See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/387418492

Leveraging Remote Sensing for Exploring Climatic and Hydro‐Geomorphic

Linkages to Flooding: A Case of Bagmati River in Central Nepal

Article in River Research and Applications · December 2024

DOI: 10.1002/rra.4413

CITATIONS READS

0 28

5 authors, including:

Suraj Lamichhane Nischal Karki

Tribhuvan University University of Alabama
32 PUBLICATIONS 372 CITATIONS 4 PUBLICATIONS 12 CITATIONS

SEE PROFILE SEE PROFILE

Vishnu Prasad Pandey Pradhumna Joshi

Pulchowk Campus. Institute of Engineering. Tribhuvan University. Nepal Tribhuvan University
142 PUBLICATIONS 3,382 CITATIONS 4 PUBLICATIONS 2 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Nischal Karki on 26 December 2024.

The user has requested enhancement of the downloaded file.

River Research and Applications

RESEARCH ARTICLE

Leveraging Remote Sensing for Exploring Climatic and

Hydro-­Geomorphic Linkages to Flooding: A Case of
Bagmati River in Central Nepal
Suraj Lamichhane1 | Nischal Karki1 | Vishnu Prasad Pandey2,3 | Pradhumna Joshi1,3 | Sarita Dawadi4

1Center for Disaster Studies, Institute of Engineering, Tribhuvan University, Lalitpur, Nepal | 2 Department of Civil Engineering, Pulchowk Campus,

Institute of Engineering, Tribhuvan University, Lalitpur, Nepal | 3Center for Water Resources Studies, Institute of Engineering, Tribhuvan University,
Lalitpur, Nepal | 4Water Resources Research and Development Center (WRRDC), Government of Nepal, Lalitpur, Nepal

Correspondence: Vishnu Prasad Pandey ([email protected])

Received: 19 October 2024 | Revised: 29 November 2024 | Accepted: 2 December 2024

Keywords: fluvial processes | Nepal | remote sensing | river channel shifting | satellite imagery

ABSTRACT
Rivers are dynamic systems that evolve in response to different geologic, hydrologic, climatic, and anthropogenic factors. This
study utilized over three decades of remote sensing imagery to quantify river planform dynamics of the Bagmati River in central
Nepal. It maps the spatiotemporal changes in the active river channel extent and lateral migration rates from 1990 to 2022, re-
vealing an alarming decline in total river channel area by about 65% as compared to the historic extent. Lateral migration rates
of the river channel varied significantly along the reach ranging between 2 and 40 m per year (m/yr). Intra-­annual analysis of
the erosion and accretion areas from river masks before and after the monsoon season revealed a stronger correlation between
peak annual flow and accretion area in the study reach. Further exploration of a flood event in July 2019 using Sentinel-­1 radar
imagery revealed that the inundated area followed the historic channel extent, and multi-­day precipitation event had a major
contribution to triggering floods in the Bagmati. This study highlights the invaluable role of satellite remote sensing in studying
river planform dynamics, which helps devise informed approaches for flood mitigation, erosion control, and myriads of other
crucial endeavours to safeguard the population in the vicinity.

1   |   Introduction Thorne, and Watson 2000; Wu et al. 2017). The inherent bal-
ance between sediment deposition and erosion shapes the river's
Rivers are excellent examples of dynamic systems, perpetually course, impacting its width, depth, and overall configuration.
in motion and subject to constant change. The intricate interplay Additionally, human interventions, such as dam construction
of various geological, hydrological, and climatic factors imparts and land use/cover (LULC) changes, can further accelerate
remarkable instability to these watercourses (Charlton 2007; river instability by disrupting the natural equilibrium and al-
Dépret et al. 2015; Mirzaee et al. 2018; Shukla, Srivastava, and tering the sediment regime (Boix-­Fayos et al. 2007; Keesstra
Singh 2012). The primary drivers of their ever-­shifting nature et al. 2005; Kong et al. 2020; Xie, Yang, and Lundström 2021;
include rainfall patterns, sediment load, and the morphology of Yadav et al. 2023). Understanding these fluvial dynamics and
the riverbed itself. Rainfall intensity and frequency influence their multifaceted drivers is essential in mitigating the potential
the volume of water entering the river, thereby affecting its flow risks associated with floods, erosion, and habitat degradation.
velocity and discharge (Kale 2003). Sediment load, consisting of
eroded material from upstream sources, significantly contributes Due to their immense significance, typical methods to quan-
to the morphological evolution of the river channel (Biedenharn, tify fluvial dynamics have relied on traditional approaches

© 2024 John Wiley & Sons Ltd.

River Research and Applications, 2024; 0:1–13 1 of 13

https://doi.org/10.1002/rra.4413
 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
involving field measurements and mathematical model- the threat of flooding in terms of frequency, timing, and ex-
ling techniques (Asselman and Middelkoop 1995; Pietroń tent (Dingle et al. 2020; Sinha et al. 2014; Slater, Singer, and
et al. 2015; Sinha et al. 2019). These methods have offered Kirchner 2015). Given these implications, many studies have
valuable insights into the complex behaviours of river systems utilized remote sensing imagery and multi-­temporal analysis
but are often limited in spatial and temporal coverage and re- to monitor and understand fluvial processes in the Himalayan
quire calibration and validation data that can be challenging foreland basins (Baniya et al. 2023; Bhatpuria et al. 2022;
to acquire comprehensively. In response, satellite remote sens- Dewan et al. 2017; Khan and Islam 2003; Rakhal et al. 2021;
ing has emerged as an indispensable resource, revolutionizing Yadav et al. 2023). Most of the literature studies focus on large-­
methodologies in fluvial geomorphology research (Gilvear and scale snow-­fed basins, however, small to medium-­sized and
Bryant 2016; Kondolf and Piégay 2016). Satellite observations rain-­fed basins also have equal possibility for geomorphic
like Landsat and Sentinel have been frequently used in this changes (Sharma and Awal 2013; Van Appledorn, Baker, and
field of research (e.g., Feng et al. 2022; Gao et al. 2021; Jarriel, Miller 2019). As literatures in under-­observed rain-­fed basins
Swartz, and Passalacqua 2021; Langhorst and Pavelsky 2023; are quite limited due to dynamic monsoon patterns (Lebel,
Nyberg et al. 2023; Pickens et al. 2020; Wang et al. 2019) due to Garden, and Lebel 2011), this study aims to address the knowl-
their multi-­temporal and multi-­spectral nature. The Landsat edge gap with a case study in Nepal. Furthermore, the use of
satellite, with its extensive historical record dating back to remote sensing techniques has shown promise in flood inun-
1970, remains widely utilized despite its coarser resolution dation analysis, especially in under-­observed transboundary
compared to Sentinel. Various indices computed using different rivers of Nepal (Lamichhane et al. 2024).
band combinations such as normalized difference water index
(NDWI, McFeeters 1996), modified normalized difference The Bagmati River, originating from the capital city of Nepal,
water index (MNDWI, Xu 2006), normalized difference veg- poses a significant threat of flooding to the population residing
etation index (NDVI, Rouse et al. 1974), enhanced vegetation on its southern plain. The increase in rainfall intensity, proba-
index (EVI, Huete et al. 2002), and a combination of thereof are bly due to climate variability, coupled with haphazard urban-
used in fluvial research (Spada et al. 2018). Furthermore, the ization causing minimal groundwater recharge and increased
advent of cloud computing platforms like Google Earth Engine surface flow are some of the key reasons behind severe flooding
(GEE, Gorelick et al. 2017) has ushered in a new era of remote witnessed recently (Danegulu et al. 2024; Shamim 2024). Thus,
sensing applications affording ease of access to petabytes of this work complements the previous studies in understanding
data from satellite and aerial imaging systems, environmen- the fluvial processes in a monsoon-­fed river of southern Nepal
tal variables, climate, land cover, topographic and socioeco- by utilizing intra-­annual analysis in GEE, a cloud computing
nomic indicators as well as fast computing resources for their platform. The main objectives of this study are to: (i) quantify
analysis (Boothroyd et al. 2021). This powerful platform has spatiotemporal changes in channel extent and migration in the
enabled many studies ranging from regional to planetary Bagmati River, (ii) understand the role of monsoon peaks in re-
scales within the fluvial environment (Feng et al. 2022; Gao shaping channel planform and (iii) identify the relative contri-
et al. 2021; Jarriel, Swartz, and Passalacqua 2021; Langhorst bution of different precipitation events in triggering of flooding
and Pavelsky 2023; Nyberg et al. 2023; Pickens et al. 2020; events.
Wang et al. 2019).

The Himalayas—often recognized as the “water tower of 2   |   Materials and Methods

Asia”—hosts some of the world's largest and most distinctive
river systems. This intricate network sustains the livelihoods 2.1   |   Study Area
of millions of downstream inhabitants primarily through
hydro-­energy generation, drinking water supply and agricul- The Bagmati River originates in the Shivapuri hills of the
tural production (Biemans et al. 2019; Immerzeel et al. 2020; Mahabharat range, located north of the capital city of Nepal,
Rasul 2014). There is a unique source-­sink linkage between Kathmandu. A few kilometres downstream of the origin, the
the upstream and downstream regions in the Hindu-­ Kush river flows through Kathmandu, Bhaktapur, and Lalitpur dis-
Himalayan (HKH) region. The young geology, active tectonics, tricts, some of the most populous and rapidly urbanizing cit-
steep relief, and intense summer monsoon provide favourable ies of Nepal (Lamichhane and Shakya 2019). The river passes
conditions for sediment generation and transport in this region through the inner Mahabharat range and reaches the plains of
(Andermann et al. 2012; Scott et al. 2019). About 80% of the Terai before merging with the Koshi River in India (Figure 1a).
annual precipitation in the Himalayan region is concentrated The river eventually meets the Ganges River before draining
during the summer monsoon season (July–September) causing into the Bay of Bengal.
erosion and mass-­wasting in the upstream and increasing sed-
iment concentration in the rivers (Kale 2003; Morin et al. 2018; The Bagmati River Basin (BRB) can be divided into three geo-
Qazi and Rai 2018; Struck et al. 2015). The fast-­flowing riv- physical regions: the upper part is Kathmandu Valley; the mid-
ers of the Himalayas not only act as perfect conveyors of these dle section is dominated by hills in the Mahabharat range, and
sediments but also contribute to their volume by toe-­cutting-­ the lower reach lies within the Indo-­Gangetic plain. The ele-
induced landslides and bank erosion. A significant propor- vation within the basin ranges from 80 to 2800 masl. We focus
tion of this sediment is deposited in the southern flatland, on the lower-­most reach of the river as the river enters the
fundamentally altering the morphological properties of the plains from the hills up to the Nepal-­India border (Figure 1b).
river channels and adjacent floodplain. This transformative The length of the reach is 57 km and its start is marked with a
sediment deposition can constrict channel capacity, elevating discharge gauging station (station index. 589.5, Bagmati River

2 of 13 River Research and Applications, 2024

 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FIGURE 1    |    Study area—(a) location of the Bagmati River and its confluence with the Koshi River and the Ganges. (b) Bagmati basin and analysis
extent used in this study and (c–e) false-­colour composite image of 2022 from Sentinel-­2 satellite imagery.

at Padherodovan) making it feasible to study the relationship c­ omposite using a median reducer for each year from 1990
between annual peak flow and channel extent changes. The to 2022.
average annual flow recorded at this station from 2000 to 2014
is 125.4 m3/s and the average annual maximum and minimum We defined an active river channel as a combination of wet-
flow are 3169.2 and 9.9 m3/s, respectively. ted channel and alluvial deposits (Boothroyd et al. 2021;
Monegaglia et al. 2018). Several indices are defined in the
literature to characterize river channels (Bormudoi 2024;
2.2   |   Active River Channel Mask Generation Kodimalar, Vidhya, and Eswar 2020). However, for the wet-
ted channel, we followed a classification scheme by Zou
We used all the Landsat surface reflectance collection-­ 2 et al. (2018) using three different multispectral indices:
(Masek et al. 2020; Wulder et al. 2022) imagery available on MNDWI, NDVI and EVI. Pixels where MNDWI values ex-
GEE to extract active river channel masks. The Landsat 5 the- ceeded both NDVI and EVI values and where EVI values were
matic mapper (TM), Landsat 7 enhanced thematic mapper less than 0.1 were classified as wetted channel. To classify al-
(ETM+), and Landsat 8 operation land imager (OLI)/thermal luvial deposits, a dual thresholding approach on the NDVI and
infrared sensor (TIRS) dataset were merged into a single MNDWI images was applied. Pixels with NDVI values below
image collection for the region of interest. A cloud masking the set threshold and MNDWI values above their designated
­a lgorithm (Foga et al. 2017) was applied to remove cloudy pix- threshold were classified as alluvial deposits. We identified
els from every image in the merged collection. The cloud-­ these parameter values using trial-­and-­error and found the
masked images were then combined into a single ­ a nnual MNDWI parameter ranging from − 0.35 to − 0.4 and the NDVI

3 of 13
 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
parameter between 0.15 and 0.2 showed good classification of where, Mcl is the centreline migration rate, Acl is the area tra-
alluvial deposits for our study region. versed by the centereline, Δ t is the time interval and Lcl is the
length of centreline.

2.3   |   Channel Centreline Delineation and Mesh

Generation 2.4   |   Inter-­Linkage Between Monsoon Floods
and Channel Morphology
We extracted the channel centreline for each individual river
channel mask using the RivGraph Python package (Schwenk and We used two approaches to observe how monsoon floods affect
Hariharan 2021). RivGraph is a Python package that automates the channel morphology in the Bagmati River. We first per-
the extraction and characterization of river channel networks from formed a correlation analysis between daily discharge and chan-
a user-­provided binary image of a channel network. These channel nel extent changes to infer alterations induced by hydrology.
centrelines were used to infer lateral migration of the river chan- Since the discharge data was only available up to 2014, a flood
nel throughout the years. In order to obtain spatiotemporal re- extent map based on Sentinel-­1 satellite imagery was also used
sults, we followed a very simple approach to creating a mesh layer to analyse if there was any correspondence between the historic
(Figure 2). We first extracted a common centreline by combining planform dynamics and flood inundation. These approaches are
river masks for all years into a single mask. We then added perpen- further described in-­depth as follows.
dicular transects at 250 m intervals along the common centreline.
The centreline was then buffered to cover all the channel masks
and the mesh was created as the intersection polygon of the tran- 2.4.1   |   Based on Observed Daily Runoff
sects and buffered centreline. Within each mesh grid, the areal ex-
tent of the active river channel for all the years was computed. The In order to understand how the monsoon flood changes the
migration rate within each mesh grid was also computed as the river channel extent, we performed an intra-­a nnual analysis
area traversed by the centreline normalized by the length of the by extracting the river mask for dry (January–May) and wet
centreline (Boothroyd et al. 2021; Langhorst and Pavelsky 2023; (June–December) periods. We further categorized the change
Schwenk and Hariharan 2021), given by in channel extent as erosion and accretion areas using a pixel-­
wise differencing approach (Langhorst and Pavelsky 2023;
Acl . Δ t Rowland et al. 2016). The comparison of channel masks
Mcl =
Lcl (1)
during wet and dry periods enabled the classification of pixels

FIGURE 2    |    Mesh generation and migration rate computation: (a) Combined channel mask of all years from 1990 to 2022, (b) extraction of com-
mon channel centreline from the combined mask, transect lines at 250 m intervals, and buffered centreline covering combined channel mask, (c)
zoomed in section showing mesh generated using the intersection of buffered centreline and transects and (d) area traversed by the centreline from
time period T1 to T2.

4 of 13 River Research and Applications, 2024

 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
unique to the wet period as accretion areas, while pixels disap- 2.5   |   Determination of Flood Drivers
pearing during the wet period were deemed erosion sites. The
accretion and erosion areas for all the years were computed We conducted an event-­based analysis to discern the underlying
at the mesh grid scale to get the spatiotemporal picture of the factors contributing to flooding within the BRB. We first identi-
channel extent change. fied flood events using the peak over threshold (POT) method,
with the threshold value of the 99th percentile of daily streamflow
We also acquired daily streamflow data at one gauging measurements. Flood events occurring within a 3-­day window
­station upstream of our analysis extent from the Department were treated as a single event. This thresholding resulted in 56
of Hydrology and Meteorology (DHM), Nepal. The data different flooding events from 1990 to 2014 in the Bagmati River.
spanned from 2000 to 2014. Our investigation examined the For each event, we examined three different precipitation-­based
correlation between the magnitude of annual flood peaks and events using (i) multiday precipitation (Pmd), (ii) Extreme precip-
variations in erosion and accretion patterns across the stud- itation (P99) and (iii) antecedent precipitation index (API). We
ied area. acquired daily precipitation measurements from ground-­based
stations within and around the basin from DHM. The Pmd was de-
fined as the cumulative sum of 3-­day precipitation and was con-
2.4.2   |   Satellite-­Based Flood Mapping sidered a driver if its value, before the flooding event, exceeded
the 95th percentile of the long-­term average 3-­day precipitation.
In July 2019, the Bagmati River experienced continuous mas- Extreme precipitation events (P99) were defined as precipitation
sive rainfall leading to a devastating flood on July 13, affecting events exceeding the 99th percentile of the values on rainy days
a large number of lives and property. We extracted a flood ex- (p > 1 mm). We also used the API value defined as the ratio of 30-­
tent map from Sentinel-­1 satellite data for the July 2019 flood- day cumulative precipitation to the long-­term average precipita-
ing event to explore potential connections between channel tion of that period as a proxy of antecedent moisture conditions in
planform dynamics and spatial flooding patterns. Radar im- the basin (Marchi et al. 2010). We considered API values greater
ages from the Sentinel-­1 mission, available on the GEE plat- than 1.5 to be the potential contributor to flooding. For each flood
form, were used for mapping flood/inundation extents. The event, we assessed the number of times Pmd, P99 or API events
Sentinel-­1 image collection in GEE contains Level-­1 Ground preceded the flooding event at each precipitation station in rela-
Range Detected (GRD) scenes that have been processed to pro- tion to the total number of flood events.
vide backscatter coefficient values in decibels (dB). To derive
the backscatter coefficient in each pixel, GEE uses a series of
pre-­processing steps, including applying an orbit file, removing 3   |   Results and Discussion
noise and invalid data, applying radiometric calibration values,
and performing terrain correction. 3.1   |   Annual River Channel Extent and Migration

The imagery available on June 13, 2019, was extracted from We found a significant reduction in channel extent across all seg-
GEE and a series of a pre-­processing step including refined-­lee ments of the study section (Figure 3). The channel experienced
speckle filtering (Lee 1981) and slope correction using MERIT a substantial drop of around 65% in extent from 1990 to 2022
DEM (Yamazaki et al. 2017) were applied to the image. We with a decline rate of 1.85 km2 per year. During the 1990s, the
selected VV-­ polarization for the surface water classification channel extent exhibited larger variability ranging from around
because preliminary analysis using VH-­ polarization did not 45 to 90 km2, with an average extent of 66.07 km2. In the subse-
perform well in our study region. To classify surface water from quent decade, the channel's extent displayed a decreasing trend.
the Sentinel-­1 image, an adaptive thresholding algorithm called Although slight fluctuations were observed in the early 2000s,
“edge-­Otsu” (Markert et al. 2020) was applied. The algorithm the extent continued to decrease, reaching around 45 km2 in
separates the land and water pixels using an initial threshold 2010. The declining trend continued steadily from 2010 to 2022,
which was found to be − 16 for our analysis through trial-­and-­ reaching a lowest of around 20 km2 in 2022.
error. The initial binary classification is then used to detect edges
and sample pixels from the buffered edges to ensure bimodality The channel occurrence frequency is shown in Figure 4a, which is
in the histogram. The final threshold for surface water classi- the percentage of a particular pixel that was classified as an active
fication was then computed by feeding the sampled histogram river channel. Two prominent sections of the river with low occur-
into Otsu's algorithm [Otsu 1979]. However, the derived surface rence frequency were identified, one in the upstream and the other
water map also includes permanent water bodies that need to be in mid-­stream sections which is also evident in the comparison of
removed in order to extract the flooded pixels only. We obtained river channel masks for 1990 and 2022 (Figure 4b).
the permanent water layer from the Dynamic World v2.0 data-
set (Brown et al. 2022), a near-­real-­time landcover classification The centreline position of the river for each year is illustrated in
obtained from Sentinel-­2 imagery. We filtered the dataset for the Figure 5a, showing the gradual shifting of the river from 1990 to its
year 2019 and reduced the collection using a mode-­reducer in original position. The circular insets depict Google Earth imagery
order to obtain permanent water bodies in the study region. In showcasing the prominent migration zones. Notably, a significant
the post-­processing step of the surface water map, we removed alteration in the active channel area arises from the separation of
the permanent water pixels as well as those water pixels for the left branch, a transformation still discernible in recent imagery.
which the height above the nearest drainage (HAND) values, In the 1990s, the channel functioned as a vital water supplier to
obtained from the MERIT-­hydro dataset (Yamazaki et al. 2019), the primary branch. However, it has since become disassociated,
exceeded 15 m. presumably due to water extraction for irrigation in the adjacent

5 of 13
 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FIGURE 3    |    (a) Annual channel extent along the river centreline. Gray lines show channel extent for individual years. (b) Total annual channel
extent across the study reach.

FIGURE 4    |    Changes in channel extent from 1990 to 2022. (a) frequency of pixels classified as channel and (b) comparison of channel mask for
years 1990 and 2022.

fields. In the downstream section, a notable and gradual move- decade up to 2002. The river further shifted to its present location
ment of the river channel is observable. This shifting process un- around 2003–2004, where it has since remained stable. Evidently,
folded in three distinct phases. The initial position, represented by remnants of the river's previous course are discernible in satellite
the solid black line, remained consistent until 1992. Subsequently, imagery, underscoring the historical path. The migration zone is
the river transitioned to the second position, illustrated by a clus- predominantly characterized by agricultural fields and scattered
ter of light blue lines, and remained in the same position for a settlements. The rate of river channel shift was computed using a

6 of 13 River Research and Applications, 2024

 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FIGURE 5    |    (a) River channel centreline positions for each year from 1990 to 2022. Circular insets show the area of major channel migration. (b)
Spatial channel migration rate along the channel centreline.

centreline migration approach extracted from the channel masks 3.2   |   The Role of Monsoon Season in Channel
within each mesh grid and stored in the centreline vector layer Dynamics
for 2022 (Figure 5b). The average migration rate ranged from 2
to 40 m/yr. across the river with higher rates in the downstream In order to understand channel dynamics driven by the monsoon
sections than the upstream. season, we examined the changes before and after the monsoon
season (inclusive) for each year. To do this, we further disaggre-
Many studies have shown the role of extreme flood in reshap- gated the total channel extent change into accretion and ero-
ing the channel through processes such as bank erosion, aggra- sion areas. The spatiotemporal variation in net channel extent
dation, incision, and avulsion (Bryndal et al. 2017; Dewan change, defined by the difference in accretion and erosion area is
et al. 2017; Rusnák and Lehotskỳ 2014; Yousefi et al. 2018). In shown in Figure 6a. Throughout the duration of our study, a gen-
the monsoon of 1993, the Bagmati River experienced a 100-­year eral trend of erosion upstream (up to around 20 km downstream
return period flood caused due to torrential rainfall leading to from the gauging station) and accretion further downstream
widespread inundation (Gautam and Kharbuja 2006; Sharma could be observed. However, reach-­w ide erosion and accretion
and Shakya 2006). There is a large possibility that the shifting could also be observed in some years, specifically in the 1990s.
of the river channel in 1993 could have been due to this flood The channel also experienced a prolonged erosion period from
event which is further explored in the next section. The west- 2005 to 2010. After this erosive period, the channel exhibited a
ward migration of the river found in our study aligns well with relatively stabilized condition, characterized by a consistent yet
similar studies in two large rivers of Nepal. Baniya et al. (2023) modest (less than 10 km2) pattern of erosion upstream and ac-
assessed the channel shifting of the Koshi River, where the cretion downstream. The variability and magnitude of both ero-
Bagmati River eventually converges, and found higher rates sion and accretion were also higher during the 1990–2005 time
of shifting in the west (21.6 m/yr) than in the east (0.8 m/yr). frame (Figure 6b,c). The largest accretion occurred in 1993 with
Similarly, Rakhal et al. (2021) also found that the Karnali River an area of around 25 km2, whereas extensive erosion occurred in
in the western region of Nepal also showed a prominent shift- 1992 with a similar extent.
ing toward the west banks. Attribution of these shifting in the
Himalayan forelands to various factors including climate, tec- The first phase of centreline migration was also found to be co-­
tonics, landforms, or human interventions could be an exciting incident with the peak flood in 1993 (Figure 7a). The overlapped
field for further research. active channel masks (Figure 7b), and before and after images

7 of 13
 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FIGURE 6    |    (a) Spatiotemporal mapping of the net change in channel extent from 1990 to 2022, (b) Annual time-­series of net area change, and (c)
Total erosion (red) and accretion (blue) areas.

FIGURE 7    |    (a) Daily flow hydrograph at Padherodovan hydrologic station, (b) accretion and erosion areas during dry (pre-­monsoon) and wet
(monsoon and post-­monsoon) seasons, and (c–f) false colour composite image showing channel planform changes before (top) and after (bottom) the
monsoon season during years 1993, 2002, 2003 and 2004 respectively.

8 of 13 River Research and Applications, 2024

 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
of the 1993 flood event (Figure 7c) clearly portray the westward implies that larger flood events are accompanied by higher
shift of the channel. This westward migration could be primar- sediment load which are deposited in these plain lowlands.
ily attributed to the activation of a small channel, which sub- The erosion extent, however, showed a negative correlation
sequently took precedence in conveying the flow, as evident in with the peak flow. This sheds light on complex erosion pro-
the pre-­2002 image (Figure 7d). In contrast, the second phase cesses occurring in the region that are not explainable through
of centreline migration observed in 2003 and persisting until hydrologic interpretation alone. Furthermore, an analysis of
2022 was not directly linked to the 2002 flood event. Instead, the July 2019 flood event using flood maps generated from
we observed that this shift resulted primarily from the abandon- Sentinel-­1 imagery revealed an intriguing pattern of the flood
ment of the eastern channel and could have been a cumulative closely following the historical channel route (Figure 8c). In
effect of the 2002 flood, followed by the low-­f low period in 2003 the middle section, there was an occurrence of channel over-­
and the subsequent peak flow in 2004. The 2002 flood showed topping, which resulted in water flowing into the migrated
continued activity in the east channel with visible sedimenta- area of the river. This observation suggests that the histori-
tion. However, the 2003 image displayed inactivity in the west cal channel still maintains a significant influence on the flow
channel, which could be further confirmed by the 2004 image, dynamics of floodwaters, despite changes in the surrounding
where accretion was observed exclusively in the left channel. landscape over time.
The deposition of sediment at the channel bifurcation during
the peak flood event could deepen the main channel, impeding
water flow into the branching channel (Kleinhans and van den 3.3   |   Identifying Drivers of Flooding
Berg 2011; Valenza et al. 2020). This effect could have been fur-
ther augmented by the low-­f low conditions experienced in 2003 We computed the relative contribution of multi-­day precipi-
rendering the right branch inactive. tation event (P md), extreme precipitation event (P 99), and API
as a proxy for antecedent wetness conditions in the basin in
To understand if the information on accretion or erosion ex- triggering flooding events. We found that the majority of the
tent is embedded in the hydrologic peak signals, we explored flood events in the Bagmati River were triggered by multi-­day
the correlation between these variables. We found a stronger precipitation (Figure 9a). The relative contribution of multi-­
correlation between annual peak flow and accretion area than day precipitation was found to be 55.4% whereas only a few
with erosion area (Figure 8a,b). The coefficient of determina- events were preceded by extreme precipitation events (12.7%)
tion value (r) was 0.70 for accretion and − 0.35 for erosion. This and antecedent conditions (15.2%) as shown in Figure 9b,c

FIGURE 8    |    The role of monsoon flood in channel extent change. Scatter plots showing a correlation between peak annual flow and (a) accretion
area and (b) erosion area. The blue line represents the regression line while the grey shaded area shows the confidence interval. (c) Flood map for the
2019 July event derived from Sentinel-­1 radar imagery.

9 of 13
 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
FIGURE 9    |    Contribution of different precipitation events occurring before flooding: (a) multi-­day precipitation (Pmd), (b) extreme precipitation
(P 99), and (c) antecedent precipitation index (API). The values in the square bracket indicate the average contribution at all precipitation stations.

respectively. This clearly shows that a single day of heavy 2005, followed by 5 years of erosion-­prone period. The recent
precipitation is unlikely to trigger flooding in the basin and decade showed relative stability with the general trend of ero-
flooding is more likely to occur when there are multiple days sion in the upstream region and accretion in the downstream.
of precipitation. Similarly, the annual peak flow had a strong correlation with
the accretion area, suggesting that high flood events deposit
The results found in our study basin are consistent with the a large amount of sediment in these lowland reaches, how-
study in large Indian basins (Nanditha and Mishra 2022) ever, the erosion processes were not properly reflected in the
where multi-­ day precipitation was found to be the prom- hydrologic signal. Furthermore, the flood extent of an event
inent driver of flooding in the majority of the study basins. in July 2019 showed that the abandoned channel was largely
However, the contributing factors used in our study are only inundated. Additionally, the multi-­day precipitation event was
based on daily precipitation data. While we used the API value found to be the highest contributor of the flooding event in
in lieu of soil moisture observations, accurate soil moisture this study region.
data and other variables such as evapotranspiration as well as
the inclusion of compound events could result in a more ac- The analyses conducted in this study provide a comprehensive
curate portrayal of drivers of flooding in the basin (Berghuijs understanding of the interactions between climate, hydrol-
et al. 2019). ogy, and geomorphology in the region. Multiday precipitation
events trigger extreme floods, which, in turn, drive significant
sediment transport from the upper to lower reaches of river
systems in the region. These sediment dynamics can reshape
4   |   Conclusions
fluvial geomorphology, causing floodplains to elevate, rivers
to shift their course, and channels to alter in size and shape.
We leverage over three decades of Landsat imagery to monitor
Such changes exacerbate flooding hazards, posing risks to
fluvial processes in the lower reach of the Bagmati River in
populations, agricultural productivity, and assets within the
Nepal. We first extracted the active channel mask, defined as
floodplain. Monitoring sediment deposition, erosion patterns,
the extent of both the wetted channel as well as alluvial depos-
and floodplain inundation is thus critical and can be effectively
its, from the annual composite of the Landsat imagery in the
achieved using remote sensing imagery on cloud-­based plat-
GEE cloud computing platform. Annual time-­series of chan-
forms. Integrating such analysis with hydrologic regionaliza-
nel extent revealed a notable decline of about 65% of the riv-
tion techniques (Karki et al. 2023) can be a valuable tool for
er's total area with a pronounced decrease beginning around
understanding geomorphic changes and their hydro-­climatic
2010. The average lateral migration rate of the river gradually
drivers and for monitoring subsequent flooding hazards, even
increases towards downstream reach showing a higher im-
in remote and ungauged basins.
pact on the population residing on the southern vicinity of the
river. Further, channel masks before and after the monsoon
season were also extracted to understand how the monsoon
season affects the channel extent and if the erosion and ac- Data Availability Statement
cretion processes are embedded in the hydrologic peak signal. The data that support the findings of this study are available from the
We found a high incidence of erosion and accretion before corresponding author upon reasonable request.

10 of 13 River Research and Applications, 2024

 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
References Foga, S., P. L. Scaramuzza, S. Guo, et al. 2017. “Cloud Detection
Algorithm Comparison and Validation for Operational Landsat Data
Andermann, C., A. Crave, R. Gloaguen, P. Davy, and S. Bonnet. 2012.
Products.” Remote Sensing of Environment 194: 379–390.
“Connecting Source and Transport: Suspended Sediments in the Nepal
Himalayas.” Earth and Planetary Science Letters 351: 158–170. Gao, S., Z. Li, M. Chen, et al. 2021. “Spatiotemporal Variability of Global
River Extent and the Natural Driving Factors Revealed by Decades of
Asselman, N. E., and H. Middelkoop. 1995. “Floodplain Sedimentation:
Landsat Observations, GRACE Gravimetry Observations, and Land
Quantities, Patterns, and Processes.” Earth Surface Processes and
Surface Model Simulations.” Remote Sensing of Environment 267:
Landforms 20: 481–499.
112725.
Baniya, S., R. Deshar, R. Chauhan, and S. Thakuri. 2023. “Assessment
Gautam, D. K., and R. G. Kharbuja. 2006. “Flood Hazard Mapping of
of Channel Migration of Koshi River in Nepal Using Remote Sensing
Bagmati River in Kathmandu Valley Using Geo-­Informatics Tools.”
and GIS.” Environmental Challenges 11: 100692.
Journal of Hydrology and Meteorology 3: 1–9.
Berghuijs, W. R., S. Harrigan, P. Molnar, L. J. Slater, and J. W. Kirchner.
Gilvear, D., and R. Bryant. 2016. “Analysis of Remotely Sensed Data
2019. “The Relative Importance of Different Flood-­ Generating
for Fluvial Geomorphology and River Science.” In Tools in Fluvial
Mechanisms Across Europe.” Water Resources Research 55: 4582–4593.
Geomorphology, 103–132. UK: John Wiley & Sons, Ltd. https://​doi.​org/​
Bhatpuria, D., K. Matheswaran, T. Piman, T. Tha, and P. Towashiraporn. 10.​1002/​97811​18648​551.​ch6.
2022. “Assessment of Large-­ Scale Seasonal River Morphological
Gorelick, N., M. Hancher, M. Dixon, S. Ilyushchenko, D. Thau, and
Changes in Ayeyarwady River Using Optical Remote Sensing Data.”
R. Moore. 2017. “Google Earth Engine: Planetary-­ Scale Geospatial
Remote Sensing 14: 3393.
Analysis for Everyone.” Remote Sensing of Environment 202: 18–27.
Biedenharn, D. S., C. R. Thorne, and C. C. Watson. 2000. “Recent
Huete, A., K. Didan, T. Miura, E. P. Rodriguez, X. Gao, and L. G. Ferreira.
Morphological Evolution of the Lower Mississippi River.”
2002. “Overview of the Radiometric and Biophysical Performance of the
Geomorphology 34: 227–249.
MODIS Vegetation Indices.” Remote Sensing of Environment 83: 195–213.
Biemans, H., C. Siderius, A. F. Lutz, et al. 2019. “Importance of Snow
Immerzeel, W. W., A. F. Lutz, M. Andrade, et al. 2020. “Importance and
and Glacier Meltwater for Agriculture on the Indo-­Gangetic Plain.”
Vulnerability of the world's Water Towers.” Nature 577: 364–369.
Nature Sustainability 2: 594–601.
Jarriel, T., J. Swartz, and P. Passalacqua. 2021. “Global Rates and
Boix-­Fayos, C., G. Barberá, F. López-­Bermúdez, and V. Castillo. 2007.
Patterns of Channel Migration in River Deltas.” Proceedings of the
“Effects of Check Dams, Reforestation, and Land-­Use Changes on River
National Academy of Sciences 118: e2103178118.
Channel Morphology: Case Study of the Rogativa Catchment (Murcia,
Spain).” Geomorphology 91: 103–123. Kale, V. S. 2003. “Geomorphic Effects of Monsoon Floods on Indian
Rivers.” In Flood Problem and Management in South Asia, 65–84.
Boothroyd, R. J., R. D. Williams, T. B. Hoey, B. Barrett, and O. A. Prasojo.
Dordrecht: Springer. https://​doi.​org/​10.​1007/​978-­​94-­​017-­​0137-­​2 _​3.
2021. “Applications of Google Earth Engine in Fluvial Geomorphology
for Detecting River Channel Change.” Wiley Interdisciplinary Reviews: Karki, N., N. M. Shakya, V. P. Pandey, L. P. Devkota, A. M. S.
Water 8: e21496. Pradhan, and S. Lamichhane. 2023. “Comparative Performance of
Regionalization Methods for Model Parameterization in Ungauged
Bormudoi, A. 2024. “Common Index-­Based Methods of Flood Mapping
Himalayan Watersheds.” Journal of Hydrology: Regional Studies 47:
by Optical Satellite Sensors: A Review of the Merits.”
101359.
Brown, C. F., S. P. Brumby, B. Guzder-­Williams, et al. 2022. “Dynamic
Keesstra, S., J. Van Huissteden, J. Vandenberghe, O. Van Dam, J. De
World, Near Real-­Time Global 10 mLand Use Land Cover Mapping.”
Gier, and I. Pleizier. 2005. “Evolution of the Morphology of the River
Scientific Data 9: 251.
Dragonja (SW Slovenia) due to Land-­Use Changes.” Geomorphology 69:
Bryndal, T., P. Franczak, R. Kroczak, W. Cabaj, and A. Kołodziej. 191–207.
2017. “The Impact of Extreme Rainfall and Flash Floods on the Flood
Khan, N. I., and A. Islam. 2003. “Quantification of Erosion Patterns
Risk Management Process and Geomorphological Changes in Small
in the Brahmaputra–Jamuna River Using Geographical Information
Carpathian Catchments: A Case Study of the Kasiniczanka River (Outer
System and Remote Sensing Techniques.” Hydrological Processes 17:
Carpathians, Poland).” Natural Hazards 88: 95–120.
959–966. Quantification of erosion patterns in the Brahmaputra–
Charlton, R. 2007. Fundamentals of Fluvial Geomorphology. London: Jamuna River using geographical information system and remote sens-
Routledge. ing techniques.
Danegulu, A., S. Karki, P. K. Bhattarai, and V. P. Pandey. 2024. Kleinhans, M. G., and J. H. van den Berg. 2011. “River Channel and Bar
“Characterizing Urban Flooding in the Kathmandu Valley, Nepal: The Patterns Explained and Predicted by an Empirical and a Physics-­Based
Influence of Urbanization and River Encroachment.” Natural Hazards Method.” Earth Surface Processes and Landforms 36: 721–738.
120: 10923–10947. https://​doi.​org/​10.​1007/​s1106​9 -­​024-­​06650​-­​w.
Kodimalar, T., R. Vidhya, and R. Eswar. 2020. “Land Surface Emissivity
Dépret, T., E. Gautier, J. Hooke, D. Grancher, C. Virmoux, and D. Retrieval From Multiple Vegetation Indices: A Comparative Study Over
Brunstein. 2015. “Hydrological Controls on the Morphogenesis of Low-­ India.” Remote Sensing Letters 11, no. 2: 176–185. https://​doi.​org/​10.​
Energy Meanders (Cher River, France).” Journal of Hydrology 531: 1080/​21507​0 4X.​2 019.​1692384.
877–891.
Kondolf, G. M., and H. Piégay. 2016. “Tools in Fluvial Geomorphology:
Dewan, A., R. Corner, A. Saleem, et al. 2017. “Assessing Channel Problem Statement and Recent Practice.” In Tools in Fluvial
Changes of the Ganges Padma River System in Bangladesh Using Geomorphology, 1–12. UK: John Wiley & Sons, Ltd. https://​doi.​org/​10.​
Landsat and Hydrological Data.” Geomorphology 276: 257–279. 1002/​97811​18648​551.​ch1.
Dingle, E., M. Creed, H. Sinclair, et al. 2020. “Dynamic Flood Kong, D., E. M. Latrubesse, C. Miao, and R. Zhou. 2020. “Morphological
Topographies in the Terai Region of Nepal.” Earth Surface Processes and Response of the Lower Yellow River to the Operation of Xiaolangdi
Landforms 45: 3092–3102. Dam, China.” Geomorphology 350: 106931.
Feng, D., C. J. Gleason, X. Yang, G. H. Allen, and T. M. Pavelsky. 2022. Lamichhane, S., N. Karki, V. P. Pandey, P. Joshi, and S. Dawadi. 2024.
“How Have Global River Widths Changed Over Time?” Water Resources “A Synergic Approach Using the Model and Remote Sensing Data for
Research 58: e2021WR031712. Flood Monitoring in Under-­Observed Transboundary Rivers.” Journal

11 of 13
 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
of Hydroinformatics 26, no. 10: 2489–2505. https://​doi.​org/​10.​2166/​ Rasul, G. 2014. “Food, Water, and Energy Security in South Asia:
hydro.​2 024.​056. A Nexus Perspective From the Hindu Kush Himalayan Region☆.”
Environmental Science & Policy 39: 35–48.
Lamichhane, S., and N. M. Shakya. 2019. “Alteration of Groundwater
Recharge Areas due to Land Use/Cover Change in Kathmandu Valley, Rouse, J. W., R. H. Haas, J. A. Schell, and D. W. Deering. 1974.
Nepal.” Journal of Hydrology: Regional Studies 26: 100635. “Monitoring Vegetation Systems in the Great Plains With Earth
Resources Technology Satellite (ERTS).” NASA Special Publication 351:
Langhorst, T., and T. Pavelsky. 2023. “Global Observations of Riverbank
309.
Erosion and Accretion From Landsat Imagery.” Journal of Geophysical
Research: Earth Surface 128: e2022JF006774. Rowland, J. C., E. Shelef, P. A. Pope, et al. 2016. “A Morphology
Independent Methodology for Quantifying Plan View River Changes
Lebel, L., P. Garden, and P. Lebel. 2011. “Managing Floods and Scarcity
and Characteristics From Remotely Sensed Imagery.” Remote Sensing of
in a Monsoon Climate.” In Integrating Science and Policy. London:
Environment 184: 212–228.
Routledge.
Rusnák, M., and M. Lehotskỳ. 2014. “Time-­ Focused Investigation
Lee, J.-­S. 1981. “Refined Filtering of Image Noise Using Local Statistics.”
of River Channel Morphological Changes due to Extreme Floods.”
Computer Graphics and Image Processing 15: 380–389.
Zeitschrift für Geomorphologie 58: 251–266.
Marchi, L., M. Borga, E. Preciso, and E. Gaume. 2010. “Characterisation
Schwenk, J., and J. Hariharan. 2021. “RivGraph: Automatic Extraction
of Selected Extreme Flash Floods in Europe and Implications for Flood
and Analysis of River and Delta Channel Network Topology.” Journal of
Risk Management.” Journal of Hydrology 394: 118–133.
Open-­Source Software 6: 2952. https://​doi.​org/​10.​21105/​​joss.​02952​.
Markert, K. N., A. M. Markert, T. Mayer, et al. 2020. “Comparing
Scott, C. A., F. Zhang, A. Mukherji, W. Immerzeel, D. Mustafa, and L.
Sentinel-­1 Surface Water Mapping Algorithms and Radiometric Terrain
Bharati. 2019. “Water in the Hindu Kush Himalaya.” In The Hindu Kush
Correction Processing in Southeast Asia Utilizing Google Earth
Himalaya Assessment: Mountains, Climate Change, Sustainability, and
Engine.” Remote Sensing 12: 2469.
People, 257–299. Cham: Springer International Publishing.
Masek, J. G., M. A. Wulder, B. Markham, et al. 2020. “Landsat 9:
Shamim, S. 2024. More Than 200 Killed: What Caused the Devastating
Empowering Open Science and Applications Through Continuity.”
Floods in Nepal? Doha: Al Jazeera. https://​w ww.​aljaz​eera.​com/​news/​
Remote Sensing of Environment 248: 111968.
2024/​10/2/​what-­​cause​d-­​the-­​devas​tatin​g-­​f lood​s-­​in-­​nepal​.
McFeeters, S. K. 1996. “The Use of the Normalized Difference Water
Sharma, R. H., and R. Awal. 2013. “Hydropower development in Nepal.”
Index (NDWI) in the Delineation of Open Water Features.” International
Renewable and Sustainable Energy Reviews 21: 684–693. https://​doi.​org/​
Journal of Remote Sensing 17: 1425–1432.
10.​1016/j.​rser.​2 013.​01.​013.
Mirzaee, S., S. Yousefi, S. Keesstra, H. R. Pourghasemi, A. Cerdà, and Sharma, R. H., and N. M. Shakya. 2006. “Hydrological Changes and Its
I. C. Fuller. 2018. “Effects of Hydrological Events on Morphological Impact on Water Resources of Bagmati Watershed, Nepal.” Journal of
Evolution of a Fluvial System.” Journal of Hydrology 563: 33–42. Hydrology 327: 315–322.
Monegaglia, F., G. Zolezzi, I. Güneralp, A. J. Henshaw, and M. Tubino. Shukla, U. K., P. Srivastava, and I. Singh. 2012. “Migration of the Ganga
2018. “Automated Extraction of Meandering River Morpho Dynamics River and Development of Cliffs in the Varanasi Region, India During the
From Multitemporal Remotely Sensed Data.” Environmental Modelling Late Quaternary: Role of Active Tectonics.” Geomorphology 171: 101–113.
& Software 105: 171–186.
Sinha, R., A. Gupta, K. Mishra, et al. 2019. “Basin-­Scale Hydrology and
Morin, G. P., J. Lavé, C. France-­Lanord, T. Rigaudier, A. P. Gajurel, and Sediment Dynamics of the Koshi River in the Himalayan Foreland.”
R. Sinha. 2018. “Annual Sediment Transport Dynamics in the Narayani Journal of Hydrology 570: 156–166.
Basin, Central Nepal: Assessing the Impacts of Erosion Processes in
the Annual Sediment Budget.” Journal of Geophysical Research: Earth Sinha, R., K. Sripriyanka, V. Jain, and M. Mukul. 2014. “Avulsion
Surface 123: 2341–2376. Threshold and Planform Dynamics of the Koshi River in North Bihar
(India) and Nepal: A GIS Framework.” Geomorphology 216: 157–170.
Nanditha, J., and V. Mishra. 2022. “Multiday Precipitation Is a
Prominent Driver of Floods in Indian River Basins.” Water Resources Slater, L. J., M. B. Singer, and J. W. Kirchner. 2015. “Hydrologic Versus
Research 58: e2022WR032723. Geomorphic Drivers of Trends in Flood Hazard.” Geophysical Research
Letters 42: 370–376.
Nyberg, B., G. Henstra, R. L. Gawthorpe, R. Ravnås, and J. Ahokas.
2023. “Global Scale Analysis on the Extent of River Channel Belts.” Spada, D., P. Molinari, W. Bertoldi, A. Vitti, and G. Zolezzi. 2018. “Multi-­
Nature Communications 14: 2163. Temporal Image Analysis for Fluvial Morphological Characterization
With Application to Albanian Rivers.” ISPRS International Journal of
Otsu, N. 1979. “A Threshold Selection Method From Gray-­ Level Geo-­Information 7: 314.
Histograms.” IEEE Transactions on Systems, Man, and Cybernetics 9: 62–66.
Struck, M., C. Andermann, N. Hovius, et al. 2015. “Monsoonal
Pickens, A. H., M. C. Hansen, M. Hancher, et al. 2020. “Mapping and Hillslope Processes Determine Grain Size-­ Specific Suspended
Sampling to Characterize Global Inland Water Dynamics From 1999 to Sediment Fluxes in a Trans-­H imalayan River.” Geophysical Research
2018 With Full Landsat Time-­Series.” Remote Sensing of Environment Letters 42: 2302–2308.
243: 111792.
Valenza, J., D. Edmonds, T. Hwang, and S. Roy. 2020. “Downstream
Pietroń, J., J. Jarsjö, A. O. Romanchenko, and S. R. Chalov. 2015. “Model Changes in River Avulsion Style Are Related to Channel Morphology.”
Analyses of the Contribution of In-­Channel Processes to Sediment Nature Communications 11: 2116.
Concentration Hysteresis Loops.” Journal of Hydrology 527: 576–589.
Van Appledorn, M., M. E. Baker, and A. J. Miller. 2019. “River-­Valley
Qazi, N. Q., and S. P. Rai. 2018. “Spatio-­Temporal Dynamics of Sediment Morphology, Basin Size, and Flow-­Event Magnitude Interact to Produce
Transport in Lesser Himalayan Catchments, India.” Hydrological Wide Variation in Flooding Dynamics.” Ecosphere 10, no. 1: e02546.
Sciences Journal 63: 50–62. https://​doi.​org/​10.​1002/​ecs2.​2546.
Rakhal, B., T. R. Adhikari, S. Sharma, and G. R. Ghimire. 2021. Wang, Y., J. Ma, X. Xiao, X. Wang, S. Dai, and B. Zhao. 2019. “Long-­
“Assessment of Channel Shifting of Karnali Mega Fan in Nepal Using Term Dynamic of Poyang Lake Surface Water: A Mapping Work Based
Remote Sensing and GIS.” Annals of GIS 27: 177–188. on the Google Earth Engine Cloud Platform.” Remote Sensing 11: 313.

12 of 13 River Research and Applications, 2024

 15351467, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/rra.4413 by Nischal Karki - University Of Alabama-Tuscaloosa , Wiley Online Library on [26/12/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Wu, X., N. Bi, J. Xu, et al. 2017. “Stepwise Morphological Evolution of
the Active Yellow River (Huang He) Delta Lobe (1976–2013): Dominant
Roles of Riverine Discharge and Sediment Grain Size.” Geomorphology
292: 115–127.
Wulder, M. A., D. P. Roy, V. C. Radeloff, et al. 2022. “Fifty Years of
Landsat Science and Impacts.” Remote Sensing of Environment 280:
113195.
Xie, Q., J. Yang, and T. S. Lundström. 2021. “Sediment and Morphological
Changes Along Yangtze river's 500 Km Between Datong and Xuliujing
Before and After Three Gorges Dam Commissioning.” Scientific Reports
11: 13662.
Xu, H. 2006. “Modification of Normalised Difference Water Index
(NDWI) to Enhance Open Water Features in Remotely Sensed Imagery.”
International Journal of Remote Sensing 27: 3025–3033.
Yadav, A., R. J. Boothroyd, G. H. Sambrook Smith, and S. Sen. 2023.
“Morphological Adjustments of the Yamuna River in the Himalayan
Foothills in Response to Natural and Anthropogenic Stresses.”
Hydrological Processes 37: e14934.
Yamazaki, D., D. Ikeshima, J. Sosa, P. D. Bates, G. H. Allen, and T. M.
Pavelsky. 2019. “MERIT Hydro: A High-­Resolution Global Hydrography
Map Based on Latest Topography Dataset.” Water Resources Research
55: 5053–5073.
Yamazaki, D., D. Ikeshima, R. Tawatari, et al. 2017. “A High-­Accuracy
Map of Global Terrain Elevations.” Geophysical Research Letters 44:
5844–5853.
Yousefi, S., S. Mirzaee, S. Keesstra, et al. 2018. “Effects of an Extreme
Flood on River Morphology (Case Study: Karoon River, Iran).”
Geomorphology 304: 30–39.
Zou, Z., X. Xiao, J. Dong, et al. 2018. “Divergent Trends of Open-­Surface
Water Body Area in the Contiguous United States From 1984 to 2016.”
Proceedings of the National Academy of Sciences 115: 3810–3815.

13 of 13

View publication stats