0 6.8 59.7 66.5
2019 January - June June - September January - September
Meander 1 (Erosion Surface) Area (m2) 363.1 82.4 445.5
Meander 2 (Erosion Surface) 240.4 7.9 248.3
Meander 1 (Erosion Surface) Volume (m3) Lowering Raising Total Lowering Raising Total Lowering Raising Total
413.0 142.7 555.7 43.1 6.3 49.4 382.9 278.3 661.2
Meander 2 (Erosion Surface) 542.6 82.3 624.9 4.1 1.8 5.9 494.1 75.8 569.9

understanding the sediment mobility. The lowering caused by fail- It could be stated that meander structures showed similar
ures will enable the determination of the filling volume needed in changes while there was no flooding in the basin. Even though
the reconstruction process. Herein, the reason for the highest the meanders have been located in the same basin, the effects of
change between January and September 2019 is due to the these meanders have been observed to differ in flood conditions.
decrease in the sediment accumulation where the eroded areas Unless there is an artificial change in the basins, there will be sed-
were submerged. Volumetric values and visual interpretation from iment deposition areas in the meander structures with the
January to June in 2019 showed that the area of erosion returned to decrease of the water level and the flow of the stream. Further-
its structure in 2018. more, the flood caused erosion and failures in meander structures
and the agricultural areas were negatively affected. After the flood,
it was observed that the flood-induced failures were restored by
5. Results and discussion natural sediment transport and artificial filling operations.

In this study, the morphological changes along the meander


structures were monitored and mapped within a 2-year period
considering the flood induced effects by means of UAV based
methods which can be regarded as a rapid and efficient mapping 6. Conclusion
solution. Field surveys showed that the expansion in the sediment
regions (i.e. deposition), was observed from January to September Herein, the UAVs were proved to be very suitable for monitor-
with the deposition process in the existing sediment regions. Espe- ing river morphology by providing multi-temporal, very high spa-
cially in January 2019, new sediment regions were formed with a tial resolution orthomosaics and DEMs to calculate morphological
decrease in water level and high volumetric changes occurred. changes. UAVs can provide high-resolution topographic data of
Despite the sudden changes in the bankline due to the flood in Jan- river basins, where it is difficult or impossible to access because
uary 2019, the bankline in June and September 2019 was returned of natural surface structures and natural disasters. Therefore, UAVs
an equivalent position to 2018. The comparison of the periods of will be a part of the works that need to be monitored and con-
June and September for both years showed that the change is less trolled especially for mapping natural hazards. Since floods do
in the Meander 2 region. It can be interpreted that the deposition not occur frequently and are not continuous natural events, it is
process is more effective in the regions, where riverbank failures hard to make damage estimation. In the study, diachronic data sets
were observed. were obtained from UAV flights that were performed periodically
The areas in yellow color indicate little or no change in the with minimum exposure to weather conditions. Comparison of
bankline, whereas the red and blue colors show the ascending the same periods in 2018 and 2019 showed that the areal and vol-
and descending volume in Fig. 6. The surface elevation changes umetric change values have similar periodic changes.
of Meander 1, Meander 2 and Meander 3 were measured as In this study, periodic changes were investigated in the mean-
2.1 m, 1.2 m and 1.4 m for 2018, respectively, while they were dering structures of the BMR basin. The changes that occurred in
2.3 m, 3.7 m, 3.4 m for 2019. Comparing the same periods in both a non-flood year and a flood year were examined. It can be
2018 and 2019, it was determined that the highest change was stated that the flood effect was different in each meander structure
in Meander 2 with 6.2 m due to the flood event. In addition, centre- and similar surface changes can be seen in the non-flood processes.
lines of the meander structures have changed due to the increase This study shows that UAVs can be regarded as a reliable and faster
in the water level following the inundation, and the SI values have measurement method to investigate morphological changes in
decreased. Thus, even a decrease in the value of SI would indicate river basins, and the damages and failures can be minimized by
flooding in the basin. After the flooding in 2019, SI values reached monitoring and taking precautions.
the SI value in 2018 again. While failures occurred in Meander 1
and Meander 2 during flooding, there were no failures occurred
in Meander 3, although it has the highest SI value. As a result,
the relationship between SI and amounts of sediment has been Declaration of Competing Interest
revealed. The correlation of areal and volumetric sediment changes
was performed for the entire meander structures used in the study. The authors declare that they have no known competing finan-
Nominately; the correlations of areal and volumetric values with SI cial interests or personal relationships that could have appeared
were found as rArea = 0.52 and rVolume = 0.47, respectively. to influence the work reported in this paper.
8
Semih Sami Akay, O. Özcan and Füsun Balık Sßanlı Engineering Science and Technology, an International Journal 27 (2022) 101016

Acknowledgments BueCh, Hautes-Alpes, France, Int. J. Appl. Earth Obs. Geoinf. 73 (2018) 428–
443, https://doi.org/10.1016/j.jag.2018.07.016.
[20] G. Duró, A. Crosato, M.G. Kleinhans, W.S.J. Uijttewaal, Bank Erosion Processes
This work was supported by the Yıldız Technical University Sci- Measured With UAV-SfM Along Complex Banklines Of A Straight Mid-Sized
entific Research Project (YTU-BAP) [grant number FDK-2019- River Reach, Earth Surf. Dyn 6 (4) (2018) 933–953, https://doi.org/10.5194/
esurf-6-933-2018.
3552]; the Istanbul Technical University Scientific Research Project
[21] A. Santo, N. Santangelo, G. Forte, M. De Falco, Post flash flood survey: the 14th
(ITU-BAP) [grant number MGA-2017-40767]. and 15th October 2015 event in the Paupisi-Solopaca area (Southern Italy), J.
Maps 13 (2) (2017) 19–25, https://doi.org/10.1080/17445647.2016.1249034.
[22] S. Cucchiaro, M. Cavalli, D. Vericat, S. Crema, M. Llena, A. Beinat, L. Marchi, F.
Cazorzi, Monitoring topographic changes through 4D-structure-from-motion
References photogrammetry: application to a debris-flow channel, Environ. Earth Sci. 77
(18) (2018), https://doi.org/10.1007/s12665-018-7817-4.
[23] C. Gomez, H. Purdie, UAV- based Photogrammetry and Geocomputing for
[1] E. Dimitriou, E. Stavroulaki, Assessment of riverine morphology and habitat
Hazards and Disaster Risk Monitoring – A Review, Geoenviron Disasters 3
regime using unmanned aerial vehicles in a mediterranean environment, Pure
(2016) 23, https://doi.org/10.1186/s40677-016-0060-y.
Appl. Geophys. 175 (9) (2018) 3247–3261, https://doi.org/10.1007/s00024-
[24] A. Izumida, S. Uchiyama, T. Sugai, Application of UAV-SfM photogrammetry
018-1929-3.
And Aerial Lidar To A Disastrous Flood: Repeated Topographic Measurement
[2] B.o. Wang, Y.J. Xu, Decadal and episodic changes in morphology and migration
Of A Newly Formed Crevasse Splay Of The Kinu River, Central Japan, Nat.
of the confluence bar of two alluvial rivers in Louisiana, USA, J. Am. Water
Hazards Earth Syst. Sci. 17 (2017) 1505–1519, https://doi.org/10.5194/nhess-
Resour. Assoc. 56 (4) (2020) 615–629, https://doi.org/10.1111/jawr.
17-1505-2017.
v56.410.1111/1752-1688.12838.
[25] X. Li, B. Xiong, Z. Yuan, K. He, X. Liu, Z. Liu, Z. Shen, Evaluating the Potentiality
[3] S. Kabdaslı, Karasu sahili erozyon probleminin incelenmesi: ön değerlendirme
_ _ of Using Control-free Images from a Mini Unmanned Aerial Vehicle (UAV) and
raporu, I.T.Ü. Insaat Fakültesi, Su ve Deniz Bilimleri ve Teknolojisi Uygulama ve
Structure-from-Motion (SfM) Photogrammetry to Measure Paleoseismic
Arasßtırma Merkezi, 2010.
Offsets, Int. J. Remote Sens. 42 (7) (2021) 2417–2439, https://doi.org/
[4] S.S. Akay, O. Özcan, F.B. Sßanlı, T. Görüm, Ö.L. Sßen, B. Bayram, S.P. Aldrich, UAV-
10.1080/01431161.2020.1862434.
based evaluation of morphological changes induced by extreme rainfall events
[26] M. Rivas Casado, T. Irvine, S. Johnson, M. Palma, P. Leinster, The use of
in meandering rivers, PLoS ONE 15 (11) (2020) e0241293, https://doi.org/
unmanned aerial vehicles to estimate direct tangible losses to residential
10.1371/journal.pone.0241293.
properties from flood events: a case study of cockermouth following the
[5] C. Raymond, R.M. Horton, J. Zscheischler, O. Martius, A. AghaKouchak, J. Balch,
desmond storm, Remote Sens., 10, 2018, 1548, https://doi.org/10.3390/
S.G. Bowen, S.J. Camargo, J. Hess, K. Kornhuber, M. Oppenheimer, A.C. Ruane, T.
rs10101548.
Wahl, K. White, Understanding and Managing Connected Extreme Events, Nat.
[27] J. Suh, Y. Choi, Mapping hazardous mining-induced sinkhole subsidence using
Clim. Chang. 10 (7) (2020) 611–621, https://doi.org/10.1038/s41558-020-
unmanned aerial vehicle (drone) photogrammetry, Environ. Earth Sci. 76
0790-4.
(2017) 144, https://doi.org/10.1007/s12665-017-6458-3.
[6] M. Mirza, Mainstreaming Climate Change for Extreme Weather Events &
[28] O. Özcan, O. Özcan, Multi-temporal UAV based repeat monitoring of rivers
Management of Disasters: An Engineering Challenge, 2006 IEEE EIC Climate
sensitive to flood, J. Maps (2020), https://doi.org/10.1080/
Change Conference, Ottawa, ON, Canada, 1-10, (2006), https://doi.org/10.1109/
17445647.2020.1820387.
EICCCC.2006.277255.
[29] N. Pineux, J. Lisein, G. Swerts, C.L. Bielders, P. Lejeune, G. Colinet, A. Degré, Can
[7] M. Abdelkader, M. Shaqura, C. G. Claudel and W. Gueaieb, A UAV based system
DEM time series produced by UAV be used to quantify diffuse erosion in an
for real time flash flood monitoring in desert environments using Lagrangian
agricultural watershed?, Geomorphology 280 (2016) 122–136, https://doi.org/
microsensors, 2013 International Conference on Unmanned Aircraft Systems
10.1016/j.geomorph.2016.12.003.
(ICUAS), Atlanta, GA, 2013, pp. 25-34, https://doi.org/10.1109/
[30] R. Eker, A. Aydın, J. Hübl, Unmanned aerial vehicle (UAV)-based monitoring of
ICUAS.2013.6564670.
a landslide: Gallenzerkogel landslide (Ybbs-Lower Austria) case study,
[8] Q. Tang, O.S. Schilling, W. Kurtz, P. Brunner, H. Vereecken, H.J. Hendricks
Environ. Monit. Assess. 190 (2018) 28, https://doi.org/10.1007/s10661-017-
Franssen, Simulating flood-induced riverbed transience using unmanned aerial
6402-8.
vehicles, physically based hydrological modeling, and the ensemble Kalman
[31] A. Eltner, P. Baumgart, H.-G. Maas, D. Faust, Multi-temporal UAV data for
filter, Water Resources Research, 54, (2018), 9342–9363, https://doi. org/
automatic measurement of rill and interrill erosion on loess soil, Earth Surf.
10.1029/2018WR023067.
Process. Landforms 40 (6) (2015) 741–755, https://doi.org/10.1002/esp.3673.
[9] S. Jana, An automated approach in estimation and prediction of riverbank
[32] F. Neugirg, M. Stark, A. Kaiser, M. Vlacilova, M. Della Seta, F. Vergari, J. Schmidt,
shifting for flood-prone middle-lower course of the Subarnarekha river, India,
M. Becht, F. Haas, Erosion processes in calanchi in the Upper Orcia Valley,
International Journal of River Basin Management, (2019), https://doi. org/
Southern Tuscany, Italy based on multitemporal high-resolution terrestrial
10.1080/15715124.2019.1695259
LiDAR and UAV surveys, Geomorphology 269 (2016) 8–22, https://doi.org/
[10] G. Sofia, E.I. Nikolopoulos, Floods and rivers: a circular causality perspective,
10.1016/j.geomorph.2016.06.027.
Sci. Rep. 10 (2020) 5175, https://doi.org/10.1038/s41598-020-61533-x.
[33] A. Büke, M. Gültekin, A. Aksoy, B.B. Dıvrak, Ç. Göcek, M.Ö. Berke, H. Çesßmeci,
[11] N. Mohamad, M.F.A. Khanan, I.A. Musliman, W.H.W. Kadir, A. Ahmad, M.Z.A.
Büyük Menderes Havza Atlası, Istanbul, 2012, ISBN NO 978-605-62927-1-2.
Rahman, M.H. Jamal, M. Zabidi, N.M. Suaib, R.M. Zain, Spatio-temporal analysis
URL https://wwftr.awsassets.panda.org/downloads/atlas_web_download.pdf.
of river morphological changes and erosion detection using very high
[34] I. Colomina, P. Molina P, Unmanned aerial systems for photogrammetry and
resolution satellite image, IOP Conf. Series: Earth and Environmental,
remote sensing: a review, ISPRS J Photogramm Remote Sensing, 92, (2012),
Science 169 (2018), https://doi.org/10.1088/1755-1315/169/1/012020
79–97, https://doi.org/10.1016/j.isprs jprs.2014.02.013.
012020.
[35] E. Stott, R.D. Williams, T.B. Hoey, Ground Control Point Distribution for
[12] E. Karamuz, R.J. Romanowicz, J. Doroszkiewicz, The Use Of Unmanned Aerial
Accurate Kilometre-Scale Topographic Mapping Using an RTK-GNSS
Vehicles In Flood Hazard Assessment, J. Flood Risk Management 13 (4) (2020),
Unmanned Aerial Vehicle and SfM Photogrammetry, Drones 4 (3) (2020) 55,
https://doi.org/10.1111/jfr3.v13.410.1111/jfr3.12622.
https://doi.org/10.3390/drones4030055.
[13] K. Appeaning Addo, P.-N. Jayson-Quashigah, S.N.A. Codjoe, F. Martey, Drone As
[36] K.N. Tahar, An Evaluation on Different Number of Ground Control Points in
A Tool For Coastal Flood Monitoring In The Volta Delta, Ghana, Geoenviron,
Unmanned Aerial Vehicle Photogrammetric Block, Int. Arch. Photogramm,
Disasters 5 (1) (2018), https://doi.org/10.1186/s40677-018-0108-2.
Remote Sens. Spatial Inf. Sci. XL-2/W2 (2013) 93–98, https://doi.org/10.5194/
[14] A. Annis, F. Nardi, A. Petroselli, C. Apollonio, E. Arcangeletti, F. Tauro, C. Belli, R.
isprsarchives-XL-2-W2-93-2013.
Bianconi, S. Grimaldi, UAV-DEMs for Small-Scale Flood Hazard Mapping,
[37] J. Zhang, M. Boutin and D. G. Aliaga, Robust Bundle Adjustment for Structure
Water 12 (2020) 1717, https://doi.org/10.3390/w12061717.
from Motion, 2006 International Conference on Image Processing, 2185-2188,
[15] J. Rodriguez, R. Macciotta, M.T. Hendry, M. Roustaei, C. Gräpel, R. Skirrow,
(2006) https://doi.org/10.1109/ICIP.2006.312973.
UAVs For Monitoring, Investigation, And Mitigation Design Of A Rock Slope
[38] Y. Chen, Y. Chen, G. Wang, Bundle Adjustment Revisited, Computer Vision and
With Multiple Failure Mechanisms—A Case Study, Landslides 17 (9) (2020)
Pattern Recognition,1912.03858, (2019), https://arxiv.org/abs/1912.03858
2027–2040, https://doi.org/10.1007/s10346-020-01416-4.
[39] USGS, Digital shoreline analysis system (DSAS) Version 5.0 user guide, (2018),
[16] N. Ngadiman, N.M. Kasan, F.H. Hamzan, S.F.S. Zakaria, Riverbank Slope Erosion
Accessed 15 February 2020 URLhttps://pubs.er.usgs.gov/publication/
Monitoring using Unmanned Aerial Vehicle (UAV), Multidiscipl. Appl. Res.
ofr20181179.
Innovat. 2 (1) (2021) 13–24. https://doi.org/10.30880/mari.2021.02.01.002.
[40] D.L. Rosgen, A classification of natural rivers, Catena 22 (3) (1994) 169–199,
[17] S.S. Akay, O. Ozcan, O.L. Sen, Modeling morphodynamic processes in a
https://doi.org/10.1016/0341-8162(94)90001-9.
meandering river with unmanned aerial vehicle-based measurements, J.
[41] B. Mistri, Hydrogeomorphic Significance of Sinuosity Index in relation to River
Appl. Remote Sensing 13 (04) (2019) 1, https://doi.org/10.1117/1.
Instability: A Case Study of Damodar River, West Bengal, India Int. J. Adv. Earth
JRS.13.044523.
Sci. 1 (2) (2012) 49–57.
[18] B.U. Meinen, D.T. Robinson, Streambank topography: an accuracy assessment
[42] V.J.J. Janes, A.P. Nicholas, A.L. Collins, T.A. Quine, Analysis of Fundamental
of UAV-based and traditional 3D reconstructions Streambank topography: an
Physical Factors Influencing Channel Bank Erosion: Results for Contrasting
accuracy assessment of UAV-based and traditional 3D reconstructions, Int. J.
Catchments in England and Wales, Environ Earth Sci. 76 (7) (2017), https://doi.
Rem. Sen. 41 (1) (2020) 1–18, https://doi.org/10.1080/
org/10.1007/s12665-017-6593-x.
01431161.2019.1597294.
[43] Gomorphic Change Detection, GD user guide, (2018), Accessed 15 April 2021
[19] S. Hemmelder, W. Marra, H. Markies, S.M. De Jong, Monitoring River
URLhttp://gcd.riverscapes.xyz/.
Morphology & Bank Erosion Using UAV Imagery–A Case Study Of The River

9
Semih Sami Akay, O. Özcan and Füsun Balık Sßanlı Engineering Science and Technology, an International Journal 27 (2022) 101016

[44] J.M. Wheaton, J. Brasington, S.E. Darby, D.A. Sear, Accounting for Uncertainty in in check-dams, Science of the Total Environment, 678, (2019), 369-382,
DEMs from Repeat Topographic Surveys: Improved Sediment Budgets, Earth https://doi.org/10.1016/j.scitotenv.2019.04.332.
Surf. Process. Landforms 35 (2010) 136–156, https://doi.org/10.1002/ [46] S.D. Hamshaw, T. Engel, D.M. Rizzo, J. O’Neil-Dunne, M.M. Dewoolkar,
esp.1886. Application of unmanned aircraft system (UAS) for monitoring bank erosion
[45] A. Alfonso-Torreño, A. Gómez-Gutiérrez, S. Schnabel, J.F.L. Contador, J. Juan de along river corridors, Geomatics, Nat. Hazard. Risk, 10, (2019), 1, 1285-1305,
Sanjosé Blasco, M.S. Fernandes, sUAS, SfM-MVS photogrammetry and a doi.org/10.1080/19475705.2019.1571533.
topographic algorithm method to quantify the volume of sediments retained

10

You might also like