Research

Urban flood risk assessment using AHP and geospatial techniques

in swat Pakistan
Muhammad Waseem1 · Sareer Ahmad1 · Izhar Ahmad1 · Hira Wahab1 · Megersa Kebede Leta2

Received: 15 May 2023 / Accepted: 13 July 2023

© The Author(s) 2023  OPEN

Abstract
The rapid urbanization and changing climate patterns in Swat, Pakistan have increased the vulnerability of urban areas
to flood events. Accurate assessment of flood risk is crucial for effective urban planning and disaster management. In
current research study flood hazard index was developed using analytic hierarchy process (AHP) technique in combina-
tion with the geographical information system (GIS) environment in Swat, Pakistan. The study integrates various data
sources, including topographic maps, land use/land cover information, rainfall data, and infrastructure data, to develop a
comprehensive flood risk assessment model. The weights obtained from the AHP analysis are combined with geospatial
data using a geographic information system (GIS) to generate flood risk maps. The flood hazard levels were categorized
into five distinct classes: very low, low, moderate, high, and very high. Using the GIS-AHP approach, higher weights were
assigned to rainfall, distance to river, elevation, and slope in comparison to NDVI, TWI, LULC, curvature, and soil type.
The flood hazard map was then reclassified for each parameter. By overlaying these maps, it was determined that 5.6%
of the total area is classified as very high flood risk, 52% as high risk, 39.3% as moderate risk, and 3.1% as low risk. The
developed comprehensive flood risk assessment model in current study can identify high-risk areas, prioritize mitigation
measures, and aid in effective urban planning and disaster management.

Article Highlights

• Flood risk assessment model using GIS and AHP iden- • GIS-generated flood hazard maps aid in prioritizing
tifies high-risk areas for effective urban planning and mitigation measures and increasing public aware-ness
disaster management. for flood risk reduction.
• Elevation, slope, rainfall, and proximity to the river are
significant factors influencing flood vulnerabil-ity in
Swat, Pakistan.

Keywords Flood risks management · Sensitivity assessment · Swat watershed flood hazard index (FHI) · Environmental
factors · Spatial analysis · Geographic information systems

* Megersa Kebede Leta, megersa.kebede@uni-rostock.de; Muhammad Waseem, muhammad.waseem@giki.edu.pk; Sareer Ahmad,

gcv2341@giki.edu.pk; Izhar Ahmad, gcv2339@giki.edu.pk; Hira Wahab, gcv2340@giki.edu.pk | 1Department of Civil Engineering, Ghulam
Ishaq Khan Institute of Engineering Sciences and Technology, Topi 23640, Pakistan. 2Faculty of Agriculture and Environmental Sciences,
University of Rostock, 18059 Rostock, Germany.

SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1

Vol.:(0123456789)
Research SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1

1 Introduction impacts, access to reliable and up-to-date information

is essential. A hazard map that depicts the likelihood of
Floods are one of the world’s most deadly risks, and more flooding is one of these essential sorts of information. To
than 1 billion people live in flood-prone areas [1]. Under evaluate and forecast flood risk, numerous investigations
projected climate change scenarios, extreme hydrologi- have used climatic variables, geomorphic and physical
cal events and flood hazards are especially anticipated watershed parameters, soil characteristics, and land use/
to be significant and to increase over time [2]. According cover determined from remote sensing data [13]. This
to the HadCM3 climate model, the probability of floods is carried out to promote sustainable development and
worldwide will rise by roughly 187% in 2050, contrasting management and effective flood disaster mitigation [12,
with a condition devoid of universal climate change [3]. 14].
Flood dangers, compared to other natural catastrophes, The most commonly used variables in environmental
affected 45% of the world’s population and resulted in studies are natural variables (such as elevation, slope,
5424 documented fatalities between 2000 and 2017 [4]. aspect, slope curvature, lithology, topographic position
Geographically, Asia, America, and Europe will experi- index, and rainfall), hydrological variables (such as drain-
ence the greatest increase in flood danger in the future. age density, river distance, topographic wetness index,
By the 2070s, populations residing in Asia’s fastest- and stream power index), and human disturbance factors
urbanizing cities, such as Bangkok, Jakarta, Dhaka, and [15–17]. Accurate modeling of flood risk threats is crucial
Mumbai, among others, will be particularly vulnerable for successful floodplain land use planning and mitigation
to coastal floods and flood hazards [5, 6]. Severe floods methods. Availability of easily accessible maps enables
pose a threat to cultural treasures and have numerous planners to identify the most hazardous locations geo-
detrimental effects on people’s well-being and the econ- graphically, which is a vital step in the mitigation process
omy [7]. Since late November 2011, Thailand has been [18, 19]. Despite the challenges involved in flood predic-
struck by some of the largest flood catastrophes in his- tion and prevention, a variety of remote sensing and GIS
tory, impacting around 10 million people, with 65 out of techniques can be employed to map areas susceptible
Thailand’s 77 provinces being designated disaster zones. to flash floods and anticipate their occurrence [20–22].
According to a World Bank report, the 2011 flooding in Multiple datasets can be obtained through the analysis
Thailand caused an estimated $80 billion in economic of remote sensing data using GIS techniques to pinpoint
damage, ranking it as the fourth most expensive envi- high-risk areas and produce flash flood danger maps. The
ronmental disaster worldwide from 1995 to 2011 [8]. GIS-weighted overlay analysis-based knowledge-driven
Floods, typhoons, heatwaves, and other extreme AHP approach has proven effective in modeling various
weather phenomena are occurring more frequently and environmental scenarios [23].
with greater intensity as a result of climate change, plac- The GIS method can forecast and find more water
ing many nations at risk. Concerns regarding hydrologi- resources by combining and analyzing vast quantities of
cal risks and water resource management are profound spatial data [23]. This method incorporates criteria into
due to climate change. Among the most hazardous natu- multiple sub-criteria and a hierarchical tree with multiple
ral disasters are flash floods, which occur in basins of var- levels. Several prediction experiments have applied this
ious sizes encompassing a confined region of hundreds multicriteria decision-making approach to generate a solu-
of square kilometers on all continents, causing severe tion based on hierarchical ordering criteria for challenging
damage and disruption to human activity [9]. These nat- decision analysis [24–26]. The input data set for the model
ural calamities are caused by the rapid onset and high- includes information such as elevation, slope, topographic
intensity rainstorms, particularly in areas with uneven witness index (TWI), stream power index (SPI), curvature,
terrain, leading to sudden and swift flooding [10]. drainage density, depressions, runoff, distance to rivers,
Flood catastrophes are among the most frequent and rainfall. In arid to semiarid regions like Pakistan, rainfall
natural hazard due to climate change and other environ- varies in both time and location. These regions also expe-
mental problems. It is only recently that the world has rience mild to warm winters and scorching to extremely
recognized the frequency with which flooding is caused hot summers [27]. Consequently, irregular flash floods
by climate change. Local rapid response teams face chal- occur, causing damage to dams, roads, people’s lives, and
lenges in predicting the timing of peak water levels, leav- property [28]. The devastating flash floods that impacted
ing little time for warnings. Consequently, the under- the Swat region in 2010 and 2022 are estimated to have
standing of flash flood prediction and control remains resulted in damage costs of RS.68 billion [29].
limited [11, 12]. To avert floods or at least decrease their

Vol:.(1234567890)
SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1 Research

Pakistan, with its fifth-largest population of approxi- 2 Materials and methods

mately 220 million people, is projected to reach 250 million
by 2025 with an annual growth rate of 2.8%. Over the past 2.1 Study area
60 years, the country has experienced nearly 19 significant
floods, resulting in the inundation of an area of 594,700 Swat is a valley and administrative district located in the
km2, affecting 166,075 communities, and causing total Khyber Pakhtunkhwa province of Pakistan. It is situated
direct cumulative losses of around $30 billion. Tragically, in the country’s northwestern region, surrounded by the
these floods also claimed the lives of 10,668 individuals Hindu Kush Mountain range to the north and west, and
[30]. While floods are a natural phenomenon that cannot be the Swat River flows through the valley. The geographic
entirely prevented, the risk can be mitigated through effec- coordinates of Swat are approximately 35.3859° N and
tive planning and mitigation strategies. Numerous studies 72.4486° E, as shown in Fig. 1 below.
have shown that the identification of flood-prone areas and The climate of Swat is moderate, with pleasant sum-
the implementation of critical risk reduction measures, both mers and freezing winters. Summer runs from May
structural and non-structural, can successfully reduce flood through September, with temperatures ranging from
losses to an acceptable level [12, 31]. 20 to 30 °C. Temperatures range from − 2 to 12 °C dur-
Researchers utilize various models to assess flood suscep- ing the winter season, which lasts from November to
tibility. According to Guo et al., traditional methodologies March. Swat receives about 800 mm of rain every year,
for mapping flood risk often have limited effectiveness due and the monsoon season lasts from July to September.
to insufficient data [32]. Conventional procedures neces- The Swat River is the major river that flows through the
sitate data collection, which, especially in poor nations, is valley. It originates from the Hindu Kush Mountain range
frequently expensive, time-consuming, and unavailable at and passes through the city of Mingora before joining
the local or regional level. On the other hand, geographic the Kabul River. The Swat River is fed by several tributar-
information systems (GIS) and remote sensing are advanced ies, including the Panjkora, Daral, and Bara rivers. The
dynamic tools that can provide a variety of data for risk man- river is an essential source of irrigation for agriculture
agement, flood zoning, and forecasting. Because natural and a significant source of hydroelectric power genera-
disasters are multidimensional phenomena with a spatial tion in the region. The topography of Swat is dominated
component, GIS is an especially useful tool for this type of by mountain ranges, with the highest peaks being Falak
investigation because it can manage massive volumes of Sher (6000 m), Mankial (5726 m), and Flir (5396 m). The
spatial data used in flood modeling [12]. Flash flood man- valley is surrounded by several passes, including the
agement strategies heavily depend on modeling the hydro- Lowari Pass (3118 m) to the west, the Malakand Pass
logical, meteorological, and topographic factors of a catch- (1207 m) to the east, and the Shangla Pass (2143 m) to
ment region to mitigate the risk of flooding. Conducting risk the south. Overall, Swat’s geographical location, climate,
analyses and employing these mapping methods should be and orography make it a unique and beautiful region in
prioritized for timely completion. The study motivation for Pakistan, attracting tourists from all over the world. Swat
selecting Swat Valley is that it has been prone to devastat- valley is situated in the northern part of Pakistan, which
ing floods in the past, leading to substantial economic and is a part of the Himalayan orogenic belt. Swat’s geology
human losses. However, the existing literature lacks compre- comprises sedimentary, metamorphic, and igneous rocks
hensive hazard maps and flood risk assessment studies spe- that formed during the Paleozoic to the Cenozoic era.
cific to this region, which are essential for effective disaster The sedimentary rocks in Swat, including sandstone,
management and urban planning. To fill this critical gap, our shale, and limestone, were deposited during the Pale-
study sought to employ Analytical Hierarchy Process (AHP) ozoic era in a shallow marine environment. Swat also
and geospatial techniques to develop a systematic and data- features metamorphic rocks such as schist, gneiss, and
driven approach for evaluating flood risks in urban areas of marble, formed as a result of tectonic activity during the
Swat. Himalayan orogeny. Additionally, there are igneous rocks
The current study uses AHP a multi-criteria decision like granite and diorite that intruded into the metamor-
analysis model and geo-information tools to assess flood phic rocks during the Cenozoic era. The hydrogeology of
risk and susceptibility mapping in Swat District for sustain- Swat is complex due to its diverse geological formations.
able agriculture, urban, and infrastructure development. The primary aquifers, located in the sedimentary rocks,
The Main objective of the presented study is to develop a are predominantly unconsolidated and porous. These
comprehensive flood risk map for Swat District in Pakistan. aquifers are replenished by rainfall and snowmelt, and

Vol.:(0123456789)
Research SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1

Fig. 1  Location map of the study area

groundwater flow is controlled by the region’s topogra- decisions need the development of an opinion, the com-
phy and geological structure. Wells and springs are the parison of alternatives, or the evaluation of each choice
primary sources of groundwater in Swat. Wells, usually considering an ideal. In contrast, intuitive decisions are
shallow and hand-dug, serve domestic and agricultural totally subjective and cannot be justified by a logical
needs. Springs, mainly situated in mountainous regions, argument. The latter choice is one that the AHP approves
provide water for drinking and irrigation. However, of. The assumptions for the study were derived from a
Swat’s hydrogeology is under threat from increasing thorough literature review and align with well-estab-
urbanization, deforestation, and pollution. These fac- lished factors influencing flood risk assessment. These
tors have caused a decline in water quality and quantity, assumptions include rainfall, distance to river, slope,
necessitating the implementation of effective manage- elevation, land use/land cover (LULC), topographic wet-
ment strategies to protect Swat’s groundwater resources. ness index (TWI), normalized difference vegetation index
(NDVI), soil type, and curvature. Each assumption is sup-
2.2 Data collection and model setup ported by existing research and contributes to a compre-
hensive understanding of flood dynamics in the study
This section discusses the components of morphomet- area. These assumptions are grounded in established
ric, topographic, and variable analysis involved in the scientific knowledge and align with recognized factors
creation of AHP-GIS-based maps. It is important to note contributing to flood risk. We have considered these
that the type of flood significantly impacts the variables assumptions to ensure a comprehensive assessment
utilized in the multi-parametric AHP model. Table 1 pre- of flood hazards. As a result, the variables evaluated in
sents the variables employed for vulnerability analysis, this study were chosen based on how important they
along with the corresponding risk categories and levels were for producing floods in the study area. Elevation
of relative importance. The selection of criteria using a and inclination, soil types, yearly rainfall distribution, the
spatial reference is a crucial step in multicriteria decision topo-graphic wetness index, and land-use/cover data are
analysis. The AHP integrates objective assessments with all taken into account land Use Cover downloaded from
actual data to determine initiatives, activities, schemes, NASA Earth Data and Environmental Systems Research
or alternatives based on quantitative criteria [13, 33]. Institute (Esri) [30]. The weighted overlay is then used to
According to Saaty [10], decisions can be divided into create the final flood risk map, as displayed in Fig. 2, is
two categories: intuitive and analytical. Analytical finally created.

Vol:.(1234567890)
SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1 Research

Table 1  Flood susceptibility criteria and sub-criteria ranges for flood susceptibility assessment
Flood criteria’s Unit Class Class ranges/flood Class Ratings Percent-
probability age
weight

Topographic Wetness Index Level 2.5–3.2 Very low 1 5

3.2–8.9 Low 2
8.9–14.6 Moderate 3
14.6–20.3 High 4
20.3–26 Very high 5
Slope % 0–13.70 Very high 5 16
13.70–27.40 High 4
26.40–41.10 Moderate 3
41.10–61.25 Low 2
> 61.25 Very low 1
Elevation Meter 709–1596 Very low 1 13
1596–2443 Low 2
2443–3310 Moderate 3
3310–4117 High 4
4117–5852 Very low 5
Distance from Rivers Meter 0–140 Very high 5 23
140–290 High 4
290–430 Moderate 3
430–570 Low 2
570–720 Very low 1
Soil Type Clay percentage 5 Very high 5 3
23 High 4
26 Moderate 3
30 Low 1
LULC Level Waterbody Very high 5 9
Trees High 4
Crops High 4
Built Area Moderate 3
Range Land Moderate 3
Snow Ice Low 2
Flooded Vegetation Low 2
Bare Land Very low 1
NDVI Level − 0.167–0.02 Very high 5 3
0.02–0.046 High 4
0.046–0.129 Moderate 3
0.129–0.218 Low 2
0.218–0.540 Very low 1
Curvature Level Flat (− 0,9–0,54) Very high 5 1
Convex (0,55–11) Moderate 3
Concave (− 10–1) Very low 1
Rainfall mm/year 225–418 Low 1 27
418–612 Moderate 3
612–805 High 4
805–1007 Very high 5

Vol.:(0123456789)
Research SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1

2.4 Land use types

The various land-use groups influence the amount of rain-

fall and runoff generated. Land cover plays a role in the fre-
quency of flood occurrences due to the ability of vegeta-
tion to absorb precipitation. High-resolution land use data
for the research area weren’t accessible, so the land cover
was determined using the ESRI land cover dataset https://​
www.​esri.​com he land uses were divided into eight dif-
ferent groups based on the vegetation cover. These eight
types of land uses include water bodies, trees, crops, built
areas, range land, snow/ice, flooded vegetation, and bar-
ren land, as shown in Fig. 4.

2.5 Distance to river

The risk posed by an area can be estimated by consider-

Fig. 2  Methodological framework of the current study ing the distance between the area and nearby rivers. The
streams were extracted using the D8 approach, and the
STRAHLER methodology was employed to identify dif-
2.3 Elevation and slope ferent levels in the extracted streams. In the STRAHLER
technique, links without tributaries are assigned an order
The slope plays a crucial role in determining surface or of one and referred to as first order [24]. It was observed
subsurface drainage in a specific area or outlet. The com- that cells located farther away from certain streams had
bination of slope angles, influenced by the slope’s shape, a lower likelihood of flooding compared to those nearby.
determines the interplay between lithology, structure, Distances from the streams were classified into five risk
soil type, and drainage. Higher slopes tend to experience zones, as depicted in Fig. 5.
surface runoff issues, while flatter surfaces are prone to
water accumulation. In contrast, local depressions, which 2.6 Topographic wetness index (TWI)
are cells in a digital elevation model with lower elevation and hydro‑lithological formations
than their surroundings, are more susceptible to pluvial
floods. This highlights the importance of understanding TWI is an index that measures the water’s ability to accu-
the relationship between ground elevation and flood risk mulate, with a high value indicating deeper water caused
[23]. In this study, a slope map was created using slope- by a low slope, as shown in Fig. 6. The hydrological and
generating tools and the ArcGIS digital elevation model lithological formations influence the infiltration resulting
(DEM). The DEM for the study area was obtained from the from rainfall-runoff generation.
Earth Data Search on the NASA website (https://​search.​
earth​data.​nasa.​gov/​search). The slope, which affects the 2.7 Soil types and precipitation
speed of runoff, determines the probability of catastrophic
floods. Higher slopes result in slower water penetration, Sandy soil absorbs water rapidly and has less drainage,
optimizing water velocity and flow. The length of the slope significantly impacting flooding. In contrast, clayey soils,
is a factor that influences both floods and the likelihood of which are less permeable and retain water for longer peri-
flooding since floods typically occur in lower areas where ods, make locations more susceptible to flooding. In the
water gradually accumulates as it flows from higher to absence of measurements, the soil’s texture and appear-
lower elevations. Elevation and flood risk are intercon- ance can be used to estimate its moisture content. It plays
nected, with cells in lower-elevation regions being more a crucial role in dividing precipitation into runoff and
prone to flooding [16]. To analyze elevation, it was clas- groundwater storage, acting as a point of contact between
sified into five categories ranging from 709 to 5852 feet. the atmosphere and the land surface. Soil moisture levels
The slope range in the study area is 0 to 89, divided into rise when sufficient precipitation compensates for losses
five classes using the Natural Break technique. The per- to streams and groundwater, leading to reduced soil ero-
centage categories are “0,”“0–13,”“13–27,”“27–41,”“41–61,” sion, slope stabilization, and plant growth. The soil types
and “ > 61”. Figure 3a, b, c and d presents the original and in a region are significant as they regulate water entry into
reclassified elevation and slope maps for the Swat District. the ground and, consequently, water flow. The structure

Vol:.(1234567890)
SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1 Research

Fig. 3  a Digital elevation model; b Reclass of elevation map; c Slope map; d Reclass of slope map

Vol.:(0123456789)
Research SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1

Fig. 4  a Land use land cover map; b Reclass of land use land cover map

and infiltration capacity of soils greatly influence their abil- satellite data. It provides information about the level of
ity to absorb water and function as sponges. Different soil photosynthetic activity occurring in a particular area. NDVI
types have varying capabilities. Increased surface runoff is determined by analyzing the reflectance of near-infrared
due to decreased soil infiltration capacity enhances the and visible light detected by satellite sensors.
likelihood of flooding. Flooding occurs when the rate of NDVI values range from -1 to + 1, with higher values
water supply exceeds the soil’s infiltration capacity, caus- indicating greater vegetation density and health (+ 1)
ing runoff downslope on sloped ground. Figure 7 illus- and lower values indicating lower vegetation density
trates the categorized soil cover and precipitation map and health (-1). These values are commonly employed
for the study area. to monitor vegetation growth, drought conditions, and
other environmental factors that can impact plant health.
2.8 Curvature and NDVI The results of the Curvature function are depicted in Fig. 8,
with the classifications determined through a combina-
The Curvature feature represents the contour and curva- tion of academic literature and previous research on flood
ture of the slope. It determines whether the surface is con- events [34].
cave or convex based on its second derivative. The results
of the Curvature function are displayed in Fig. 8. The classi-
fication of the curvature classes was established through a 2.9 Data acquisition
combination of academic literature and insights obtained
from previous flood research [34]. A three-grade scale was The maps in this research article were generated using
employed to standardize the classes, with each class being geographic information system (GIS) software, specifically
assigned a positive whole number. ArcGIS version 10.4 developed by Esri (). Table 2 below
The normalized difference vegetation index (NDVI) shows the source of the data acquired for the current
evaluates the density and health of vegetation by utilizing research work.

Vol:.(1234567890)
SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1 Research

Fig. 5  a Distance to river map; b Reclass of distance to river map

2.10 Analytical hierarchical process (AHP) a meaningful interpretation of the criteria. AHP can be
applied with various criteria, and once the hierarchy is
AHP, which stands for Analytic Hierarchy Process, is a cru- established, it compares the two criteria to determine
cial decision-making approach that finds applications in their relative importance.
various industries such as business, banking, education, An AHP pair-wise comparison matrix was used to
politics, and engineering. It was initially developed by establish the relative weights of the pertinent com-
Saaty in 1980 as a multicriteria decision-making method, ponents. As the criteria are prioritized, their relative
drawing inspiration from the work of Myers and Alpert weights are established. On a scale of 1 to 9, where 9
in 1968, and has since become one of the most widely is the most significant component, there is a relative
used approaches in this field [26]. The AHP is highly flex- importance scale.
ible and allows for the consideration of both quantitative The final map of flood risk zones was created by com-
and qualitative factors in decision-making, taking into bining the weights and multiplying them by the rate of
account the priorities of the user or group implementing each individual element, as given in Eq. (1).
the model. In the AHP, the comparison matrix is trans-
∑
n
formed into a priority vector, and a consistency ratio H= wi xi (1)
is calculated by assigning an arbitrary index value. The 1
highest level in the hierarchy represents the final state of
the decision, followed by its criteria, and then the least where n is the number of components, w(i) is each factor’s
preferred options at the lower level. Decisions can be weight, and x(i) is each factor’s individual rating, and H is
made at the lowest level of the hierarchy. It is crucial to the flood risk level.
accurately define the total number of criteria and each The pair-wise comparison matrix indicated in the
discrete criterion to ensure consistent pairwise compari- following equation (Eq. 2) theoretically revealed the
sons. Grouping similar criteria together is important for weighting coefficients of the stated flood risk compo-
nents in the AHP approach.

Vol.:(0123456789)
Research SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1

Fig. 6  a Topographic wetness index (TWI) map; b Reclass of TWI map

⎛ 1 aij … a1n ⎞ 𝝀max − n

� �⎜ 1� ⎟
CI = (3)
n−1
A = aij ⎜ aij 1 a2n ⎟ (2)
⎜ 1∕ 1 ⎟ where n represents the number of classes.
⎝ a1n ∕a2n 1 ⎠
Consistency ratio, CR is a pair-wise comparison matrix,
The scale of Relative Importance in the Analytic Hier- which is calculated with Eq. (4)
archy Process (AHP) is utilized to allocate weights or
CI
priorities to various criteria or factors, considering their CR = (4)
RI
respective significance or preference, to support deci-
sion-making processes. This scale is shown in Table 3 for RI donates random consistency index.
the current study. The decision matrix in AHP is a tool to If the CR is less than 0.10, the comparison matrix is con-
compare alternatives based on criteria and determine sidered to have appropriate consistency; otherwise, the
their relative importance as illustrated in Table 4. The decision-making process is repeated until consistency is
pair-wise assessment matrix is shown in Table 5 using achieved. CR = 0.00 is the best consistency value.
an 9 × 9 matrix with diagonal elements equal to 1. The The consistency (CR) ratio of 5% indicates that the AHP
ratings are computed by comparing each row’s values calculation using the decision matrix is correct because it
to each column’s values. Rainfall intensity, for example, falls within the range. A CR value of less than 10% is con-
is far more important than land use and thus has a value sidered acceptable.
of one in category. Row talks about the importance of Although AHP has been used extensively to create reli-
land use. Priorities can be calculated and placed in Arc- able hazard maps, the findings are mostly dependent on
GIS weighted overlay function after assigning values as expert judgement, which could cause errors in the deter-
in a decision matrix, with the results displayed on a map mination of the criteria weight. Sensitivity analysis was uti-
as shown in Fig. 9. lized in this study to look at the sensitivity requirements for
Consistency Index (CI), calculated using Eq. (3) mapping flood dangers. The weight of a single element is

Vol:.(1234567890)
SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1 Research

Fig. 7  a Soil map; b Reclass of soil map; c Precipitation map; d Reclass of precipitation map

Vol.:(0123456789)
Research SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1

Fig. 8  a Curvature map; b Reclass of curvature map; c NDVI map; d Reclass of NDVI map

Vol:.(1234567890)
SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1 Research

Table 2  Summary of the data used in this study and the sources from where the data is obtained
Sr. No Data Type Description Source

1 DEM ASTER global digital elevation model with 30 m cell size Earth Data Search, NASA [35]
2 Soil Data Digital Soil Map of the World- ESRI Shapefile format with FAO Map Catalog, Food and Agricultural Organization of
resolution of 5’ × 5’ the United Nations [36]
3 LULC Data Sentinel-2 10-Meter Land Use/Land Cover Esri Land Cover—ArcGIS Living Atlas [37]
4 NDVI Data Landsat Collection 2 Level-1 (Landsat 8–9 OLI/TIRS C2 L1) USGS Earth Explorer. United States Geological Survey [38]
5 Rainfall Data CRU TS Monthly precipitation high-resolution datasets CRU UAE Data, Climate Research Unit [39]
with 0.5 × 0.5 degree resolution

Table 3  The fundamental scale for pair-wise comparison changed over a range of percentages when using the one-
Intensity Definition
at-a-time (OAT) method, and the effects of the change are
then evaluated as shown in Fig. 9. In this study, the range of
1 Equal importance percent change (RCP) was applied to all criteria weights with
2 Relative importance a 5% increase in the percent change (IPC) (i.e., plus or minus
3 Moderate importance 5%) from a 20% beginning criterion weight. The weights of
4 Relative importance the primary criterion at various degrees of percent change
5 Strong importance were calculated using Eq. (5), which is displayed in Fig. 9. The
6 Relative importance interpolation table aids in assigning average values to crite-
7 Very strong importance ria within a specified range, ensuring precise and standard-
8 Relative importance ized evaluations in the decision-making process as shown
9 Extreme importance in Fig. 10.

Table 4  Decision matrix used Category Type Priority Rank (+) (-)
in the current study
1 Rainfall 27.6% 1 12.4% 12.4%
2 Distance to river 23.9% 2 11.7% 11.7%
3 Slope 15.80% 3 5.3% 5.3%
4 Elevation 12.90% 4 3.8% 3.8%
5 LULC 8.90% 5 3.4% 3.4%
6 TWI 4.00% 6 1.7% 1.7%
7 NDVI 2.6% 7 1.0% 1.0%
8 Soil type 2.70% 8 0.7% 0.7%
9 Curvature 1.60% 9 0.7% 0.7%
The Rank column is color-coded is to indicate the priority order of the nine parameters

Table 5  Priority table for each parameter

Distance to Eleva-
Criteria Rainfall Slope LULC TWI NDVI Soil type Curvature
river tion
Rainfall 1 2 3 3 4 5 6 7 9
Distance to River 0.50 1 2 4 4 6 6 8 8
Slope 0.33 0.5 1 1 3 6 7 7 9
Elevation 0.33 0.25 1 1 2 4 7 6 8
LULC 0.25 0.25 0.33 0.50 1 3 6 5 7
TWI 0.20 0.17 0.17 0.25 0.33 1 2 1 5
NDVI 0.17 0.17 0.14 0.14 0.17 0.50 1 1 2
Soil type 0.14 0.12 0.14 0.17 0.20 1 1 1 2
Curvature 0.11 0.12 0.11 0.12 0.14 0.20 0.50 0.50 1
CFR: % 5.4

The diagonal values are highlighted with color to represent the relative importance of each parameter to itself. This convention simplifies
the table’s interpretation and aids in comprehension

Vol.:(0123456789)
Research SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1

The total number of criteria is n, the RCP’s minimum and

maximum values are RPC min and RPC max, respectively
(Fig. 11).

3 Results

The study’s findings indicated that elevation, slope, rain-

fall, and distance to the river were the primary factors
influencing the occurrence of flooding in the area. The
elevation map revealed that regions with lower elevations
were more susceptible to flooding. This is due to the natu-
ral flow of water towards lower areas, making them prone
to inundation during intense rainfall events. The slope
map demonstrated that areas with steeper slopes were
at a higher risk of flash floods. Analysis of the rainfall data
Fig. 9  AHP model for all parameter
indicated that the study area experienced significant pre-
cipitation levels during the monsoon season, which played
a crucial role in the occurrence of flooding. Figure 12 fur-
𝐖(𝐂𝐦, 𝐩𝐜)= 𝐖(𝐂𝐦, 𝟎)+𝐖(𝐂𝐦, 𝟎) 𝐱𝐩 (5)
ther illustrated that proximity to the river increased the
vulnerability to flooding.
W (Cm,0) denotes the primary criterion’s weight, Cm in Less impact was observed on flooding from parameters
the case of a base run, and pc denotes the percent change’s such as curvature, NDVI, TWI, land use/land cover (LULC),
level. and soil type. The curvature map indicated that concave
In order to alter the weights of the other criterion propor- areas were more susceptible to flooding as they tended
tionally at any percent change level, Eqs. 6 and 7 were used; to retain water. The NDVI map revealed that regions with
as a consequence, all criteria weights must add up to one. negative NDVI values (indicating very poor or no vegeta-
tion, almost bare land) were at risk of flooding as there
∑
𝐧
𝐖(𝐩𝐜)= 𝐖(𝐂𝐢, 𝐩𝐜)=𝟏, 𝐑𝐏𝐂𝐦𝐢𝐧≤ 𝐩𝐜≤𝐑𝐏𝐂𝐦𝐚𝐱 (6) were no obstacles to impede the flow of water, potentially
𝐢=𝟏 resulting in significant soil erosion and landslides. Areas
with high TWI values indicated a higher water accumu-
W (Ci, pc) is the weight of the ith criterion ci at a certain lation capacity, thereby increasing the risk of flooding.
percent change level, Concerning LULC, areas occupied by water bodies and
W(Ci, pc) = (1 − W(Cm, pc)x W(Ci, 0)∕(1 − W(Cm, 0) (7) vegetation near rivers were more prone to flooding com-
pared to other land cover conditions. Clayey soil, due to
its low drainage rate compared to other soil types, had an

Fig. 10  Selected parameters 45

and their impact comparison
for study area 40

35
% Weightage

max. result min.

30

25

20

15

10

0
Rainfall Dsitance Slope Elevation LULC TWI NDVI Soil type Curvature
to river

Vol:.(1234567890)
SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1 Research

The flood hazard map was generated by considering

nine parameters: elevation, slope, rainfall, distance to river,
curvature, NDVI, TWI, LULC, and soil type. The AHP process
was employed to determine their relative weights, and it
yielded a consistency ratio (CR) of 5.4%, which is below the
acceptable threshold of 10%. This indicates the reliability
of the results. Table 1 presents the recommended weights
and ratings assigned to each of the nine criteria. The flood
hazard levels were categorized into five distinct classes:
very low, low, moderate, high, and very high. Using the
GIS-AHP approach, higher weights were assigned to rain-
fall, distance to river, elevation, and slope in comparison to
NDVI, TWI, LULC, curvature, and soil type. The flood hazard
map was then reclassified for each parameter. Overlaying
these maps determined that 5.6% of the total area is clas-
sified as very high flood risk, 52% as high risk, 39.3% as
moderate risk, and 3.1% as low risk. Further details on the
breakdown of the area categorized under different risk
zones can be found in Table 7.

3.1 Flood hazard map

Flood hazard map was produced from the weighted over-

lay of nine parameters. The standard deviation approach
was employed for the classification of the maps. The study
Fig. 11  Flood hazard map from AHP method for study area region is divided into four groups for the flood hazard
assessment: "Very high," "High," "Moderate," and "Low"
flood risk zones, as illustrated in Fig. 11. The high flood
increased potential to retain water, thereby elevating the risk zone covers the Southeast and Southwest parts of the
flood risk in clayey soil areas. study area, which is the most populated area in Swat dis-
trict. The flood risk is either moderate or low in the middle

Fig. 12  Validated map of Study Area

Vol.:(0123456789)
Research SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1

Table 6  Estimated flood-prone area of Swat reduced, ensuring the safety and protection of the popula-
Study area (District) 2
Total area (­ km ) 2
Area under high risk ­(km ) tion and infrastructure in the high-risk areas.

Swat 5,337 3074.112

3.2 Model validation

Table 7  Classification of flood prone areas in form of categories Model validation is a critical process that involves sys-
Criteria (Flood risk) Area(km2) Percentage
tematically comparing the outputs of a model to inde-
pendent observations in the real world. This comparison
Very High 298.872 5.6% helps assess the accuracy and reliability of the model’s
High 2775.24 52% predictions in terms of both quantity and quality. In the
Moderate 2092.104 39.3% context of flood susceptibility assessment, researchers
Low 165.447 3.1% utilize various models to evaluate the vulnerability of dif-
Total 5337 100% ferent regions to flooding. However, it is essential to vali-
date these models by comparing their outputs to actual
observed data or ground truth information. By calibrating
of the map. The flood risk is high for the Northeast and and validating the model using observable data, research-
Northwest sides as those areas are susceptible to flash ers can ensure that the model accurately reflects the real-
floods. world conditions and improves its reliability for future
The Southern part of Swat valley, where the majority of flood risk assessments.
the urban population resides and major road infrastruc- In September 2011, a disaster risk mapping project was
ture is located, is identified as being at high risk of flood- conducted in Swat District under the supervision of the
ing according to the flood hazard map. This is a significant National Rural Support Program (NRSP) and Diakonie, a
concern as it suggests that a significant loss of infrastruc- German-based INGO. The objective was to assess the vul-
ture may occur during flooding events, as observed in nerability of 11 selected villages in the district to flood-
the 2010 flood incident [40]. Another crucial factor influ- ing. These villages were chosen based on their extensive
encing the flood risk is rainfall. The speed of water flow damages during the 2010 flood event [41]. The geographic
downstream plays a vital role in determining the extent locations of these villages were plotted on the current
of flood damage. The slope and elevation of the terrain flood hazard map developed in this study. It was found
greatly influence the velocity and direction of water flow that all of the selected villages fell within the moderate
within the study area, as gravity plays a significant role to high flood risk zones, thus validating the study’s find-
in flood occurrences. As rainfall is relatively higher and ings with field observation data. Another study titled
more frequent in the Northern parts of Swat valley, the "Flood Disasters and Land Use Planning in Swat Valley"
rainwater gains velocity as it flows down the slope, result- was conducted by Atta-ur-Rehman et al. This study aimed
ing in greater damage in the Southern areas of the valley. to analyze flood disasters and the effectiveness of land
The methods and techniques used to develop the hazard use planning. They also developed a flood hazard zona-
map have the advantage of providing a comprehensive tion map using GIS tools [40]. A comparison was made
and easy-to-understand representation of the flood haz- between their flood hazard zonation map and the one
ards. This hazard map can be particularly useful in areas developed in the current study. Although the emphasis of
where primary data is unavailable. Tables 6 and 7 present their study was more on land use analysis, the results were
the calculated categorization of the area into flood risk comparable. However, the current study incorporated nine
zones, ranging from high to low. different parameters into the development of the flood
Table 7 indicates that over 50% of the area is classified as hazard map, making it more comprehensive and reliable.
having a high flood risk. To mitigate the negative impacts The comparison of both flood hazard maps validated the
of flooding, it is crucial to implement strategic measures results of the current study and indicated their reasona-
for planned construction in these areas. Encroachments, bleness. Swat Valley, like many other districts in Pakistan,
particularly those near the river banks, should be removed experiences periodic severe flooding [42]. Areas located
to prevent further vulnerability. Constructing a flood pro- along the river banks have been particularly vulnerable
tection wall is recommended in regions where it will effec- to floods since 1973 [43]. Figure 12 shows illustrations of
tively mitigate the risk, with the southern site being the some flood events superimposed on the flood hazard map
most feasible option. By implementing these measures, developed in the current study, highlighting the accuracy
the potential damage from flooding can be significantly of the hazard map in depicting the flood-prone areas.

Vol:.(1234567890)
SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1 Research

3.3 Limitations of the methodology geographical location and topographic characteristics

and recommendation respectively. Hence, it was selected as the research area
for the study.. Based on the data, the extent of each dan-
While the current study has its limitations, such as the ger zone varies as the rainfall factor increases. As rainfall
quality of the medium resolution image (SRTM 30 m) increases, the intensity of floods also increases. However,
used and the subjectivity in assigning coastlines to certain the significance of the rainfall peak coefficient parameter
parameters, the obtained flood susceptibility map of the in determining the position of the greatest danger zone
Swat District KPK watershed remains a valuable tool for decreases as the return period of precipitation increases
development, planning, and decision-making by admin- [44]. The urban region constitutes approximately 33% of
istrative authorities and decentralized territorial commu- the entire study area. During intense storm events, when
nities. It will contribute to improving the quality of life for the runoff exceeds the capacity of the drainage system, six
the densely populated area and addressing environmen- factors were identified as significant contributors to water-
tal management and sanitation issues. It is important to course-related issues. These factors include slope, eleva-
acknowledge the limitations and inconsistencies in future tion, proximity to open channel streams, hydro-lithology,
studies, incorporating them into the analysis and explor- distance from fully covered streams, and human-induced
ing more effective models to reduce error margins. This factors such as urban areas and additional land uses. The
study marks the first attempt of its kind to assess the vul- AHP technique and GIS were employed to develop a flood
nerability of the Swat District KPK watershed to flooding hazard assessment map, which assesses the likelihood of
in the context of global change, utilizing a wide range of a flood occurring within a specific timeframe. However,
environmental data including SRTM, rainfall, and geology. the current study did not consider the evaluation of flood
These findings serve as a starting point in addressing the hazards in relation to return times [26]. As a result, a flood
flooding problems in the Swat area of Pakistan. Further- susceptibility map was created to indicate the likelihood of
more, the outcomes of this study will facilitate coordina- flooding in specific areas. This map classified the potential
tion among the government, decentralized territorial com- flood hazards in the study region into five levels of sus-
munities, and other development actors in their efforts to ceptibility: very low, low, moderate, high, and very high.
prevent, mitigate, and respond to floods. From a scientific According to Table 7, approximately 40% of the region falls
perspective, this endeavor enhances our understanding of within two high-risk zones. In Swat, the fluvial systems are
the ecology of the Swat District KPK watershed and ena- mainly composed of fast-flowing streams, which can lead
bles the identification and evaluation of variables influenc- to severe but infrequent flash floods. The presence of small
ing flood risk. drainage basins and the unpredictable behavior of streams
in the area contribute to the occurrence of floods. Runoff
tends to increase in the upper parts of watersheds, which
4 Discussion raises the risk of flooding in urban areas downstream. By
assigning weights to each element or indicator involved in
The study’s findings demonstrated the efficiency of the the flood process, the AHP approach enhances our under-
Analytic Hierarchy Process (AHP) and Geographic Infor- standing of their individual roles [45].
mation System (GIS) tools in determining the danger of However, the use of data from multiple sources, inter-
urban flooding in Pakistan’s Swat District’s KPK water- polation, and cross-referencing in a GIS at the same resolu-
shed. Elevation slope and rainfall were shown to be the tion can introduce bias when processing and analyzing the
most significant factors influencing flood risk using the data. To mitigate bias and uncertainty in the final results,
AHP technique. In the meanwhile, flood risk maps were it is necessary to normalize and weight these parameters
produced, and the findings were shown using GIS tools. appropriately. The subjective nature of selecting indicator
The survey also discovered that although the majority of weights based on random expert judgments is another
people in the KPK watershed believed that there was a reason why the AHP approach may not be effective. In
significant danger of flooding, very few of them had taken general, an integrated approach combining AHP and
any precautions. The GIS maps revealed that low-lying GIS techniques can yield satisfactory outcomes [46]. The
locations and those with a lot of impermeable surfaces, proposed technique offers a comprehensive analysis of
such as roads and buildings, were those most at danger flood-prone areas using consistent methodologies [34,
of flooding. Extreme rainfall events have become more 47]. The findings provide valuable information that can
frequent and severe due to climate change and increased be applied in different contexts. Therefore, this concept
urbanization, resulting in regular urban flooding disasters. is beneficial when combined with probability attribution
Swat District, being Pakistan’s most populated district, is methodologies. However, there are challenges in imple-
susceptible to urban floods and flash floods due to its menting this technique. In our study, due to a lack of data

Vol.:(0123456789)
Research SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1

on the underground drainage system, we employed a these stakeholders to make informed decisions about
simplified method of flooding simulation by generalizing land use planning, emergency response strategies, and
the network. While there were some errors in the simula- targeted mitigation measures. By utilizing our findings,
tion results, they did not significantly impact the assess- they can proactively enhance the district’s resilience and
ment of flood risk levels [48]. He specific objectives of the promote sustainable development, effectively minimizing
investigation determine the required level of resolution the impact of flood hazards.
for considering physical, social, and environmental factors.
To enhance the comprehensibility of the findings, data is
presented in a raster format of the same size as the text. 5 Conclusions
Future studies should aim to employ indicators with con-
sistent spatial resolution. Due to the concentration of lowlands with moderate
Measuring urban flood risk consistently is difficult due slopes, ephemeral streams with torrential behavior,
to the dynamic nature of the factors involved, such as pre- completely covered streams, and urban areas, the east-
cipitation intensity and volume, population density, and ern and southern parts of the research region exhibit a
building layout. In our research, we focused on examining high concentration of very high and high flood danger
and analyzing a subset of critical indicators relevant to the zones. The locations of these flood-risk zones showed
research topic. Despite the inherent uncertainty, it is cru- relatively little variation, as indicated by the uncertainty
cial to precisely and thoroughly compute the weighting analysis. The flood danger map effectively illustrates
of these indicators as it significantly impacts the final esti- the relationship between flood hazard zones and areas
mation of flood risk. Researchers have extensively studied prone to flooding. The validation process demonstrated
the challenges faced during rescue operations following the reliability and precision of the hazard map. The cur-
natural disasters. In the future, a comprehensive investiga- rent urban flood hazard map and spatial distribution
tion of urban flooding is necessary to establish scientific were analyzed to identify potential sites for implement-
criteria for urban emergency rescue systems [49]. ing flood protection measures. The proposed method,
The findings of this study have several implications for which provides a comprehensive analysis of the area, can
flood risk management in the KPK watershed and neigh- be easily implemented in locations where primary data
boring areas. First and foremost, the research highlights is limited and produces reasonable results. Therefore,
the significance of land use planning and management in scientists, stakeholders, engineers, and decision makers
mitigating flood risk. The presence of numerous roads and involved in future land use planning initiatives can adopt
impermeable surfaces in the study area leads to increased this method. It can also contribute to the development
runoff during rainfall events, increasing the likelihood of of flood mitigation techniques and be applied to disas-
floods. Therefore, it is crucial to make informed land use ter preparedness planning and post-fire management,
decisions to minimize the risk of flooding. The research enhancing their effectiveness.
highlights the need for improved drainage infrastructure
and maintenance to mitigate the impacts of floods. It Acknowledgements The ESRI and USGS are gratefully acknowledged
by the study’s authors for contributing the digital elevation model
was observed that the drainage system in the study area and land use and cover data, respectively.
was inadequate, which heightened the risk of flooding.
Therefore, investing in drainage infrastructure and main- Author Contributions Conceptualization, MW, SA, IA, and HW; meth-
tenance can help minimize the impacts of floods in the odology, SA, IA, HW, and MW; software, IA, SA, MW, and HW; valida-
future. Furthermore, the study emphasizes the importance tion, IA, MW, SA, and HW; formal analysis, SA, MW, IA, and HW; investi-
gation, MW; data curation, SS, IA, HW, and MW; writing original draft,
of educating and increasing public awareness regarding SA, IA; writing—review and editing, MW, and MKL; visualization, MW,
flood risk and preparedness. Although many people in and MKL. All authors have read and agreed to the submitted version
the study region acknowledged the significant likelihood of the manuscript.
of flooding, only a few had taken any preparatory meas-
ures. To mitigate the impacts of floods, initiatives should Funding Open Access funding enabled and organized by Projekt
DEAL. This study received no external funding. We are thankful to
be undertaken to enhance public awareness of flood risk the University of Rostock, Open access program for their willingness
and preparedness. to pay the article processing charges.
The significance of this research lies in its practical
implications. The hazard maps produced by our study Data availability The datasets generated and/or analyzed during the
serve as valuable tools for urban planners, local authori- current study are not publicly available due to privacy reasons but
are available from the corresponding author on reasonable request.
ties, and disaster management agencies. They enable

Vol:.(1234567890)
SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1 Research

Declarations and its practical applications : The AHP/ANP approach. Opera-

tions Res. https://​doi.​org/​10.​1287/​opre.​2013.​1197
Competing interests The authors declare no competing interests. 11. A. Lahrach, 2019 Water security and sustainable develop-
ment * Zineb Moumen 1, Najiba El Amrani El Idrissi 2, Manuela
Ethical approval Not applicable. Tvaronavičienė 3 Abderrahim Lahrach 4 14 1
12. Tabari H (2020) Climate change impact on flood and extreme
Consent to participate Not applicable. precipitation increases with water availability. Sci Rep. https://​
doi.​org/​10.​1038/​s41598-​020-​70816-2
Consent for publication Not applicable. 13. Al-Aizari AR et al (2022) “Assessment analysis of flood suscep-
tibility in tropical desert area: a case study of Yemen. Remote
Open Access This article is licensed under a Creative Commons Attri- Sens. https://​doi.​org/​10.​3390/​rs141​64050
bution 4.0 International License, which permits use, sharing, adap- 14. Allen MR, Ingram WJ (2002) Constraints on future changes in
tation, distribution and reproduction in any medium or format, as climate and the hydrologic cycle. Nature. https://​doi.​org/​10.​
long as you give appropriate credit to the original author(s) and the 1038/​natur​e01092
source, provide a link to the Creative Commons licence, and indicate 15. Muzzamil S et al (2020) A review of the flood hazard and risk
if changes were made. The images or other third party material in this management in the South Asian Region, particularly Pakistan.
article are included in the article’s Creative Commons licence, unless Sci. African 10:e00651. https://​doi.​org/​10.​1016/j.​sciaf.​2020.​
indicated otherwise in a credit line to the material. If material is not e00651
included in the article’s Creative Commons licence and your intended 16. Chauhdry KT, Javed U (2019) Climate change and water secu-
use is not permitted by statutory regulation or exceeds the permitted rity: focus on pakistan. ISSRA Papers 11(II):57–70
use, you will need to obtain permission directly from the copyright 17. Nsangou D et al (2022) Urban flood susceptibility modelling
holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​ using AHP and GIS approach: case of the Mfoundi water-
org/​licen​ses/​by/4.​0/. shed at Yaoundé in the South-Cameroon plateau. Sci African
15:e01043. https://​doi.​org/​10.​1016/j.​sciaf.​2021.​e01043
18. Dottori F et al (2018) Increased human and economic losses
from river flooding with anthropogenic warming. Nat Clim
References Chang. https://​doi.​org/​10.​1038/​s41558-​018-​0257-z
19. Nofal OM, Van De Lindt JW (2021) International Journal of Disas-
1. Wingfield T, Macdonald N, Peters K, Spees J, Potter K (2019) ter Risk Reduction High-resolution flood risk approach to quan-
Natural Flood Management: Beyond the evidence debate. tify the impact of policy change on flood losses at community-
Area. https://​doi.​org/​10.​1111/​area.​12535 level. Int J Disaster Risk Reduct 62:102429. https://​doi.​org/​10.​
2. Munawar S et al (2022) Future climate projections using SDSM 1016/j.​ijdrr.​2021.​102429
and LARS-WG downscaling methods for CMIP5 GCMs over the 20. Ullah K, Id JZ (2020) GIS-based flood hazard mapping using
transboundary jhelum river basin of the Himalayas Region. relative frequency ratio method : A case study of Panjkora River
Atmosphere (Basel). https://​doi.​org/​10.​3390/​atmos​13060​898 Basin eastern Hindu Kush Pakistan. PLoS ONE. https://​doi.​org/​
3. Arnell NW, Gosling SN (2016) The impacts of climate change 10.​1371/​journ​al.​pone.​02291​53
on river flood risk at the global scale. Clim Change 134(3):387– 21. Sinha R (2005) GIS in flood hazard mapping : a case study of Kosi
401. https://​doi.​org/​10.​1007/​s10584-​014-​1084-5 River Basin India. GIS Develop Weekly. https://d ​ oi.o
​ rg/1
​ 0.1
​ 3140/​
4. CRED, “2022 Disaster in Numbers,” 2023, [Online]. Available: RG.2.​1.​1492.​2720
https://​cred.​be/​sites/​defau​lt/​files/​2022_​EMDAT_​report.​pdf 22. Bhatt GD, Sinha K, Deka PK, Kumar A (2014) Flood Hazard and
5. Hanson S et al (2011) A global ranking of port cities with high Risk Assessment in Chamoli District, Uttarakhand Using Satel-
exposure to climate extremes. Clim Change 104(1):89–111. lite Remote Sensing and GIS Techniques. Int J Innov Res Sci Eng
https://​doi.​org/​10.​1007/​s10584-​010-​9977-4 Technol 3(8):15348–15356. https://​doi.​org/​10.​15680/​IJIRS​E T.​
6. Adb, Addressing Climate Change and Migration in Asia and the 2014.​03080​39
Pacific. 2012. [Online]. Available: http://​w ww.​adb.​org/​sites/​ 23. Anthropiques P, Les SUR, La MDE (2012) Impacts conjugués
defau​lt/​files/​publi​cation/​29662/​addre​ssing-​clima​te-​change-​ des changements climatiques et des pressions anthropiques
migra​tion.​pdf sur les modifications de la couverture végétale dans le bassin
7. M. Drdácký, L. Binda, I. Herle, and L. Lanza, 2007 Protecting the versant du n’zi -bandama (cô te d’ivoire). Rev Ivoir Sci Technol
cultural heritage from natural disasters, Adv. Res. Cent. Cult. 20:124–146
Herit. Feb, p. 100, , [Online]. Available: http://​schol​ar.​google.​ 24. Rinc D, Khan UT, Armenakis C (2018) Flood risk mapping using
com/​schol​ar?​hl=​en&​btnG=​Searc​h&q=​intit​le:​Prote​cting+​the+​ gis and multi-criteria analysis : a greater toronto area case study.
cultu​r al+​h erit​a ge+​f rom+​n atur​a l+​d isas​t ers#3%​5 Cnht​t p://​ Geosciences. https://​doi.​org/​10.​3390/​geosc​ience​s8080​275
www.​europ​arl.​europa.​eu/​activ​ities/​commi​ttees 25. Guo E, Zhang J, Ren X, Zhang Q (2014) Integrated risk assess-
8. World Bank and R. T. Government, 2011 Rapid Assessment ment of flood disaster based on improved set pair analysis and
for Resilient Recovery and Reconstruction Planning, p.377, the variable fuzzy set theory in central Liaoning Province China.
[Online]. Available: http://​www.​gfdrr.​org/​sites/​gfdrr.​org/​files/​ Nat Hazards. https://​doi.​org/​10.​1007/​s11069-​014-​1238-9
publi​cation/​Thai_​Flood_​2011_2.​pdf 26. Hussain M, Butt AR, Uzma F (2020) A comprehensive review of
9. Ullah K, Wang Y, Fang Z, Wang L, Rahman M (2022) Multi-haz- climate change impacts adaptation and mitigation on environ-
ard susceptibility mapping based on convolutional neural mental and natural calamities in Pakistan. Environ Monit Assess.
networks. Geosci Front. https://​doi.​org/​10.​1016/j.​gsf.​2022.​ https://​doi.​org/​10.​1007/​s10661-​019-​7956-4
101425 27. Ouma YO, Tateishi R (2014) Urban flood vulnerability and risk
10. Saaty TL, Saaty TL (2013) The modern science of multicriteria mapping using integrated multi-parametric ahp and gis: meth-
decision making and its practical applications : The AHP / ANP odological overview and case study assessment. Water. https://​
approach the modern science of multicriteria decision making doi.​org/​10.​3390/​w6061​515
28. Hagos YG, Andualem TG, Yibeltal M, Mengie MA (2022) Flood
hazard assessment and mapping using GIS integrated with

Vol.:(0123456789)
Research SN Applied Sciences (2023) 5:215 | https://doi.org/10.1007/s42452-023-05445-1

multi - criteria decision analysis in upper Awash River basin. Hindu Kush. Land use management in disaster risk reduction.
Appl Water Sci. https://​doi.​org/​10.​1007/​s13201-​022-​01674-8 Springer, Tokyo, pp 179–195
29. Sweeney G, Hand M, Kaiser M, Clark JK, Rogers C, Spees C (2015) 41. A. Mathematics, 2016 Disaster Risk Mapping District Swat 1–23
The state of food mapping : academic literature since 2008 and 42. “FloodsIn Pakistan Maps, Graphs And Key Data | Critical Threats.”
review of online gis-based food mapping resources. J Plan Liter. https://​w ww.​criti​calth​reats.​org/​analy​sis/​floods-​in-​pakis​tan-​
https://​doi.​org/​10.​1177/​08854​12215​599425 maps-​graphs-​and-​key-​data accessed May 11, 2023
30. Patrikaki O, Kazakis N, Kougias I, Theodossiou N, Voudouris K, 43. “As Floodwaters Recede In Pakistan, Swat Valley Residents Come
Patsialis T (2018) Assessing flood hazard at river basin scale with To Grips With Climate Change .” https://​www.​r ferl.​org/a/​pakis​
an index-based approach : the case. Geosciences. https://​doi.​ tan-​floods-​swat-​clima​te-​change/​32055​964.​html Accessed May
org/​10.​3390/​geosc​ience​s8020​050 11, 2023
31. Bengal W, Chakraborty S, Mukhopadhyay S (2019) Assessing 44. Li G, Liu J (2022) Flood risk assessment using TELEMAC-2D mod-
flood risk using analytical hierarchy process ( AHP ) and geo- els integrated with multi-index analysis in Shenzhen River Basin
graphical information system ( GIS ): application. Nat Hazards. China. Water 14(16):2513
https://​doi.​org/​10.​1007/​s11069-​019-​03737-7 45. Schoenherr T, Tummala VMR, Harrison TP (2008) Assessing sup-
32. Region L, Wang Y, Li Z, Tang Z (2011) A GIS-based spatial multi- ply chain risks with the analytic hierarchy process : providing
criteria approach for flood risk assessment in the dongting. decision support for the offshoring decision by a US manufac-
Water Res Manage. https://​doi.​org/​10.​1007/​s11269-​011-​9866-2 turing company. J Purch Supply Manage 14:100–111. https://​
33. Tayyab M et al (2021) Gis-based urban flood resilience assess- doi.​org/​10.​1016/j.​pursup.​2008.​01.​008
ment using urban flood resilience model: A case study of pesha- 46. Boroushaki S, Malczewski J (2010) Computers & geosciences
war city khyber pakhtunkhwa Pakistan. Remote Sens. https://​ using the fuzzy majority approach for GIS-based multicriteria
doi.​org/​10.​3390/​rs131​01864 group. Comput Geosci 36(3):302–312. https://d ​ oi.o
​ rg/1
​ 0.1
​ 016/j.​
34. Z. Khatoon, S. Gulzar, A. Shah, and T. Education, Manuscript Info. cageo.​2009.​05.​011
1(2320): 329–344 47. Sinha R, Singh GVBLK, Rath B (2008) Flood risk analysis in the
35. “Earthdata Search | Earthdata Search.” https://​search.​earth​data.​ Kosi river basin, north Bihar using multi- parametric approach
nasa.​gov/​search Accessed Apr. 10, 2023 of analytical hierarchy process (AHP). J Indian Soci Remote Sens
36. “FAOMapCatalog.” https://​data.​apps.​fao.​org/​map/​catal​og/​srv/​ 36:335–349
eng/c​ atalo​ g.s​ earch;j​ sessi​ onid=A
​ 9D9F0
​ D0DE8 ​ 7CAE7 ​ 10645
​ 3B0B​ 48. Lin K et al (2020) Assessment of flash flood risk based on
93D1C9 ​ D?n​ ode=s​ rv#/s​ earch?f​ acet.q=t​ ype%2
​ Fdata​ set Accessed improved analytic hierarchy process method and integrated
Apr. 10, 2023 maximum likelihood clustering algorithm. J Hydrol 584:1–57.
37. “Esri | Sentinel-2 Land Cover Explorer.” https://​livin​gatlas.​arcgis.​ https://​doi.​org/​10.​1016/j.​jhydr​ol.​2020.​124696
com/​landc​overe​xplor​er/#​mapCe​nter=-​83.​21%​2C34.​332%​2C4&​ 49. Barroca B, Bernardara P, Mouchel JM, Hubert G (2006) Indicators
mode=​step&​timeE​x tent=​2017%​2C202​1&​year=​2017&​downl​ for identification of urban flooding vulnerability. Nat Hazards
oadMo​de=​true Accessed Apr. 10, 2023 Earth Syst Sci. https://​doi.​org/​10.​5194/​nhess-6-​553-​2006
38. “EarthExplorer.” https://​earth​explo​rer.​usgs.​gov/ Accessed Apr.
10, 2023 Publisher’s Note Springer Nature remains neutral with regard to
39. “CRU TS Version 4.06.” https://​cruda​ta.​uea.​ac.​uk/​cru/​data/​hrg/​ jurisdictional claims in published maps and institutional affiliations.
cru_​ts_4.​06/ Accessed Apr. 10, 2023
40. Atta-ur-Rahman Farzana (2017) G. Rahman, and R. Shaw, 2017
Flood Disasters and Land Use Planning in Swat Valley Eastern

Vol:.(1234567890)