Identifying Most Suitable Priority Areas for Soil-Water Conservation Using

Coupling Mechanism in Guwahati Urban Watershed

Ishita Afreen Ahmed a , Swapan Talukdar a, Mohd Waseem Naikoo a , Shahfahad a, Ayesha
Parvez b , Swades Pal c , Shakeel Ahmed d , Atiqur Rahman a, Abu Reza Md Towfiqul Islam e ,
Amir Mosavi f

a Department of Geography, Faculty of Natural Sciences, Jamia Millia Islamia, New Delhi, 110025, India
b Henry Samueli School of Engineering University of California, Irvine 5200 Engineering Service Rd, Irvine, CA, 92617, USA
c Department of Geography, University of Gour Banga, Mokdumpur, Malda, West Bengal, 732103, India
d School of Sciences Maulana Azad National Urdu University (MANUU), Hyderabad, 500032, India
e Department of Disaster Management, Begum Rokeya University, Rangpur, 5400, Bangladesh
f Faculty of Informatics, Obuda University, Budapest, Hungary

Abstract

The urban watershed of Guwahati, Assam, has witnessed a rapid urbanization in recent years, making it to one of the
most deteriorated cities in terms of environmental issues. Therefore, this work focused on finding the most suitable
soil-water conservation areas at a micro-scale using coupling coordination mechanism. Principal component analysis
(PCA) was used to create a priority model for sub-watershed basin based on nineteen morphometric parameters. We
then quantified the soil erosion using the revised universal soil loss equation (RUSLE) for current and future
scenarios based on the representative concentration pathway (RCP) 2.6 and 8.5 models (RCP2.6 and 8.5). In
addition, we proposed the fuzzy logic and analytical hierarchy process (AHP) model-based soil-water conservation
suitability (SWPC) model for current and future scenarios. Finally, the most suitable soil-water conservation areas
for current and future scenarios were identified using coupling coordination degree model (CCD). To the author’s
knowledge, this is the first in-depth study that identifies the most suitable conservation areas by analyzing the
watershed prioritization, soil erosion, and topographic, hydrologic, land cover, and climatic parameters-based soil-
water conservation suitability models. Sub-watersheds comprising Silsako, Bharalu, Deepor Beel, and North
Guwahati have been identified as high-priority sub-watersheds. According to the current RUSLE model, soil erosion
in the study area varies from 140 to 181.64 tonnes per hectare per year. In contrast, soil erosion would continue to
increase in the future as per the RCP8.5 model, which varies from 305 to 332 tonnes per hectare per year. The
current SWPC model predicted 46.92 km2 area as high and 54.40 km2 area as very high suitable zones. However,
under the RCP 2.6 and RCP 8.5 models, the high and very high SWPC zones will experience extended areas in the
future due to increased soil erosion intensity. According to the CCDM results, Bharalu, Deepor Beel, and North-
Guwahati sub-watersheds have observed a very high to medium coupling degrees, which are considered the most
suitable areas for conservation. The findings of this study will significantly help stakeholders and experts in long-
term land-water resource management and effectively address environmental degradation issues in urban watersheds
around the world.
Keywords: Watershed priority zone; artificial intelligence; soft computing; soil erosion; soil conservation; revised
universal soil loss equation; coupling mechanism; fuzzy logic
1. Introduction

River erosion, land use change, and incompetent management contribute to the loss of tons of soil per year,
causing permanent damage in agricultural and mountain areas (Kulimushi et al., 2021, Kebede et al., 2021). Many
countries have developed river basin management plans that prioritize watersheds based on their drainage and
ecological importance (Patle et al., 2020; Arefin et al., 2020; Rodrigues et al., 2022). The drainage system of an area
and the characteristics of the watershed are crucial to the long-term development, management and restoration of
soil-water process, as they allow accurate estimation of water flow, soil loss, and management alternatives (Naqvi et
al., 2019; Hembram and Saha, 2020). Recently, numerous studies have used remote sensing (RS) coupled with
geographic information systems (GIS) to reliably calculate morphological features and analyze basin attributes
(Resmi et al., 2019; Bogale, 2021). In recent years, experts and researchers worldwide have developed methods for
automatic extraction of drainage networks and their properties utilizing remote sensing and GIS (Arubalaji and
Padmalal, 2020; Hibi et al., 2021; Ahad et al., 2022). Furthermore, several scholars have reported about the
application of GIS for morphometric analysis with interaction of drainage morphometric features of various
watersheds of India (Kumar et al., 2021: Shekar et al., 2022). Therefore, RS can be considered a valuable technique
for morphometric analysis, and prioritizing watersheds, which are crucial indicators to adopt any different relevant
conservation actions. However, prioritization of watershed and landscape planning are often considered for
developing and implementing robust management and restoration strategies (Sharda et al., 2021). Prioritization is
key to identifying areas that need attention (Yu et al., 2021; Abdeta et al., 2020). Due to time and cost constraints, it
is challenging to undertake restoration and soil and water conservation work at all sites simultaneously in a
watershed management program. Therefore, watershed prioritization for robust soil-water conservation can be a
feasible approach to solve the mentioned issues (Zhang et al., 2021; Meshram et al., 2022). Although soil erosion
(SE) can occur naturally, human activities such as land use change, agriculture, livestock grazing, and deforestation
have worsened erosion and soil deterioration (Gong et al., 2022). As a result, SE is seen as a natural and human-
caused issue with severe economic and environmental consequences in many cultures. Despite the significant
impacts of SE on soil sustainability, there is a scarcity of data on existing and future situations. The lack of this data
is because of the complexity of erosion processes, making SE estimates costly, time-consuming, and challenging
(Cunha at al., 2022; Ferreira et al., 2022). Because of this difficulty, several models and techniques have been
developed to simplify SE modelling and better our knowledge of the pattern and processes of SE. Previous
researchers have found that RUSLE model outperformed other models in terms of accuracy and simplicity to
estimate the potential soil erosion (Mengie at al., 2022; Somasiri et al., 2022). Therefore, in the present study, we
used the RUSLE model to estimate the potential soil erosion for current and future scenarios.
However, previous researchers have developed future soil erosion scenarios by considering the R factor
(erosive factor) for current and future times (Pal et al., 2022). They have computed R factor from future rainfall
datasets. The Intergovernmental Panel on Climate Change (IPCC) has released the Representative Concentration
Pathways (RCPs), which comprises four future scenarios for earth’s greenhouse gas (GHG) emission (2.6, 4.5, 6,
and 8.5), for future climatic variables. The RCP2.6 model simulates low GHG emissions, whereas the RCP4.5 and
RCP6 models represent stability scenarios, and the RCP8.5 model simulates high GHG emissions. Using several
CMIP5-RCP scenarios, research has been undertaken to forecast the influence of future climate on soil erosion;
however, relatively few studies have been conducted on the Indian subcontinent (Choudhury et al., 2022; Raj et al.,
2022). In this research, we considered minimum and maximum RCP scenarios to compute the R factor for future
rainfall scenarios. The soil erosion estimation using RUSLE model is very common, not the future soil erosion
models, but the research on the conservation strategies for reducing soil erosion and its impact is very rare. Also, it
is very challenging to identify the areas to be conserved and monitored for soil erosion based on limited resources.
However, to do so, several parameters related to soil erosion are required under the framework of multi-criteria
decision making (MCDM) process. Therefore, this research aimed to predict robust suitable soil-water conservation
zone, not only for current situation, but also for future soil erosion scenarios, under the MCDM framework. Previous
researchers have already applied several MCDM techniques to find the flood, landslide susceptible zones, wetland
habitat zone, urban surface ecological health condition zones (Rehman et al., 2022; Craciun et al., 2022). Few
popular MCDM techniques are analytical hierarchy process (AHP) (Kumar et al., 2022; Roshani et al., 2022),
analytical network process (ANP) (Dahri et al., 2022; Abedini et al., 2022), weights of evidence (WOE) (Bopche &
Rege, 2022; Behera & Panigrahi, 2022), evidential belief function (EBF) (Ramesh & Iqbal, 2020; Zhao et al., 2022)
etc. However, the MCDM techniques have suffered from biasness due to subjective weights, time consuming and
slow computational process. To prevent these drawbacks, the fuzzy logic model, a semi machine learning algorithm,
has been emerged and frequently used for solving several environmental issues (Aghaloo and Chiu, 2020;
Baharvand et al., 2020; Jha et al., 2020; Noori et al., 2019). Because of its capacity to handle linguistic data
efficiently, Vema et al. (2019) employed a fuzzy inference system (FIS) for the selection of sites concerning the
facilities related to harvesting of water (e.g., ponds, check dams, and tanks). Similarly, Kumar and Anbalagan
(2015) and Ramesh and Anbazhagan (2015) identified landslide prone regions using landslide frequency ratio and
fuzzy logic based on GIS. According to the conclusions of these investigations, the fuzzy logic technique exceeds
the other methods in terms of accuracy. Literature review also validates fuzzy-based work for groundwater as well
as land suitability analysis (Bikdeli, 2020; Moonjun et al., 2020; Tafreshi et al., 2018). Therefore, based on previous
literature, we used advanced fuzzy logic and conventional AHP model to prepared SWPC model for current and
future scenarios. Furthermore, the association between watershed prioritizing and soil-water appropriateness is a
new and emerging idea (Melaku et al., 2018). The current study links these two systems to comprehend the
influence of physical and drainage features on soil and water resources in a location. Coupling coordination theory
can explain the degree to which two or more subsystems interact. The degree of coupling describes the strength of
the contact between two subsystems, whereas the degree of coordination indicates the intensity of cooperative
growth between them (Lai et al., 2020). We selected the coupling coordination theory because of its capacity to
provide a complete assessment system and its intuitiveness, but it has not been frequently employed in empirical
applications. Only a few researchers have recently applied the coupling coordination theory to investigate the
coordination between two or more parameters (Ye et al., 2021; He et al., 2021; Dong et al., 2021). Yuan et al. (2021)
for example, investigated the coordination among the regional environment, economy, and tourism in western
Hunan Province, China, using a coordinated development model. In addition, Yang et al. (2021) employed the
CCDM (coupling coordination model) and GIS spatial methodologies to examine the correlation, coupling, and
coordination degrees between the geo-ecological environment and urbanization in Chongqing
Municipality.Guwahati Metropolitan Area is one of the most urbanized areas in India, on the bank of the
Brahmaputra River. Therefore, because of the urbanization, this river has experienced flow of domestic waste or
sewages and industrial debris from the city, such as hazardous, chemical, industrial solid, and solid waste generated
by municipalities (Pondari et al., 2020; Tsering et al., 2020). Little work has been done so far in the urban watershed
of Guwahati, focusing on soil-water conservation modelling. Therefore, to fill the research gaps, this study proposed
a robust way to find the soil-water conservation zones by considering PCA-NDVI based watershed prioritization,
RUSLE based soil erosion and fuzzy logic and AHP based-soil water potential conservation (SWPC) zones under
current and future scenarios at RCP2.6 and 8.5. The novelty of the present study is to identify the highly accurate
areas to be conserved or monitored in order to reduce the area coverage and impact of soil erosion for current and
future scenarios through a coupling coordination mechanism in the Guwahati watershed To the author’s knowledge,
this has been the first study that considered different aspects for mitigating soil erosion by identifying the
conservation areas. In this study, we provide an alternate method for watershed prioritization in the six sub-
watersheds of the Guwahati Metropolitan Area, Assam, such as Deepor Beel, Bharalu, Silsako, North Guwahati,
Foreshore, and Kalmoni, by leveraging factors with strong influence (as determined by PCA) on soil degradation in
each of these micro-watersheds. Furthermore, for the first time, the coupling coordination evaluation model (CCD)
has been used to examine the degree of coupling and coordination among watershed prioritized model, soil erosion
model, and soil-water suitability model. The present research will provide scientific support and insights for
mitigating the severity of soil erosion by identifying accurate conservation areas for current and future scenarios
(2040). The main objectives of the present study are:
1. to provide an alternative method for watershed prioritization by effectively utilizing factors that have a
strong influence (as determined by PCA) on soil degradation in the urban watershed of Guwahati,
2. to estimate soil erosion using RUSLE model for current and future scenarios (2040) at RCP2.6 and 8.5,
3. to develop the sub-watershed zones-wise soil-water conservation suitability model (SWPC) in the study
area for current and future scenarios (2040) at RCP2.6 and 8.5,
4. to identify the highly suitable soil-water conservation areas through the coupling degree and degree of
coordination among the priority, soil erosion and SWPC models for current and future scenarios (2040) at
RCP2.6 and 8.5.
2. Material and Methodology

2.1 Study Area

The urban watershed of Guwahati was chosen as the study area as it is the fastest-growing city in Northeast
India. Guwahati is located in Assam's Kamrup (Metro) district. Guwahati's urban watershed, known as the Guwahati
Metropolitan Region, includes the Guwahati Municipal Corporation (GMC) area, the North Guwahati Town
Committee (NGTC) area, the Amingaon Census Town (ACT), and 21 revenue villages (GMA). It has a total area of
340.67 km2. Guwahati's overall population has grown from 500,000 in 1991 to 8, 14,575 in 2008, and 963,429 in
2011 (Census of India, 2001 & 2011). The study area is on an undulating plain with elevations ranging from 50 to
55 meters above mean sea level (MSL). Guwahati's urban watershed is further expanded in an east-west ward trend,
spanning 45 kilometres between Narengi in the east and Lokpriya Gopinath Bordoloi (LGB) International Airport in
the west, as well as both sides of the Brahmaputra River in the north and the slopes of the Shillong Plateau in the
south. The urban watershed of Guwahati is divided into six sub-watersheds: Bharalu, Silsako, Kalmoni, Deepor
Beel, Foreshore, and North Guwahati (Fig. 1). In recent years, all the basins being part of the primary drainage route
have been undergoing increasing urbanization. As a result, they have become the most degraded regions due to
encroachment and dumping of solid and liquid wastes into these watersheds. Thus, watershed prioritization of these
six micro-watersheds will considerably alleviate the problem of soil degradation and water management in the given
area.

Figure 1 Location of the study area

2.2 Database

The morphometric analysis of the urban watershed of Guwahati was conducted using the 1:50,000 scale
topographical maps collected from the Survey of India (SOI). Landsat 8 (OLI/TIRS) image acquired from the
website of the United States Geological Survey (USGS). The analysis comprises delineating the study area into six
sub-watersheds and then calculating nineteen morphometric parameters based on each watershed's areal, linear, and
relief aspects (e.g., stream number, stream length ratio, elongation ratio, ruggedness number, bifurcation ratio,
drainage density, etc.). In addition, the drainage network of the urban watershed of Guwahati was evaluated using
Horton's (1945) and Strahler's (1945, 1964) methodologies. SoI topographic maps were georeferenced in ArcGIS
desktop version 10.4 using the WGS84 datum and UTM zone 46 N. The SRTM digital elevation map (DEM) was
downloaded from earth explorer of USGS. The rainfall map was collected from meteorological stations of IMD for
the periods of 2000-2018. The road data was collected from DIVA GIS.

2.3 Method for morphometric analysis

A detailed description of the methodology is provided in the supplementary material and supplementary table
1.

2.3.1 Methods for watershed prioritization

The areal, linear, and relief morphometric parameters were designated as the significant aspects for prioritizing
sub-watersheds (Kulimushi et al., 2021; Kale and Deshmukh, 2020). Soil erosion of an area directly relates to linear
(e.g., stream number, stream order, etc.), as well as relief parameters (e.g., ruggedness number, relief ratio, etc.). As
a result, higher ranks are assigned to a higher value of linear and relief factors. Areal parameters (e.g., basin area,
drainage texture, elongation ratio) have opposite relationship with soil erodibility (Hema et al., 2021). Areal
parameters with lower values have a more significant influence on the degradation of the soil of a watershed. Hence,
the ranking of the erodibility of soil was computed through assigning priority ranks. In this research, we used three
methods to prepare the watershed priority model i.e., compound factor, PCA, and NDVI. The theoretical
background of these methods has been given below:

Compound Factor

For the final watershed prioritization map, three basic weighted sum ranking systems were used: compound
parameter (CP), priority rank (NP), and priority degree (GP) (Chauhan et al. 2016; Prabhakar et al. 2019). At first,
CP for each sub-watershed was determined through computing the mean of all morphological parameters. The
second phase involves assigning NP. The sub-watershed with the lowest compound parameter was given the highest
NP, and so on. Furthermore, the priority degree was calculated using eq. (1), (2), and (3), and they were categorized
into high, medium, and low priority ranks.

𝐶𝑃𝑀𝐴𝑋 −𝐶𝑃𝑀𝐼𝑁
𝐺𝑃𝐻𝑖𝑔ℎ = [𝐶𝑃𝑀𝐼𝑁 , 𝐶𝑃𝑀𝐼𝑁 + ( )] [1]
3

𝐶𝑃𝑀𝐴𝑋 −𝐶𝑃𝑀𝐼𝑁 𝐶𝑃𝑀𝐴𝑋 −𝐶𝑃𝑀𝐼𝑁

𝐺𝑃𝑀𝑒𝑑𝑖𝑢𝑚 = [𝐶𝑃𝑀𝐼𝑁 + ( ) , 𝐶𝑃𝑀𝐼𝑁 ± 2 ∗ ( )] [2]
3 3

𝐶𝑃𝑀𝐴𝑋 −𝐶𝑃𝑀𝐼𝑁
𝐺𝑃𝐿𝑜𝑤 = [𝐶𝑃𝑀𝐼𝑁 + 2 ∗ ( ) , 𝐶𝑃𝑀𝐴𝑋 ] [3]
3
 In the same way, the NDVI was calculated, and its values were classified according to the methods proposed
by Gómez-Almonte (2005) and Merg et al. (2011) (Supplementary Table 2).

Principal Component Analysis

The PCA was primarily used in this study to reduce data and represent fundamental data properties with
only a few principal components (Pourghasemi et al., 2021; Mohamed and Worku, 2021; Perez and Reynoso, 2021).
Using PCA, we computed the correlation matrix, the principal component loading matrix, and the respective
Eigenvalues (Singh et al. 2021; Patle et al., 2020; Meshram and Sharma, 2017)) based on nineteen morphometric
parameters. Finally, we used PCA to assess the structural correlations of the morphometric characteristics and
classify the basins based on component values.

Normalized difference vegetation index

Vegetation indices, which utilize the transformation of electromagnetic spectrum (EMS) reflected from the
earth's surface to satellite sensors, are the most prevalent and widely studied remote sensing products. For example,
the normalized differential vegetation index (NDVI) measures the difference between near-infrared and red bands to
quantify vegetation (Tucker, 1979).

2.3.2 Method for current and future soil-water conservation model

In this research, to compute the current and future soil-water conservation model, we prepared several components,
such as current and future rainfall, soil erosion model using RUSLE, fuzzy logic based SWPC model, and final
robust SWPC model using CCD mechanism. The methods for these components have been presented below
sections:

2.3.3 Method for estimation of current and future rainfall

The circulation patterns of the atmospheric and ocean currents have changed because of global warming. The
geographical and temporal patterns of precipitation and temperature throughout the planet have altered because of
this. As a result, rising groundwater depletion is linked to global warming (IPCC, 2011). Increased flooding risks in
particular landscapes, which are entirely unfavorable for groundwater recharge, may be directly connected to a
change in the pattern of extreme rainfall events (IPCC, 2011). Researchers have attempted to create these
connections in a stochastic manner. Therefore, future climates may be forecasted using GCMs in conjunction with
expected changes in land use to predict future SWPC models. To consider climate change in the research field, we
extracted precipitation data of RCP 2.6 and 8.5 climate scenarios from CCSM-4 of the 4th IPCC Assessment Report
based on AR5. These data are freely available from the NCAR GIS Initiative Climate Change Scenario portal (https:
/gisclimatechange.ucar.edu) in vector format (Shapefile). Representative pathways of concentration (RPCs) show
the concentrations of greenhouse gases (GHGs) in the atmosphere and the routes that may be taken to get there. The
GHG emission will be 2.6 W/m2 at RCP2.6 by 2100, followed by 4.5 W/m2 at RCP4.5, 6.0 W/m2 at RCP 6 and 8.5
W/m2 at RCP8.5. In the present study, we produced rainfall data for the period of 2020-2040 based on minimum and
maximum concentration of greenhouse in atmosphere (RCP2.6 and RCP8.5). In the present study, we prepared R-
factor based on the rainfall obtained from RCP 2.6 and RCP 8.5 for the period of 2020-2040.

2.3.4 Method for potential soil erosion estimation for current and future scenario using RUSLE

This research uses the RUSLE to estimate the soil erosion in an urban watershed. It needs significant input
variables, which can be readily obtained from accessible data sources. RUSLE measures the long-term erosion
process of rills and inter-rills over a wide range of agricultural and forest catchments (Ostovari et al., 2021; Biswas
and Pani 2015). It is modified version of USLE model, which has issues of sensitivity (Renard et al., 1997). The
RUSLE model has widely been used for its accurate estimation of average annual soil erosion. RUSLE estimates the
average annual soil loss in an area using equation 5.

𝑆𝐸 = 𝑅 × 𝐾 × 𝐿𝑆 × 𝐶 × 𝑃 [5]

SE is the projected soil loss (t ha-1 year-1) in this equation, and R is the rainfall-erosivity index (MJ mm ha−1
h−1 year−1). K is the predicted soil erodibility factor (t h ha MJ−1 ha−1 mm−1). C is the land use-specific cover
management component; P is the conservation support practice element that includes slope length and slope
steepness. For a detailed explanation of the parameters used for RUSLE modeling, readers can follow Behera et al.
(2020) and Teng et al. (2018).

2.3.5 Soil-water conservation suitability zone modeling for current and future scenario

To identify robust soil-water conservation areas, topographic, hydrologic, climatic, land cover, and artificial
structure-related parameters are required. The application of fuzzy logic and AHP to create SWPC model has not
been done so far in any study, so the approach used in this paper is novel. Physical as well as drainage parameters
are strongly connected to soil and water conservational activities. DEM, Aspect, Slope, LULC, Topographic
Wetness Index (TWI), drainage density (Dd), length of overland flow (Lg), basin relief (Bh), ruggedness number
(Rn), and rainfall are all standard criteria for assessing soil-water suitability. The affluence, as well as the influence
of these parameters, varies depending on the land characteristics. Also, the future conservation areas can be
modelled using fuzzy logic as well as an analytical hierarchy process (AHP) under the scenario of climate change in
terms of rainfall (2020-2040).

Fuzzy logic based SWPC modelling

The SWPC model was prepared by integrating eleven parameters using fuzzy logic and AHP methods. To
implement fuzzy logic, the 'Membership function' and 'Fuzzy overlay' analysis were implemented in ArcGIS 10.4
environment. There are two important aspects related to selecting a membership function (MF): the midpoint and
spread. Initially, it is possible to choose from several membership functions, such as linear and non-linear fuzzy MF,
including large, small, Gaussian, and linear.

Non-linear membership function "MsSmall" (large membership values are assigned to a small input raster
value), as well as "MsLarge" (small membership values are set to an enormous input raster value), were used to
assess soil-water suitability for all the nine parameters. Furthermore, five distinct combinations of fuzzy operators
are described in the literature (fuzzy OR, fuzzy AND, fuzzy Product, fuzzy Sum, and fuzzy Gamma). The fuzzy
Gamma operator was used in this study because it is determined by all input layers' maximum and minimum
membership values (Ki and Ray, 2014; Tafreshi et al., 2018).

Analytical Hierarchical Process (AHP) based modeling

Multi-criteria decision analysis, such as the Analytical Hierarchical Process (AHP), is a well-known method
that analyzes different environmental parameters and involves ranking based on certain criteria to solve a very
complex decision-making process (Saaty and Vargas, 2001). It is the most common method for delineating potential
zones of various natural phenomena (Rahaman et al., 2022). The hierarchy of the parameters of SWPC was
determined by its interdependence and interplay with other parameters, and the hierarchy was built from the top
down. In the decision-making process, this strategy employs subjective and objective criteria (Baig et al., 2020). It
also uses a pairwise comparison of a distinct parameter affecting the study area's soil and water. The pairwise
comparison was done with the assistance of a parameter matrix and assigning weight to the parameter concerning
their effect on other parameters stated in a numerical scale and consistency ratio (Saaty, 1980). Each parameter's
weights were assigned using Saaty's scale (1-9) of relative relevance value.

Estimation of important variables using Random Forest

To calculate the importance of each soil erosion parameter for defining soil loss in the study area, the
methodology of random forests (a classifier that combines many single decision trees) is used. In this study, we have
used the two-variable importance measures of the RF algorithm: mean decrease in accuracy (MDA) and mean
decrease in Gini (MDG). The Gini index measures node impurity; the mean decrease in Gini indicates an
improvement in the splitting criterion, which measures the reduction in class impurity from partitioning the data set
(Myles et al., 2004). At the same time, the mean decrease in accuracy is a permutation-based measure of variable
importance derived from evaluating a variable's contribution to prediction accuracy. The mean decrease in Gini also
measures of variable significance by permuting the values of each soil erosion controlling factor in the out-of-bag
sample. The factors with the highest mean Gini importance in the model were primarily responsible for flood
prediction in the study area.

2.3.6 Method for identification of precise soil-water conservation areas using coupling coordination mechanism

The coupling coordination degree model has been used to compute the agreement between two sub-systems
through interaction and physical process. Therefore, in this research, the watershed prioritization, potential soil
erosion, and soil-water conservation potential areas have been integrated using CCD model to find the agreement
areas of all models. The application of CCD for identifying suitable model by considering three different system is
totally new concept, also provides quite accurate results. The CCDM is derived using equation 5.

𝐶 = √((𝑆𝑊 × 𝑊𝑃)/(𝑆𝑊 + 𝑊𝑃)2 ) -----------------------Eq. (5)

 Where SW is the comprehensive index of soil-water conservation level, WP is the watershed prioritization
index, and C is the coupling degree value of the coupled system of soil-water conservation and watershed
prioritization. The CCDM degree value ranges from 0≤ C ≤ 1, higher the value, more precise the soil-water
conservation suitable zones and vice versa (Supplementary Table 3). The high priority watershed, areas with higher
soil erosion rates, and areas with higher chances to be conserved potentiality can have the high value of coupling
degree and coupling coordination degree. These areas should be monitored closely for implementing management
plans. In this research, several sophisticated methods have been applied to identify different aspects for soil erosion
conservation modeling. Therefore, to explain the methodology in a bird-eye-view, a hierarchical methodology chart
has been given (Figure 2).
 Figure 2 Hierarchical methodological framework of the study

3. Result
3.1 Watershed prioritization modeling
In the present study, the watershed prioritization has been done based on the parameters of areal aspect, linear
aspect, and relief aspect. The parameters of these three morphometric aspects have been integrated at ArcGIS
platform. The details of these aspects and computation of watershed priority model is given below:

3.1.1 Areal aspect of sub-watersheds

Morphometric analysis of the sub-watershed areal aspects reveals a dendritic drainage pattern strongly affected
by the terrain of the study area. The urban watershed of Guwahati is divided into six sub-watersheds. The total
drainage area of the watershed is 339.25 km2, while the perimeter is 922.16 km2, and the basin length is 42.81 km2,
respectively. Further, the total length (Lu) of the urban watershed is 316.58 km. Deepor Beel (98.14 km2) is the
largest sub-watershed of the 6, while Kalmoni was the smallest (8.75 km2). The highest and lowest perimeters
among the sub-watersheds were Deepor Beel, and North Guwahati are 90.40 km and 25.52 km, respectively.
Supplementary Fig. 1 (a), (b) & (c) depicts the variations in the basin area, perimeter and length. The basin length
corresponds to the total area of the drainage basin, with a maximum length of 10.50 km observed in Deepor Beel,
followed by Foreshore (Supplementary Table 4). Dd ranges from 0.84 km/km2 in North Guwahati to 1.43 km/km2 in
the Kalmoni sub-watershed in the study area, indicating abundant vegetation and permeable surface material
(Supplementary Table 4). The study shows that higher Fs values are observed in the southern part of the
Brahmaputra River (Supplementary Fig. 1 e), primarily in sub-watersheds Kalmoni (3.89) and Foreshore (2.33)
(Supplementary Table 4). As per the result, the textures of all the sub-watersheds are classified as very coarse,
indicating the presence of vegetation-protected massive and resistant rocks in the study area resulting in low soil
degradation (Supplementary Fig. 1 f). The Lg value varies from 0.59 to 0.35, which indicates that Lg moves over
longer distance before joining the respective channel in the study area. Such a situation arises primarily due to a
gentle slope gradient, abundant vegetation, more infiltration, and reduced runoff (Supplementary Fig. 1g). The Ff
ranges from 2.87 to 0.17 for all watersheds (Supplementary Table 4). The sub-watershed of Bharalu (2.87) and
Silsako (1.48) have high Ff. These values represent a rounded basin quickly, followed by high peak flow resulting in
soil degradation condition (Supplementary Table 4). The Re value ranges from 0.47 to 1.11, with moderate to
slightly steep slope, Kalmoni and Foreshore have the lowest Re, which indicate that the basin is elongated and has
less chance of erosion. In contrast, Deepor Beel, Bharalu, Silsako, and North Guwahati have high Re values
indicating a circular basin and will be more prone to flooding (Supplementary Fig. 1i).

3.1.2 Linear aspect of sub-watersheds

The study area is a fourth-order river based on Strahler's hierarchical ranking (Strahler, 1964). Silsako
watershed is the only fourth-order (Supplementary Table 5), while most sub-watersheds are categorized into three
orders. Out of the total stream in the watersheds, the 1st streams account for 59.4 percent (243 Nu), while the 2nd
order accounts for 24.2 percent (99 Nu), the third-order accounts for 13.6 percent (56 Nu), and the fourth-order
account for 2.68 percent (11 Nu) (Supplementary Fig.3a). Similarly, to stream number, first-order streams have
accounted for 64.22 percent (212.8 km) of total length (Supplementary Fig. 2b). Supplementary table 5 clearly
shows more information about the sub-watershed’s characteristics (e.g., area, perimeter, length, stream length, etc.).
The Lsm of the sub-watersheds in the urban watershed of Guwahati ranges from 0.27 to 1.15. Supplementary Table 5
clearly shows that the Lsm of a given stream order is greater than the lower order but less than the next higher order.
The Rl in the urban watershed of Guwahati is highest for the third-order and second-order streams. The highest Rl is
from second to third-order streams (73.64) in the Kalmoni sub-watershed (Supplementary Table 5). A high Rl
indicates an increased susceptibility to soil erosion in the study area. In the present study, the sub-watersheds of the
urban watershed of Guwahati have high Rb values ranging from 0.72 to 5.4, indicating early hydrograph peaks, flash
flood potential, and soil erosion susceptibility.

3.1.3 Relief aspects of sub-watersheds

The highest and the lowest maximum height of basin mouth (Z) is observed in the sub-watershed of Silsako
(352 meters) and Kalmoni (166 meters), respectively. While the highest and lowest minimum height of basin mouth
(z) is observed in Bharalu (0.00 meters), Silsako (0.00 meters), and Deepor Beel (-10.00 meters) above mean sea
level, respectively (Supplementary Table 6). The more the basin mouth rises, the more vulnerable the area becomes
to soil erosion. Supplementary Fig. 3 a & b shows the variation in relief across the sub-watersheds. The highest
relief is observed in the sub-watershed of Silsako (352 meters), while the lowest is observed in Kalmoni (166
meters). Rh values in the urban watershed range from 28.17 to 83.33 (Supplementary Table 6). In the study area,
sub-watersheds with high relief features and steep slope gradients have high Rh values and more susceptibility to soil
degradation, while regions with low relief and a gentle to uniform slope have low Rh (Supplementary Fig. 3d). The
Rn of the watersheds is shown in table11. Rn in this study ranged from 0.24 to 0.39. The ruggedness values were
highest in the Silsako sub-watershed and lowest in the Kalmoni sub-watersheds (Supplementary Fig. 3e). It suggests
that the Silsako sub-watershed (0.39) is more vulnerable to soil erosion than the others (Asfaw and Workineh, 2019;
Tiwari and Kushwaha, 2021).

3.1.4 Computation of watershed prioritization

Supplementary Table 7 summarizes the Pearson correlation coefficient (r) between all the morphometric
parameters. According to the findings, 19 parameters had very high positive correlations (with r greater than 0.70).
In contrast, we discovered 17 negative correlations with a high significance level (at 0.05 level of significance
level). In addition, it is observed that 14 parameters that have no significant correlation with each other (below
0.10). Furthermore, the analysis discovered that seven morphometric parameters, have significant correlations, four
are associated with areal aspects (drainage density, drainage texture, stream frequency, length of overland flow), and
three are associated with relief aspects (ruggedness number, basin relief, and height of basin mouth). Although these
correlations permit for recognition of high and effective relationships amongst the parameters, they do not yet
qualify for the grouping of specific parameters into components and the assignment of physical significance. PCA
was used in the present study because there were many correlations among these morphometric parameters. PCA
helped determine the morphometric parameters with a more significant influence on watershed behavior.
Supplementary Table 8 depicts the various morphometric parameters analyzed by PCA methods that influence the
soil erosion in the study area. According to the results, the first five principal components (PC01, PC02, PC03,
PC04, and PC05) had a total variance of 100%. But alone, PC01, PC02, and PC03 account for 88% of the total
variance (Supplementary Table 9). The result showed a high correlation of PC01, PC02 and PC03 with the
parameters. The analysis also shows that PC01 is associated with linear parameters e.g. Nu (0.744 & 0.655), Ls
(0.748 & 0.611), Lsm (0.854 & 0.085) and Fs (0.747 & 0.655). At the same time, PC02 has a high correlation with
basin perimeter, Dd, and Dt, and moderately with Ls, basin area, and basin length, indicating that these parameters are
associated with the continuous soil erosion process of the watershed over time. Also, it was observed that PC03 has
a suitable correlation with Re and Ff, along with a minimum height of basin mouth (Supplementary Table 8). The
PCA determined that basin perimeter, Dd, Fs, Dt, Lg, Ff, Re, and Rn were relevant variables to assess the watershed
prioritization of the study area (Table 1). These variables were chosen due to their highest values correlated with the
three principal components.

Table 1 Watershed prioritization based on morphometric parameter

Sub- Basin Dd Fs Dt Lg Ff Re Rn CP NP GP
watersheds perimeter
Deepor Beel 1 3 6 4 3 4 4 2 3.375 3 Medium
Bharalu 3 2 5 3 4 1 1 3 2.75 1 High
Foreshore 2 5 2 6 2 5 5 5 4 5 Low
Silsako 4 4 3 1 5 2 2 1 2.75 2 High
Kalmoni 5 1 1 5 6 6 6 6 4.5 6 Low
North 6 6 4 2 1 3 3 4 3.625 4 Medium
Guwahati
Where, Dd drainage density, Fs stream frequency, Dt drainage texture, Lg length of overland flow, Re elongation ratio, Rn Ruggedness number,
CP compound parameter, NP priority rank and GP priority degree
 Table 1 shows the sub-watersheds CP, NP, and GP based on basin perimeter, Dd, Fs, Dt, Lg, Ff, Re, and Rn.
Bharalu, Silsako, Deepor Beel, and North Guwahati have the highest and medium priority ranks. On the other hand,
the priority rank of sub-watershed Foreshore increases due to the urban areas' presence and scarcity of vegetation
cover (Supplementary Fig. 4). Supplementary table 10 shows the priority index of NDVI-estimated vegetation
cover. The degradation of the sub-watersheds caused by concentrated flow channels and downslope flow of water is
described by NDVI. Based on vegetation NDVI, the highest priority is allotted to North Guwahati and Foreshore
due to the lowest vegetation cover in these sub-watersheds. However, the rank of Bharalu and North Guwahati has
decreased when seen in the result of ranking based on morphometric parameters than that of NDVI ranking. The
NDVI indicates that the vegetation cover in these sub-watersheds is at its peak. Finally, the watershed's final
prioritization (GPf) demonstrates that sub-watershed Silsako and Bharalu have very high susceptibility to soil
degradation with NPf 1 and 2, respectively. In contrast, Deepor Beel and North Guwahati have a high susceptibility
to soil erosion with NPf 3 and 4 (Table 2). Furthermore, table 2 shows the rank of priority and final priority for sub-
watersheds as a component of morphometric parameters and NDVI values. The final priority (GPf) analysis clearly
states that the highest priority comprises Silsako and Bharalu, followed by Deepor Beel and North Guwahati (Fig.
3). The priority rank clearly shows the degradation condition of the sub-watersheds. The slopes are steeper in the
sub-watersheds with high-priority ranks, and the profiles constitute deeper channels. With the descent of the slope
degree of the sub-watersheds, the steepness, depth of the channel profile, and erodibility of the areas also reduce.

Table 2. Watershed prioritization based on final priority index

Sub-watersheds CPf NPf GPf

Deepor Beel 3.4375 4 Medium
Bharalu 3.25 2 High
Foreshore 3.75 5 Low
Silsako 3.125 1 High
Kalmoni 4.125 6 Low
North Guwahati 3.3125 3 Medium
Where, CP compound parameter, NP priority rank, GP priority degree, f mean morphometry, NDVI and final
 Figure 3 Watershed prioritization map of the urban watershed of Guwahati

3.2 Estimation of potential soil erosion with future insight

In the present study, we estimated potential soil erosion using the RUSLE model for the current scenario and
under the climate change scenario for 2040. Therefore, to do so, we prepared future climatic change-related
parameters, like rainfall for 2020-2040, under RCP2.6 and RCP8.5. Figure 4 shows the annual average rainfall data
for the current time and 2040 for RCP2.6 and RCP8.5. The maximum rainfall for the present scenario (2000-2018)
is 186.7 mm, and the minimum is 156 mm. but, maximum and minimum rainfall will increase by 42% and 27% in
future (2040) if we consider both methods (RCP2.6 and RCP8.5) based on the minimum and maximum
concentration of greenhouse gases (Figure 4b and 4c).

Fig. 4 Rainfall Modelling (a) Present rainfall (b) Rainfall 2.6, and (c) Rainfall 8.5

In the present study, we employed the RUSLE model to estimate soil erosion in the present and future
scenarios using five factors (K, C, P, LS, and R) (Figure 5). Figure 5a shows that the K factors vary from 0.02 to
0.212 tons Ha MJ−1mm−1 across the watershed. High soil erodibility indicates the higher potentiality of soil erosion
and vice versa. Also, the sub-watersheds like Silsako, Bharalu, North Guwahati, and Foreshore show very high to
high K factors, as these regions have sand-dominated soil texture. On the other hand, Deepor Beel has very low K
values, indicating lower soil erosion sensitivity. In the case of the LS factor, the value ranges from -0.61 to 46.23.
High values indicate higher chances of eroding and vice versa (Figure 5b). The value of the P factor ranges from 0.5
to 1, while low values imply higher sensitivity to be eroded (Figure 5c). The value of the C factor ranges from 0.01
to 0.35, where a low value indicates a higher potentiality to be eroded by rainfall and runoff. Most of the Silsako and
Bharalu sub-watershed areas have been covered by low values of the C factor, which indicates that these regions are
highly susceptible to soil erosion. The present study computed the R factor for current and future scenarios (2040)
using RCP2.6 and RCP8.5. The R factor of the current scenario ranges from 25.84 to 34.49 MJ mm/ha/year. Higher
R factor or rainfall erosivity can have higher erosional ability and vice versa. Very high and high values of the R
factor have been observed in the Silsako, Bharalu, and Foreshore. Similarly, the R factor for 2040 (RCP8.5) will be
increased at 42 MJ mm/ha/year, which can cause severe damage to the soil.

Figure 5 RUSLE parameters (a) K factor (b) LS factor (c) C factor (d) P factor (e) Present R factor (f) R factor at
RCP2.6 (2040), and (g) R factor at RCP8.5 (2040)
 The RUSLE models were created after integrating all the parameters in the raster calculator of ArcGIS (version
10.8) software. The final RUSLE models have been generated for three time periods to estimate soil erosion. The
current RUSLE model shows that the soil erosion of the study area ranges from 140 to 181.64 tonnes hectare per
year. The RUSLE model for 2040 (RCP8.5) estimates the soil erosion ranges from 305 to 332 tonnes hectare per
year. While, the RUSLE model for 2042.0 (RCP2.6) estimates the soil erosion ranges from 267 to 302 tonnes
hectare per year (Figure 6). Therefore, the models show that the soil erosion will be increased manifold in the future.
After deriving the RUSLE soil erosion models, we classified the models using Jenk's natural break algorithms into
five classes: very high, high, moderate, low, and very low. Very high and high soil erosion zones have been
identified in Silsako, Bharalu, North Guwahati, and Foreshore. While very low soil erosion zone has identified in
Kalmoni sub-watershed and some parts of Deepor Beel.

Figure 6 Estimation of soil erosion using RUSLE Models for (a) current scenario (b) future scenario (2040)
(RCP2.6), and (c) future scenario (2040) (RCP8.5)

3.3 Proposing potential soil-water conservation models for current and future scenarios
Estimating and identifying the degree of soil erosion is insufficient and is just for quantification and
visualization. As a result, the high priority region should be designated to avoid soil erosion. Hence, we also
proposed a fuzzy logic and AHP-based model for detecting potential soil-water conservation zones for the present
and the future (by 2040) under a climate change scenario to minimize future soil erosion. Topographical,
hydrological, and land cover-related parameters were used to create the SWPC model (Supplementary Figures 5 and
6). Since each parameter has a distinct dimension and direction, all parameters were normalized before modelling.
As a result, among various membership functions, we used two fuzzy membership functions in this study: MsLarge
(Large considers large values of the parameters, would have a low degree of membership with the destination (soil
conservation), and MsSmall (small considers small value of the parameters, will have a high degree of membership
with destination). After that, we used fuzzy operator, like GAMMA 0.9 to integrate all the fuzzified parameters to
create fuzzy logic based SWPC model. The value of SWPC model ranges from 0 to 1, where close to 1 indicates
highly suitable areas for conservation and vice versa. Then, we used Jenkin’s natural break algorithm to classify the
SWPC model into five classes, such as very high, high, moderate, low, and very low soil conservation zones.

Figure 7 shows the locations of the sites with the highest soil-water suitability using fuzzy logic and AHP.
According to the fuzzy logic present SWPC, most of the study area is in a very poor soil-water conservation
potential zone. Approximately 125.82 km2 of the total SWPC area is in the moderate suitability zone, followed by
the low suitability zone (73.51km2). At the same time, high (46.92 km2) and very high (54.40 km2) suitability zones
cover a sizable area (Figure. 7a). In RCP 2.6 (129.98 km2) and RCP 8.5 (128.99 km2), the area under the moderate
SWPC increased (Fig. 7b-c). In terms of sub-watersheds, the three SWPC maps show that Deepor Beel, Kalmoni,
and North-Guwahati have the lowest potential for soil-water conservation, followed by Kalmoni and North-
Guwahati. On the other hand, the highest SWPC is found in the Bharalu and Silsako sub-watersheds, with nearly
80% of the area falling into the high and very high SWPC zones. Also, the three maps show that around 41% of the
total area (Figure 7a-c) is suitable for moderate SWPC, while 33% of the watershed is compatible with high and
very high SWC sites. When the final fuzzy models were compared to the watershed prioritization map (Fig. 3), it
was obvious that the SWPC maps roughly corresponded to the prioritization map. For example, the sub-watersheds
of Bharalu and Silsako had high priority values on the watershed prioritization map (Fig. 3), corresponding to higher
values on the soil-water conservation maps (Fig. 7a). To compare the findings of fuzzy logic, we applied MCDM
technique (AHP) to prepare SWPC models for current to future periods. We used twelve parameters to compute
pairwise comparison matrix of AHP for generating weights of each parameter. We obtained weights for all
parameters, such as drainage density (0.85), ruggedness number 0.75), basin relief (0.4), Lg (0.25), LULC (0.2),
TWI (0.18), elevation (0.15) and so on. The CR value for this matrix was 0.08. Then, the weights have been
assigned to the parameters in raster calculator to produce the SWPC models under current and future scenarios. The
generated SWPC model for current scenario shows 81.2km2 as low suitability zone, and 49km2 as very high SWPC
zones (Figure 7d). Similarly, the AHP based SWPC models under RCP2.6 and 8.5 scenarios showed quite similar
results like fuzzy logic based SWPC models (Figure 7e & f).
 Figure 7 Soil-water suitability zones for (a, d) Present SWPC (b, e) SWPC in 2040 (RCP2.6), and (c, f) SWPC in
2040 (RCP8.5) using fuzzy logic and AHP

The variable importance for SWPC model needs to be assessed because the management plans should be
carried out based on the important variables. Therefore, this study used random forest (RF) based sensitivity
technique to compute the variable importance to the model (Figure 8). In the present study, we used this sensitivity
technique on the fuzzy logic based SWPC model because we consider it as base model of this study. The figure
shows that drainage density has highest influence to the model, followed by ruggedness number, aspect, TWI, basin
relief, and elevation. Length of overland flow has lowest influence to the output, followed by LULC, and NDVI.

Figure 8 Computation of importance variable for current SWPC model using RF based MDA and MDG technique

3.4 Coupling coordination degree mechanism for identifying accurate soil-water conservation zone
 Figures 9(a-c) and 10(a-c) depict the degree of coupling coordination between watershed prioritization, soil-
water conservation and soil loss in the urban watershed of Guwahati. It identified the agreement areas among three
sub-systems based on coupling interaction. Therefore, the agreement areas have been quantified in terms of coupling
degree and coupling coordination degree model. Higher the agreement indicates high value of three sub-systems,
which reflects the high priority zones and vice versa. In general, the coupling coordination degrees in the region are
classified into very poor, poor, medium, high, and very high. Figure 9a shows that the high agreement value for
coupling degree model under present scenario has been observed in the sub-watersheds of Bharalu and North-
Guwahati. In comparison, the Deepor Beel and Kalmoni sub-watersheds have very low coupling degrees. However,
RCP 2.6 projections for 2040 show a moderate coupling degree for Deepor Beel and a very poor coupling degree for
Silsako (Figure 9b). Furthermore, the high agreement areas for coupling degree model under RCP 8.5 scenario has
observed a slight difference from the mentioned two models (Figure 9c). Foreshore and North Guwahati have the
highest coupling degree, as shown in Fig. 9c. The coupling models clearly show that the maximum extent of the
urban watershed has very high to moderate coupling. The coupling degree for all the sub-watershed has varying
agreement values in all three models (Figure 9a-c). Although, a very poor coupling degree is primarily observed in
the sub-watershed of Deepor Beel and Kalmoni, indicating a serious imbalance between watershed prioritization,
soil-water suitability and soil loss.

Figure 9 Coupling degree between watershed priority, SWPC and RUSLE model for (a) present scenario (b) future
scenario at RCP 2.6 (2040), and (c) future scenario at RCP 8.5 (2040) of the urban watershed of Guwahati

Further analysis of the coordination degree is required to comprehend the coordinated relationship between
subsystems. As a result, we calculated the coordination degrees of the six sub-watersheds based on the present RCP
2.6 and RCP 8.5 and classified the coordination degree values into ten classes, followed by Liu et al., (2018) (Figure
10a-c). The CCD value ranges from 0 to 1, suggesting the lowest to the highest level of coordination. Also, in the
present coordination degree model, the highest coordination ( 0.8 ≤ C ≤ 1) is observed in three sub-watersheds:
Silsako, North Guwahati, and some parts of Bharalu. Due to increased soil loss, these sub-watersheds also have high
prioritization and are primarily suitable for soil-water conservation activities (Figure 9a-c). On the other hand, most
parts of the Deepor Beel sub-watershed have the lowest coordination degree (0 ≤ C ≤ 0.2) in all three maps,
indicating a severe imbalance in the study area between soil-water suitability, soil loss, and watershed prioritization.
Figure 9a also depicts that the coordination of the variables stretched a higher degree to the northern, northeastern
and central portions of the watershed due to extreme urbanization. Due to the existing vegetation and water
resources, the southwestern region of the urban watershed exhibits a lower degree of coordination (0 ≤ C ≤ 0.3).
Therefore, the areas with high coupling and coordination degrees should be monitored continuously. Also,
appropriate soil-water conservation management plans should be proposed and implemented to mitigate soil erosion.

Figure 10 CCD model between watershed priority, SWPC and RUSLE model (a) Present (b) RCP 2.6 (2040), and
(c) RCP 8.5 (2040) of the urban watershed of Guwahati

4. Discussion

In the present study, we intended to build a highly robust model of potential soil-water conservation, which can
provide precise soil-water conservation areas. To do so, we first prioritized the sub-watersheds through advanced
statistical techniques, then measured soil erosion using the RUSLE model for current and future scenarios based on
RCP2.6 and RCP8.5, then proposed a fuzzy logic-based potential soil-water conservation zone. Finally, we
introduced coupling coordination mechanism among watershed priority, soil erosion model, and potential soil-water
conservation model for present and future scenarios to identify precise areas for soil-water conservation (Fig. 2).
According to the author's knowledge, this is the first comprehensive analysis considering different dimensions and
integrating all aspects through standard methods to identify the soil-water conservation areas. In this study, six sub-
watersheds in the urban watershed of Guwahati, Assam, were delineated to prioritize watersheds based on
morphometric parameters and NDVI values. The assessment of potential soil erosion vulnerability necessitates the
consideration of several drainages and relief parameters. Morphometric properties significantly influence runoff and
infiltration capacity (Pathare and Pathare, 2021; Hembram and Saha, 2020). The urban watershed of Guwahati
covers a total area of 339.25 km2. It has a perimeter of 42.81 km2 km and is textured with 409 streams. First-order
streams are approximately 243, accounting for 59.4 per cent of all streams, indicating the catchment's erosional risk.
Nineteen morphometric indices (linear, areal, and relief parameters) (Supplementary Table 1) were encompassed to
prioritize six sub-watersheds based on erodibility risk. Furthermore, the findings of this study state that the higher
values of Fs are mostly found in the southern sub-watersheds of the Brahmaputra River, primarily in the sub-
watersheds Kalmoni (3.89) and Foreshore (2.33), indicating soils prone to erosion and degradation (Supplementary
Table 4). The Nilachal hills on the sub-watershed of Foreshore have massive landslides, soil erosion, and urban flash
floods (Hussain and Goswami, 2016). Due to increased urbanization, dense vegetative cover, primarily in the hills of
the Guwahati Metropolitan Area (GMA), has degraded at an alarming rate (Borthakur and Nath 2012; Patowary and
Sarma, 2018). However, the linear aspects of an area directly correlate with soil erosion, with a higher value
indicating greater erodibility. Similarly, the sub-watersheds of Guwahati's urban watershed have high Rb values
ranging from 0.72 to 5.4. These high Rb values are connected to high peak flows, the potentiality of flash floods
(Siddiqui et al., 2020), and susceptibility to soil erosion, all of which are influenced by geomorphological, as well as
lithological conditions such as deformation forms, tectonism, as well as the mass movements (Kamberis et al.,
2012). In the present study, PCA was very helpful in categorizing variables with physical relevance and
differentiating morphometric aspects with little influence. The result shows that the first principal component
(PC01) is linked to the linear parameters. In contrast, the second (PC02) is related to the watershed's areal aspects,
and the third component (PC03) is directly correlated with the relief parameters. The values of PC01 and PC02 for
the sub-watersheds suggest that soil erosion, as well as run-off, can be associated with linear features of
morphometric study (e.g., Nu (0.744 & 0.655), Ls (0.748 & 0.611), Lsm (0.854 & 0.085), and Fs (0.747 & 0.655). Our
result aligns with what Kamberis et al. (2012) found; components like soil characteristics, LULC, and geology are
highly influenced by parameters such as Dd and Fs. Additionally, using these morphometric parameters combined
with NDVI values improves the quality of watershed prioritization results. The morphometric parameters and the
NDVI-estimated vegetation cover result are shown in supplementary table 10. The degradation of a watershed
caused by concentrated water in channels and Lg on hillslopes is described by morphometry and NDVI, respectively.
These sub-watersheds are very highly urbanized and often face scarcity of vegetation cover, leading to soil erosion
susceptibility (Fig. 3). Hence, proper management of these sub-watersheds is necessary to safeguard against soil and
water degradation in these areas. The present study results are also likely to help the local government and decision-
makers identify the priority of the sub-watersheds that require prompt actions on soil degradation, management, and
conservation practices in the study region.

Furthermore, based on the combined input fuzzy membership values, three SWPC maps are created based on
present and future scenarios at RCP 2.6 and RCP 8.5. The final output shows the locations of the sites with the
highest soil-water conservation suitability (Fig.7a-c). According to the three models, approximately 41% of the total
area (Figure 7a-c) is suitable for moderate SWPC. In contrast, 33% of the watershed is compatible with high and
very high SWC sites. According to the present SWPC, most of the study area is in a low soil-water conservation
potential zone. The moderate suitability zone covers approximately 125.82 km 2 of the total SWPC area, followed by
the low suitability zone (73.51km2). Simultaneously, high (46.92 km2) and very high (54.40 km2) suitability zones
also cover a large area (Fig. 7a). Moreover, the area covered by the moderate SWPC increased in RCP 2.6 (129.98
km2) and RCP 8.5 (128.99 km2). Deepor Beel, Kalmoni, and North-Guwahati have the lowest potential for soil-
water conservation, according to the three SWPC maps, followed by Kalmoni and North-Guwahati, due to the
existing vegetational resources and low urban development (Nath et al., 2021). The coupling maps also depicted that
the SWPC maps roughly corresponded to the soil erosion and sub-watershed prioritization maps. The sub-watershed
of Bharalu and Silsako had high priority values, which resembled higher values on the soil-water conservation maps
and the soil loss models. The results suggest that the sub-watershed with the maximum likelihood of soil erosion is
the zones that need the most attention in soil-water conservation practices. Moreover, the three-coupling
coordination degree model (CCDM) as per the present, and RCP 2.6 and RCP 8.5 indicates that the coupling degree
is most significant in the Bharalu and North-Guwahati sub-watersheds. Deepor Beel and Kalmoni sub-watersheds,
on the other hand, have very low coupling degrees. On the other hand, RCP 2.6 projections for 2040 show a
moderate coupling degree for Deepor Beel and a very poor coupling degree for Silsako. In addition, the coupling
degree for RCP 8.5 has changed and depicts a high coupling degree in Foreshore and North Guwahati (Figure 8a-c).
Deepor Beel, on the other hand, has the lowest coordination degree, indicating a severe imbalance in the study area
between soil-water suitability, soil loss, and watershed prioritization. On the other hand, Deepor Beel's lowest
coordination degree suggests a severe imbalance between soil-water suitability, soil loss and watershed prioritization
in the study area. Figure 9a-c clearly shows that due to extreme urbanization, the coordination of the variables
extended to a greater extent to the northern, northeastern, and central parts of the Guwahati urban watershed (Pawe
and Saikia, 2017). On the other hand, a lower degree of coordination is observed in the southwestern region of the
urban watershed. The findings reveal several notable characteristics in the analyzed relationships. Also, for future
development, there is a need to promote the management of sub-watersheds and their soil-water resources at the
sub-watershed level.

5. Conclusion

In this study an attempt has been made to identify precise and mathematically robust areas for potential soil-
water conservation through coupling mechanisms in the urban watershed of Guwahati, Assam. To accomplish this,
we used advanced statistical approaches to select the sub-watersheds, and then evaluated soil erosion using the
RUSLE model for present and future scenarios based on RCP2.6 and RCP8.5, and lastly proposed a fuzzy logic-
based perspective soil-water conservation zone. Finally, we established a coupling-coordination mechanism for
present and future scenarios between watershed priority, soil erosion model, and potential soil-water conservation
model to identify exact sites for soil-water conservation. The result shows that the sub-watersheds Silsako and
Bharalu is highly susceptible to soil erosion, followed by Deepor Beel and North Guwahati. This is because these
sub-watersheds are highly urbanized which led to decrease in vegetation cover, making the sub-watersheds
vulnerable to soil erosion. Thereafter, we estimated the soil erosion using RUSLE model for current (2020) and
future scenario i.e. 2040 (RCP2.6 and RCP8.5). According to present situation by applying RUSLE model, soil
erosion in the study area varies from 140 to 181.64 tonnes per hectare per year but by 2040, the study predicts soil
erosion to a tune of 267 to 302 and 305 to 332 tonnes per hectare per year by using RCP2.6 and RCP8.5
respectively. As a result, it is seen that that soil erosion will fast increase in future. Then, we proposed fuzzy logic-
based soil-water conservation suitability zones based on topographic, hydrologic, land cover, and climatic variables
for current and future scenarios (2040) at RCP2.6 and RCP8.5. The result signifies high soil-water potential zones in
the sub-watersheds of Silsako, Bharalu, and North Guwahati. Whereas, Deepor Beel sub-watershed shows least
potentiality for conservation due to high urban development in the region in recent decades. Finally, we identified
the precise areas that need to be conserved for soil-water erosion for current and future scenarios (2040) at RCP2.6
and RCP8.5 based on coupling mechanism with the help of watershed priority model, soil erosion models, and soil-
water conservation suitability models. Moreover, when the final SWPC and soil erosion models for current and
future scenarios were compared to the watershed priority map with the CCDM model, extremely accurate
conservation areas for current and future scenarios were determined. Lastly, the degree of coupling and coordination
reflected the intensity of cooperative development in the study area. Appropriate measures to reduce soil erosion and
activities are required in these sub-watersheds to safeguard the remaining fertile land. To the author's knowledge,
this is the first comprehensive research that considers several dimensions and incorporates all elements using
standard approaches to identify soil-water conservation zones. This research proposes a novel strategy for
identifying sites with high-priority soil-water conservation in urban watersheds. The outcome of this may widen
research areas as a reference for future studies on assessing soil erosion in urban watersheds with similar geographic
conditions. The outcome may help decision-makers to prepare sustainable urban development plans and
policies. Although this research provides several scientific insights, it has some limitations too, which need to be
done in the future research. This research used moderate 30 m spatial resolution satellite imageries, but by using
higher resolution images like 10 m Sentinel-2 data we may get micro level soil erosion zones. Further, the sensitivity
analysis may be improved by employing deep learning-based sensitivity and uncertainty analysis.

Acknowledgement

The first author is thankful to the University Grants Commission for providing Maulana Azad National Fellowship
(MANF) for Minorities for Doctoral research. In addition, the authors are indebted to the Survey of India for giving
the Toposheet from which maps of the study area were obtained, and the Indian Meteorological Department,
Regional Meteorological, Guwahati, Assam for providing climatic data of the study area.

Conflict of Interest

The authors declare that they have no conflict of interest.

Data availability Statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author
on reasonable request.

References

Abdeta, G. C., Tesemma, A. B., Tura, A. L., & Atlabachew, G. H. (2020). Morphometric analysis for prioritizing
sub-watersheds and management planning and practices in Gidabo Basin, Southern Rift Valley of Ethiopia.
Applied Water Science, 10(7), 1–15. https://doi.org/10.1007/S13201-020-01239-7/FIGURES/6

Abedini, A., Aram, F., Khalili, A., & Mirzaei, E. (2022). Recognition and Evaluating the Indicators of Urban
Resilient by Using the Network Analysis Process. Urban Science, 6(2), 31.

Achyuthan, H. (2019). Quantitative analysis of the drainage and morphometric characteristics of the Palar River
basin, Southern Peninsular India; using bAd calculator (bearing azimuth and drainage) and GIS.
https://doi.org/10.1080/24749508.2018.1563750

Aghaloo, K., & Chiu, Y. R. (2020). Identifying Optimal Sites for a Rainwater-Harvesting Agricultural Scheme in
Iran Using the Best-Worst Method and Fuzzy Logic in a GIS-Based Decision Support System. Water 2020, Vol.
12, Page 1913, 12(7), 1913. https://doi.org/10.3390/W12071913

Ahad, U., Shah, A. R., & Ali, U. (2022). Quantitative estimation of drainage characteristics of the Pohru Catchment,
Kashmir valley, India: a remote sensing and GIS based approach. https://doi.org/10.1080/10106049.2022.2082559

Arefin, R., Mohirul, M., Mohir, I., &Alam, · Jahangir. (2020). Watershed prioritization for soil and water
conservation aspect using GIS and remote sensing: PCA-based approach at northern elevated tract Bangladesh.
Applied Water Science, 10, 91. https://doi.org/10.1007/s13201-020-1176-5

Arulbalaji, P., & Padmalal, D. (2020). Sub-watershed Prioritization Based on Drainage Morphometric Analysis: A
Case Study of Cauvery River Basin in South India. Journal of the Geological Society of India 2020 95:1, 95(1),
25–35. https://doi.org/10.1007/S12594-020-1383-6

Asfaw, D., & Workineh, G. (2019). Quantitative analysis of morphometry on Ribb and Gumara watersheds:
Implications for soil and water conservation. International Soil and Water Conservation Research, 7(2), 150–157.
https://doi.org/10.1016/J.ISWCR.2019.02.003

Baharvand, S., Rahnamarad, J., Soori, S., &Saadatkhah, N. (2020). Landslide susceptibility zoning in a catchment of
Zagros Mountains using fuzzy logic and GIS. Environmental Earth Sciences, 79(10), 1–10.
https://doi.org/10.1007/S12665-020-08957-W/FIGURES/7

Baig, M. R. I., Shahfahad, Ahmad, I. A., Tayyab, M., Asgher, M. S., & Rahman, A. (2021). Coastal Vulnerability
Mapping by Integrating Geospatial Techniques and Analytical Hierarchy Process (AHP) along the
Vishakhapatnam Coastal Tract, Andhra Pradesh, India. Journal of the Indian Society of Remote Sensing, 49(2),
215–231. https://doi.org/10.1007/S12524-020-01204-6
Behera, M., Sena, D. R., Mandal, U., Kashyap, P. S., & Dash, S. S. (2020). Integrated GIS-based RUSLE approach
for quantification of potential soil erosion under future climate change scenarios. Environmental Monitoring and
Assessment, 192(11), 1–18. https://doi.org/10.1007/S10661-020-08688-2/FIGURES/10

Behera, S., & Panigrahi, M. K. (2022). Gold prospectivity mapping and exploration targeting in Hutti-Maski schist
belt, India: Synergistic application of Weights-of-Evidence (WOE), Fuzzy Logic (FL) and hybrid (WOE-FL)
models. Journal of Geochemical Exploration, 235, 106963. https://doi.org/10.1016/J.GEXPLO.2022.106963

Bikdeli, S. (2020). Redevelopment modeling for land suitability evaluation of the suburb brown-fields using fuzzy
logic and GIS, northeastern Iran. 22, 6213–6232.

Biswas, S. S., & Pani, P. (2015). Estimation of soil erosion using RUSLE and GIS techniques: a case study of
Barakar River basin, Jharkhand, India. Modeling Earth Systems and Environment, 1(4), 1–13.
https://doi.org/10.1007/S40808-015-0040-3/FIGURES/9

Bogale, A. (2021). Morphometric analysis of a drainage basin using geographical information system in Gilgel
Abay watershed, Lake Tana Basin, upper Blue Nile Basin, Ethiopia. Applied Water Science, 11(7), 1–7.

Bopche, L., & Rege, P. P. (2022). Landslide Susceptibility Mapping: An Integrated Approach using Geographic
Information Value, Remote Sensing, and Weight of Evidence Method. Geotechnical and Geological Engineering,
40(6), 2935–2947. https://doi.org/10.1007/S10706-022-02070-4/TABLES/3

Borthakur, M., & Nath, B. K. (2012). A study of changing urban landscape and heat island phenomenon in
Guwahati metropolitan area. Int J Sci Res Publ, 2(11), 1-6.

Chauhan, R., Datta, A., Ramanathan, A. L., & Adhya, T. K. (2015). Factors influencing spatio-temporal variation of
methane and nitrous oxide emission from a tropical mangrove of eastern coast of India. Atmospheric
Environment, 107, 95–106. https://doi.org/10.1016/J.ATMOSENV.2015.02.006

Choudhury, B. U., Nengzouzam, G., & Islam, A. (2022). Runoff and soil erosion in the integrated farming systems
based on micro-watersheds under projected climate change scenarios and adaptation strategies in the eastern
Himalayan mountain ecosystem (India). Journal of Environmental Management, 309, 114667.
https://doi.org/10.1016/J.JENVMAN.2022.114667

Crăciun, A., Costache, R., Bărbulescu, A., Pal, S. C., Costache, I., & Dumitriu, C. Ștefan. (2022). Modern
Techniques for Flood Susceptibility Estimation across the Deltaic Region (Danube Delta) from the Black
Sea’s Romanian Sector. Journal of Marine Science and Engineering 2022, Vol. 10, Page 1149, 10(8),
1149. https://doi.org/10.3390/JMSE10081149

Cunha, E. R. da, Santos, C. A. G., Silva, R. M. da, Panachuki, E., de Oliveira, P. T. S., Oliveira, N. de S., & Falcão,
K. dos S. (2022). Assessment of current and future land use/cover changes in soil erosion in the Rio da Prata basin
(Brazil). Science of The Total Environment, 818, 151811. https://doi.org/10.1016/J.SCITOTENV.2021.151811
Dahri, N., Yousfi, R., Bouamrane, A., Abida, H., Pham, Q. B., & Derdous, O. (2022). Comparison of analytic
network process and artificial neural network models for flash flood susceptibility assessment. Journal of African
Earth Sciences, 193, 104576. https://doi.org/10.1016/J.JAFREARSCI.2022.104576

David Raj, A., Kumar, S., & Sooryamol, K. R. (2022). Modelling climate change impact on soil loss and erosion
vulnerability in a watershed of Shiwalik Himalayas. CATENA, 214, 106279.
https://doi.org/10.1016/J.CATENA.2022.106279

Dong, L., Shang, J., Ali, R., & Rehman, R. U. (2021). The Coupling Coordinated Relationship Between New-type
Urbanization, Eco-Environment and its Driving Mechanism: A Case of Guanzhong, China. Frontiers in
Environmental Science, 9, 128. https://doi.org/10.3389/FENVS.2021.638891/BIBTEX

Ferreira, C. S. S., Seifollahi-Aghmiuni, S., Destouni, G., Ghajarnia, N., & Kalantari, Z. (2022). Soil degradation in
the European Mediterranean region: Processes, status and consequences. Science of The Total Environment, 805,
150106. https://doi.org/10.1016/J.SCITOTENV.2021.150106

Gómez-Almonte MK (2005) Índice de Vegetación en Áreas del Bosque Seco del Noroeste del Perú a Partir de
Imágenes Satelitales. Universidad de Piura, Perú, Dissertation, pp 131

Gong, W., Liu, T., Duan, X., Sun, Y., Zhang, Y., Tong, X., & Qiu, Z. (2022). Estimating the Soil Erosion Response
to Land-Use Land-Cover Change Using GIS-Based RUSLE and Remote Sensing: A Case Study of Miyun
Reservoir, North China. Water 2022, Vol. 14, Page 742, 14(5), 742. https://doi.org/10.3390/W14050742

Gregory, K. J., & Wallingford, D. E. (1974). Drainage Basin Form and Process A Geomorphological Approach. Soil
Science Society of America Journal, 38(4), vi–vi.

hc, Hema., S, G., Srikanth, L., & Surendra, H. J. (2020). Prioritization of sub-watersheds of the Kanakapura
Watershed in the Arkavathi River Basin, Karnataka, India- using Remote sensing and GIS. 5(2), 149–160.
https://doi.org/10.1080/24749508.2020.1846841

He, D., Hou, K., Wen, J. F., Wu, S. Q., & Wu, Z. P. (2021). A coupled study of ecological security and land use
change based on GIS and entropy method—a typical region in Northwest China, Lanzhou. Environmental Science
and Pollution Research, 1–13. https://doi.org/10.1007/S11356-021-16080-X/FIGURES/10

Hembram, T. K., & Saha, S. (2020). Prioritization of sub-watersheds for soil erosion based on morphometric
attributes using fuzzy AHP and compound factor in Jainti River basin, Jharkhand, Eastern India. Environment,
Development and Sustainability, 22(2), 1241–1268. https://doi.org/10.1007/S10668-018-0247-3/FIGURES/11

Hibi, A., Gouaidia, L., &Guefaifia, O. (2021). Investigation of Groundwater Potential Using Remote Sensing and
Hydro-geophysical Techniques: A Case Study of the Telidjene Basin (Eastern Algeria). Environmental Research,
Engineering and Management , 77(4), 99–121. https://doi.org/10.5755/J01.EREM.77.4.29560

Horton, R. E. (1932). Drainage-basin characteristics. Eos, Transactions American Geophysical Union, 13(1), 350–
361. https://doi.org/10.1029/TR013I001P00350
Jha, M. K., Shekhar, A., & Jenifer, M. A. (2020). Assessing groundwater quality for drinking water supply using
hybrid fuzzy-GIS-based water quality index. Water Research, 179, 115867.
https://doi.org/10.1016/J.WATRES.2020.115867

Kale, H., Kale, H. S., & Deshmukh, S. B. (2020). Morphometric Analysis of WGKD Sub-watershed using Remote
Sensing and GIS Techniques. 5(1), 35–42. http://earthexplorer.usgs.gov

Kamberis, E., Bathrellos, G. D., Kokinou, E., & Skilodimou, H. D. (2012). Correlation between the structural
pattern and the development of the hydrographic network in a portion of the western thessaly basin (Greece).
Central European Journal of Geosciences, 4(3), 416–424. https://doi.org/10.2478/S13533-011-0074-
7/MACHINEREADABLECITATION/RIS

Kebede, Y. S., Endalamaw, N. T., Sinshaw, B. G., &Atinkut, H. B. (2021). Modeling soil erosion using RUSLE and
GIS at watershed level in the upper beles, Ethiopia. Environmental Challenges, 2, 100009.
https://doi.org/10.1016/J.ENVC.2020.100009

Ki, S. J., & Ray, C. (2014). Using fuzzy logic analysis for siting decisions of infiltration trenches for highway runoff
control. Science of The Total Environment, 493, 44–53. https://doi.org/10.1016/J.SCITOTENV.2014.05.121

Kulimushi, L. C., Choudhari, P., Mubalama, L. K., &Banswe, G. T. (2021). GIS and remote sensing-based
assessment of soil erosion risk using RUSLE model in South-Kivu province, eastern, Democratic Republic of
Congo. 12(1), 961–987. https://doi.org/10.1080/19475705.2021.1906759

Kumar, A., Singh, S., Pramanik, M., Chaudhary, S., & Negi, M. S. (2022). Soil erodibility mapping using watershed
prioritization and morphometric parameters in conjunction with WSA, SPR and AHP-TOPSIS models in
Mandakini basin, India. Https://Doi.Org/10.1080/15715124.2022.2114485, 1–67.
https://doi.org/10.1080/15715124.2022.2114485

Kumar, A., Singh, S., Pramanik, M., Chaudhary, S., Maurya, A. K., & Kumar, M. (2021). Watershed prioritization
for soil erosion mapping in the Lesser Himalayan Indian basin using PCA and WSA methods in conjunction with
morphometric parameters and GIS-based approach. Environment, Development and Sustainability, 1–39.
https://doi.org/10.1007/S10668-021-01586-8/TABLES/12

Kumar, R., & Anbalagan, R. (2016). Landslide susceptibility mapping using analytical hierarchy process (AHP) in
Tehri reservoir rim region, Uttarakhand. Journal of the Geological Society of India 2016 87:3, 87(3), 271–286.
https://doi.org/10.1007/S12594-016-0395-8

Lai, Z., Ge, D., Xia, H., Yue, Y., &Wangid, Z. (2020). Coupling coordination between environment, economy and
tourism: A case study of China. https://doi.org/10.1371/journal.pone.0228426

López-Pérez, A., & Fernández-Reynoso, D. S. (2021). Watershed prioritization using morphometric analysis and
vegetation index: a case study of Huehuetan river sub-basin, Mexico. Arabian Journal of Geosciences, 14(18), 1–
21. https://doi.org/10.1007/S12517-021-08212-X/FIGURES/9
Melaku, N. D., Renschler, C. S., Flagler, J., Bayu, W., & Klik, A. (2018). Integrated impact assessment of soil and
water conservation structures on runoff and sediment yield through measurements and modeling in the Northern
Ethiopian highlands. CATENA, 169, 140–150. https://doi.org/10.1016/J.CATENA.2018.05.035

Mengie, M. A., Hagos, Y. G., Malede, D. A., & Andualem, T. G. (2022). Assessment of soil loss rate using GIS–
RUSLE interface in Tashat Watershed, Northwestern Ethiopia. Journal of Sedimentary Environments 2022 7:3,
7(3), 617–631. https://doi.org/10.1007/S43217-022-00112-8

Merg, C., Petri, D., Bodoira, F., Nini, M., Fernández Díez, M. J., Schmindt, F., Montalva, R., Guzmán, L.,
Rodríguez, K., Blanco, F., & Selzer, F. (2011). Mapas digitales regionales de lluvias, índice estandarizado de
precipitación e índice verde. Pilquen - Sección Agronomía, ISSN-e 1851-2852, No. 11, 2011, 11, 5.
https://dialnet.unirioja.es/servlet/articulo?codigo=3788351&info=resumen&idioma=ENG

Meshram, S. G., & Sharma, S. K. (2017). Prioritization of watershed through morphometric parameters: a PCA-
based approach. Applied Water Science, 7(3), 1505–1519. https://doi.org/10.1007/S13201-015-0332-
9/FIGURES/6

Meshram, S. G., Tirivarombo, S., Meshram, C., & Alvandi, E. (2022). Prioritization of soil erosion-prone sub-
watersheds using fuzzy-based multi-criteria decision-making methods in Narmada basin watershed, India.
International Journal of Environmental Science and Technology, 1–12. https://doi.org/10.1007/S13762-022-
04044-8/FIGURES/4

Miller, V. (1953). A Quantitative Geomorphic Study Of Drainage Basin Characteristics In The Clinch Mountain
Area Virginia And Tennessee. https://apps.dtic.mil/sti/pdfs/AD0057755.pdf

Mohamed, A., & Worku, H. (n.d.). Urban land cover and morphometric analysis for flash flood vulnerability
mapping and riparian landscape conservation in Kebena River watershed, Addis Ababa.
https://doi.org/10.1007/s12518-020-00318-3

Mohebbi Tafreshi, A., Mohebbi Tafreshi, G., &Bijeh Keshavarzi, M. H. (2018). Qualitative zoning of groundwater
to assessment suitable drinking water using fuzzy logic spatial modelling via GIS. Water and Environment
Journal, 32(4), 607–620. https://doi.org/10.1111/WEJ.12358

Moonjun, R., Shrestha, D. P., & Jetten, V. G. (2020). Fuzzy logic for fine-scale soil mapping: A case study in
Thailand. CATENA, 190, 104456. https://doi.org/10.1016/J.CATENA.2020.104456

MR, R., C, B., & Achyuthan, H. (2019). Quantitative analysis of the drainage and morphometric characteristics of
the Palar River basin, Southern Peninsular India; using bAd calculator (bearing azimuth and drainage) and GIS.
3(4), 295–307. https://doi.org/10.1080/24749508.2018.1563750

Myles, A. J., Feudale, R. N., Liu, Y., Woody, N. A., & Brown, S. D. (2004). An introduction to decision tree
modeling. Journal of Chemometrics, 18(6), 275–285. https://doi.org/10.1002/CEM.873
Naqvi, H. R., Athick, A. S. M. A., Siddiqui, L., & Siddiqui, M. A. (2019). Multiple modeling to estimate sediment
loss and transport capacity employing hourly rainfall and In-Situ data: A prioritization of highland watershed in
Awash River basin, Ethiopia. CATENA, 182, 104173. https://doi.org/10.1016/J.CATENA.2019.104173

Nath, B., Ni-Meister, W., & Choudhury, R. (2021). Impact of urbanization on land use and land cover change in
Guwahati city, India and its implication on declining groundwater level. Groundwater for Sustainable
Development, 12, 100500. https://doi.org/10.1016/J.GSD.2020.100500

Noori, A. M., Pradhan, B., & Ajaj, Q. M. (2019). Dam site suitability assessment at the Greater Zab River in
northern Iraq using remote sensing data and GIS. Journal of Hydrology, 574, 964–979.
https://doi.org/10.1016/J.JHYDROL.2019.05.001

Ostovari, Y., Moosavi, A. A., Mozaffari, H., & Pourghasemi, H. R. (2021). RUSLE model coupled with RS-GIS for
soil erosion evaluation compared with T value in Southwest Iran. Arabian Journal of Geosciences, 14(2), 1–15.
https://doi.org/10.1007/S12517-020-06405-4/TABLES/7

Pal, S., Paul, S., & Debanshi, S. (2022). Identifying sensitivity of factor cluster based gully erosion susceptibility
models. Environmental Science and Pollution Research, 1–20. https://doi.org/10.1007/S11356-022-22063-
3/TABLES/7

Pathare, J. A., & Pathare, A. R. (2021). Watershed prioritization for soil and water conservation in Darna River
basin: a PCA approach. Sustainable Water Resources Management, 7(4), 1–15. https://doi.org/10.1007/S40899-
021-00531-X/FIGURES/7

Patle, D., Rao, J. H., & Dubey, S. (2020). Morphometric Analysis And Prioritization Of Sub-Watersheds In Nahra
Watershed Of Balaghat District, Madhya Pradesh: A Remote Sensing And Gis Perspective. Journal of
Experimental Biology and Agricultural Sciences, 4, 447–455. https://doi.org/10.18006/2020.8(4).447.455

Patowary, S., & Sarma, A. K. (2018). GIS-Based Estimation of Soil Loss from Hilly Urban Area Incorporating Hill
Cut Factor into RUSLE. Water Resources Management, 32(10), 3535–3547. https://doi.org/10.1007/S11269-018-
2006-5/TABLES/4

Pawe, C. K., & Saikia, A. (2017). Unplanned urban growth: land use/land cover change in the Guwahati
Metropolitan Area, India. Https://Doi.Org/10.1080/00167223.2017.1405357, 118(1), 88–100.
https://doi.org/10.1080/00167223.2017.1405357

Pondari, S., Dandabathula, G., Bera, A. K., Nagamani, P. V., & Amminedu, E. (2020). Characterization of drainage
network of Brahmaputra river basin in Indian sub-continent using geospatial technologies. Journal of Science
Technology and Environment Informatics, 8(1), 583–594. https://doi.org/10.18801/JSTEI.080120.60

Pourghasemi, H. R., Honarmandnejad, F., Rezaei, M., Tarazkar, M. H., & Sadhasivam, N. (2021). Prioritization of
water erosion–prone sub-watersheds using three ensemble methods in Qareaghaj catchment, southern Iran.
Environmental Science and Pollution Research 2021 28:28, 28(28), 37894–37917.
https://doi.org/10.1007/S11356-021-13300-2
Prabhakar, A. K., Singh, K. K., Lohani, A. K., & Chandniha, S. K. (2019). Study of Champua watershed for
management of resources by using morphometric analysis and satellite imagery. Applied Water Science 2019 9:5,
9(5), 1–16. https://doi.org/10.1007/S13201-019-1003-Z

Rahaman, M. H., Sajjad, H., Roshani, Masroor, M., Bhuyan, N., & Rehman, S. (2022). Delineating groundwater
potential zones using geospatial techniques and fuzzy analytical hierarchy process (FAHP) ensemble in the data-
scarce region: evidence from the lower Thoubal river watershed of Manipur, India. Arabian Journal of
Geosciences, 15(8). https://doi.org/10.1007/S12517-022-09946-Y

Ramesh, V., & Anbazhagan, S. (2015). Landslide susceptibility mapping along Kolli hills Ghat road section (India)
using frequency ratio, relative effect and fuzzy logic models. Environmental Earth Sciences, 73(12), 8009–8021.
https://doi.org/10.1007/S12665-014-3954-6/FIGURES/6

Ramesh, V., & Anbazhagan, S. (2015). Landslide susceptibility mapping along Kolli hills Ghat road section (India)
using frequency ratio, relative effect and fuzzy logic models. Environmental Earth Sciences, 73(12), 8009–8021.
https://doi.org/10.1007/S12665-014-3954-6/FIGURES/6

Ramesh, V., & Iqbal, S. S. (2020). Urban flood susceptibility zonation mapping using evidential belief function,
frequency ratio and fuzzy gamma operator models in GIS: a case study of Greater Mumbai, Maharashtra, India.
Https://Doi.Org/10.1080/10106049.2020.1730448, 37(2), 581–606.

Rehman, A., Song, J., Haq, F., Mahmood, S., Ahamad, M. I., Basharat, M., Sajid, M., & Mehmood, M. S. (2022).
Multi-Hazard Susceptibility Assessment Using the Analytical Hierarchy Process and Frequency Ratio Techniques
in the Northwest Himalayas, Pakistan. Remote Sensing 2022, Vol. 14, Page 554, 14(3), 554.
https://doi.org/10.3390/RS14030554

Renard, K. G. (1997). Predicting soil erosion by water: a guide to conservation planning with the Revised Universal
Soil Loss Equation (RUSLE). United States Government Printing.

Rodrigues, M. V. C., Guimarães, D. V., Galvão, R. B., Patrick, E., & Fernandes, F. (2022). Urban watershed
management prioritization using the rapid impact assessment matrix (RIAM-UWMAP), GIS and field survey.
Environmental Impact Assessment Review, 94, 106759. https://doi.org/10.1016/J.EIAR.2022.106759

Roshani, Sajjad, H., Rahaman, M. H., Rehman, S., Masroor, M., & Ahmed, R. (2022). Assessing forest health using
remote sensing-based indicators and fuzzy analytic hierarchy process in Valmiki Tiger Reserve, India.
International Journal of Environmental Science and Technology. https://doi.org/10.1007/S13762-022-04512-1

Saaty, T. (1980, November). The analytic hierarchy process (AHP) for decision making. In Kobe, Japan (pp. 1-69).

Saaty, T. L., & Vargas, L. G. (2001). The Decision by the US Congress on China's Trade Status: A Multicriteria
Analysis. 305–317. https://doi.org/10.1007/978-1-4615-1665-1_22

Saaty, T. L., & Vargas, L. G. (2001). Models, methods, concepts, and applications of the analytic hierarchy
process (1st ed., p. 333). Boston: Kluwer Academic.
Schumm, S., science, R. L.-A. journal of, & 1965, undefined. (n.d.). Time, space, and causality in geomorphology.
Faculty.Washington. Edu. Retrieved March 19, 2022, from
http://faculty.washington.edu/cet6/pub/Temp/CFR521e/Schumm_1965.pdf

Sharda, V. N., Mandal, D., & Dogra, P. (2021). Prioritizing soil conservation measures based on water erosion risk
and production and bio-energy losses in peninsular South Indian states. CATENA, 202, 105263.
https://doi.org/10.1016/J.CATENA.2021.105263

Sharma, D., Saha, A., & Sarma, B. (2021). Morphometric Analysis of Deepor Beel Basin Using GIS Related papers
Use of Geographical Information System in Hypsometric Analysis of Watershed. International Journal for
Research in Engineering Application & Management (IJREAM), 06, 2454–9150.

Sheik Mohideen, A. R. (2021). Morphometric assessment of hydrogeomorphic processes and landscape evolution in
the Kallar watershed (Western Ghats, India): regionalization and prioritization. Arabian Journal of Geosciences,
14(18), 1–28. https://doi.org/10.1007/S12517-021-08105-Z/FIGURES/14

Shekar, P. R., & Mathew, A. (2022). Evaluation of Morphometric and Hypsometric Analysis of the Bagh River
Basin using Remote Sensing and Geographic Information System Techniques. Energy Nexus, 7, 100104.
https://doi.org/10.1016/J.NEXUS.2022.100104

Siddiqui, R., Said, S., & Shakeel, M. (2020). Nagmati River Sub-watershed Prioritization Using PCA, Integrated
PCWS, and AHP: A Case Study. Natural Resources Research, 29(4), 2411–2430. https://doi.org/10.1007/S11053-
020-09622-6/FIGURES/7

Singh, W. R., Barman, S., & Tirkey, G. (2021). Morphometric analysis and watershed prioritization in relation to
soil erosion in Dudhnai Watershed. Applied Water Science, 11(9), 1–12.

Somasiri, I. S., Hewawasam, T., & Rambukkange, M. P. (2022). Adaptation of the revised universal soil loss
equation to map spatial distribution of soil erosion in tropical watersheds: a GIS/RS-based study of the Upper
Mahaweli River Catchment of Sri Lanka. Modeling Earth Systems and Environment, 8(2), 2627–2645.

Strahler, A. N. (1945). Hypotheses of stream development in the folded Appalachians of Pennsylvania. Geological
Society of America Bulletin, 56(1), 45-88.

Strahler, A. N. (1957). Quantitative analysis of watershed geomorphology. Eos, Transactions American Geophysical
Union, 38(6), 913–920. https://doi.org/10.1029/TR038I006P00913

Strahler, A. N. (1964). Quantitative geomorphology of drainage basin and channel networks. Handbook of applied
hydrology.

Talukdar, S., & Pal, S. (2019). Wetland habitat vulnerability of lower Punarbhaba river basin of the uplifted Barind
region of Indo-Bangladesh. 35(8), 857–886. https://doi.org/10.1080/10106049.2018.1533594
Teng, H., Liang, Z., Chen, S., Liu, Y., Viscarra Rossel, R. A., Chappell, A., Yu, W., & Shi, Z. (2018). Current and
future assessments of soil erosion by water on the Tibetan Plateau based on RUSLE and CMIP5 climate models.
Science of The Total Environment, 635, 673–686. https://doi.org/10.1016/J.SCITOTENV.2018.04.146

Tien Bui, D., Pradhan, B., Lofman, O., Revhaug, I., & Dick, O. B. (2012). Spatial prediction of landslide hazards in
Hoa Binh province (Vietnam): A comparative assessment of the efficacy of evidential belief functions and fuzzy
logic models. CATENA, 96, 28–40.

Tiwari, R. N., & Kushwaha, V. K. (2021). Watershed Prioritization Based on Morphometric Parameters and PCA
Technique: A Case Study of Deonar River Sub Basin, Sidhi Area, Madhya Pradesh, India. Journal of the
Geological Society of India 2021 97:4, 97(4), 396–404. https://doi.org/10.1007/S12594-021-1697-Z

Tsering, T., Sillanpää, M., Reinikainen, S. P., & Abdel Wahed, M. S. M. (2020). Metal Fractionation in Surface
Sediments of the Brahmaputra River and Implications for Their Mobilization. International Journal of
Environmental Research and Public Health 2020, Vol. 17, Page 9214, 17(24), 9214.
https://doi.org/10.3390/IJERPH17249214

Tucker, C. J. (1979). Red and photographic infrared linear combinations for monitoring vegetation. Remote Sensing
of Environment, 8(2), 127–150. https://doi.org/10.1016/0034-4257(79)90013-0

Vema, V., Sudheer, K. P., & Chaubey, I. (2019). Fuzzy inference system for site suitability evaluation of water
harvesting structures in rainfed regions. Agricultural Water Management, 218, 82–93.
https://doi.org/10.1016/J.AGWAT.2019.03.028

Vinutha, D. N., & Janardhana, M. R. (2014). Morphometry of The payaswini watershed, coorg district, karnataka,
India, using remote sensing and GIS techniques. International Journal of Innovative Research in Science,
Engineering and Technology, 3(5), 516-524.

Yang, C., Zeng, W., & Yang, X. (2020). Coupling coordination evaluation and sustainable development pattern of
geo-ecological environment and urbanization in Chongqing municipality, China. Sustainable Cities and Society,
61, 102271. https://doi.org/10.1016/J.SCS.2020.102271

Ye, Y., & Qiu, H. (2021). Environmental and social benefits, and their coupling coordination in urban wetland
parks. Urban Forestry & Urban Greening, 60, 127043. https://doi.org/10.1016/J.UFUG.2021.127043

Yu, D., Dong, X., Xie, P., Wei, C., Liu, J., Hu, X., Wang, K., Xu, S., Wan, H., & Su, Z. (2021). Prioritization of
critical source areas for soil and water conservation by using a one-at-a-time removal approach in the upper
Huaihe River basin. Land Degradation & Development, 32(3), 1513–1524. https://doi.org/10.1002/LDR.3814

Yuan, Y., Jin, M., Ren, J., Hu, M., & Ren, P. (2014). The Dynamic Coordinated Development of a Regional
Environment-Tourism-Economy System: A Case Study from Western Hunan Province, China. Sustainability
2014, Vol. 6, Pages 5231-5251, 6(8), 5231–5251. https://doi.org/10.3390/SU6085231

Zadeh, L. A. (1965). Fuzzy sets. Information and Control, 8(3), 338–353. https://doi.org/10.1016/S0019-
9958(65)90241-X
Zhang, Y., Chen, J., Wang, Q., Tan, C., Li, Y., Sun, X., & Li, Y. (2021). GIS-models with fuzzy logic for
Susceptibility Maps of debris flow using multiple types of parameters: A Case Study in Pinggu District of Beijing,
China. Natural Hazards and Earth System Sciences, 1–23. https://doi.org/10.5194/NHESS-2021-254

Zhao, Q., Chen, W., Peng, C., Wang, D., Xue, W., & Bian, H. (2022). Modeling landslide susceptibility using an
evidential belief function-based multiclass alternating decision tree and logistic model tree. Environmental Earth
Sciences, 81(15), 1–16. https://doi.org/10.1007/S12665-022-10525-3/FIGURES/13

Mosavi, A., Sajedi-Hosseini, F., Choubin, B., Taromideh, F., Rahi, G., & Dineva, A. A. (2020). Susceptibility
mapping of soil water erosion using machine learning models. Water, 12(7), 1995.

Kalbasi, R., Jahangiri, M., Mosavi, A., Dehshiri, S. J. H., Dehshiri, S. S. H., Ebrahimi, S., Karimipour, A. (2021).
Finding the best station in Belgium to use residential-scale solar heating, one-year dynamic simulation with
considering all system losses: economic analysis of using ETSW. Sustainable Energy Technologies and
Assessments, 45, 101097.

Mosavi, A., Golshan, M., Janizadeh, S., Choubin, B., Melesse, A. M., & Dineva, A. A. (2022). Ensemble models of
GLM, FDA, MARS, and RF for flood and erosion susceptibility mapping: a priority assessment of sub-basins.
Geocarto International, 37(9), 2541-2560.

Ayoobi, N., Sharifrazi, D., Alizadehsani, R., Shoeibi, A., Gorriz, J. M., Moosaei, H., Mosavi, A. (2021). Time series
forecasting of new cases and new deaths rate for COVID-19 using deep learning methods. Results in Physics, 27,
104495.