Integration hierarchical analysis process (AHP) and geographic information systems to map flood risk levels in the province of Utcubamba, Peru

Abstract

Floods are one of the most common and destructive natural disasters globally, causing significant economic losses and loss of life. In Peru, this problem is particularly critical, affecting nearly 1.51 million people living in flood-prone areas. This study aims to identify and map flood-prone areas in the province of Utcubamba, Amazonas, Peru, using a semi-quantitative approach that integrates the Analytic Hierarchy Process (AHP) with Geographic Information System (GIS) tools and Weighted Linear Combination (WLC). Fifteen conditioning variables grouped into five factors—topographic, geomorphological, hydrological, climatic, and anthropogenic—were selected, weighted, and spatially combined. The AHP consistency ratio (CR = 0.076) confirmed the methodological validity of the model. Results indicate that approximately 401.39 km² (10.09%) of the provincial territory presents very high flood risk, concentrated primarily in Bagua Grande, the most populated city in Amazonas, while 824.25 km² (22.47%) correspond to high risk. Model validation was conducted using two complementary approaches: (1) the Relative Mean Error (RME) and Root Mean Square Error (RMSE) method applied to 585 georeferenced flood occurrence records from the INDECI/SINPAD national emergency database, yielding RME = 12.82% (good performance) and RMSE = 0.7845; and (2) Receiver Operating Characteristic (ROC) curve analysis, which produced an Area Under the Curve (AUC) of 0.816 (95% CI: 0.721–0.895), confirming good discriminatory capacity of the model. These results underscore the urgent need for territorial planning and disaster prevention measures in the region, while demonstrating the robustness and replicability of the AHP-GIS methodology for flood risk assessment in data-scarce tropical regions.

1 Introduction

Natural disasters are considered the greatest challenge to be examined at the global, regional, and local levels (Danumah et al., 2016). In recent decades, its frequency has increased considerably due to various human and natural factors, such as environmental degradation, climate change, rapid population growth, and intensified and inappropriate land use (Noori and Bonakdari, 2023). According to a survey conducted by the World Economic Forum for the Global Risks Report 2025, extreme weather events rank second among global risks in the last 2 years and first in the last decade (Foro Económico Mundial, 2025). Floods are one of the natural disasters with the greatest impact worldwide, causing significant and irreversible damage to property and communication infrastructure (Chaudhary et al., 2021). They also result in the loss of numerous human lives and large areas of farmland and livestock, generating social, economic, and environmental consequences (Swain et al., 2020).

The increase in the intensity and frequency of rainfall and cyclonic events is a determining factor in the occurrence of disasters such as floods (Xu et al., 2022). This risk is intensified by exposure to anthropogenic actions such as unplanned urbanization and the expansion of settlements in flood-prone areas, which increase vulnerability and therefore adverse impacts (Cabrera and Lee, 2019). Risks are thus intensified during climatic events such as El Niño, which have negative and long-term impacts on livelihoods, especially in developing countries where vulnerability is linked to poverty (Cobian Alvarez and Resosudarmo, 2025); Peru is one of the most affected countries, with approximately 1.51 million people residing in areas susceptible to flooding (Adger et al., 2004; Smith et al., 2021). Between 1981 and 2021, there were 414 floods that destroyed 42,628 homes and affected another 75,680. The most catastrophic of these occurred in 2017 with the coastal El Niño event, which caused extreme rainfall, severe flooding, the loss of 107 lives, and significant economic disruption in northern Peru (Sardon et al., 2022).

Climate change is expected to further intensify flood risk in tropical mountain basins by modifying rainfall intensity, hydrological response, and the frequency of extreme precipitation events. Recent studies based on downscaled CMIP6 climate projections indicate that flood-prone areas may expand or intensify under future warming scenarios, particularly in tropical watersheds affected by climatic variability, steep terrain, and rapid land-use change (Hidalgo et al., 2024). In this context, multi-criteria approaches that incorporate climate-change conditions provide a useful framework for anticipating future flood susceptibility and supporting adaptive territorial planning, as shown in recent applications in Central America, including Panama (Petiangma et al., 2025).

Flood susceptibility in tropical mountain basins is controlled by the interaction between rainfall, terrain morphology, drainage organization, slope dynamics, lithology, land use, and exposure patterns. In these environments, steep slopes, incised channels, alluvial terraces, and floodplains favor rapid runoff concentration and spatially heterogeneous flood behavior, making geomorphological analysis essential for identifying areas prone to water accumulation, channel overflow, and flood propagation (Quesada-Román, 2021). This perspective is particularly relevant in tropical regions with limited hydrometeorological data, rapid urban expansion, climatic variability, and increasing disaster risk, where geomorphological and morphometric information derived from digital elevation models, remote sensing, and field interpretation helps explain the spatial distribution of flood-prone areas (Quesada-Román and Peralta-Reyes, 2023). Accordingly, integrated geomorphological approaches provide a methodological basis for flood and landslide risk mapping by combining terrain analysis, hydrological indicators, environmental variables, and socio-territorial factors within spatial decision-support frameworks (Acosta-Quesada and Quesada-Román, 2024).

Due to the complexity of floods, their high frequency of occurrence, and the magnitude of the damage they cause worldwide, there has been increased scientific attention aimed at understanding the interactions between hazard, exposure, and vulnerability in order to support more effective mitigation and management strategies (Dandapat and Panda, 2017; Gain et al., 2015; Guo et al., 2014). In this context, it is important to be able to assess flood risk and develop maps that can be used to support various applications, such as land use planning and infrastructure location (Zhang et al., 2020); Remote sensing and geographic information systems (GIS) are key tools that contribute to risk management assessment and geospatial analysis, as well as facilitating the integration of multi-criteria decision models in the identification and evaluation of susceptible areas (Das et al., 2017; Das et al., 2018).

In Central America, disaster risk reduction requires strengthening risk assessment at various territorial scales, as well as improving data availability, hazard mapping, and exposure and vulnerability studies to guide public decision-making (Quesada-Roman and Peralta-Reyes 2023). In this vein, Quesada-Román et al. (2025a) conducted a disaster risk analysis in Panama, demonstrating the utility of historical databases and the spatial identification of the most affected districts to support risk management, territorial planning, and environmental management at the municipal and regional levels. Similarly, in Nicaragua (Quesada-Román et al., 2026a), they show that the spatiotemporal patterns of risk and its socioeconomic determinants allow for the identification of critical municipalities and areas with the highest incidence of disasters, providing relevant evidence for developing tropical countries with limited baseline information.

In this context, various studies have shown that the integration of geographic information systems (GIS) and remote sensing techniques provides a solid basis for the development of flood risk maps (Lyu et al., 2025). In addition to this, statistical and machine learning methods are also used, which have expanded the possibilities for analysis and improved the accuracy in the delimitation of susceptible areas (Hadipour et al., 2020). Among these approaches, the following stand out: frecuency ratio (Rahmati et al., 2016), weight of evidence (Tehrany et al., 2014), analityc hierarchy process (AHP)(Gigović et al., 2017; Kazakis et al., 2015), logistic regression (Youssef et al., 2016), multicriterial decisión making (Wang et al., 2020), fuzzy weight of evidence (Hong et al., 2018), artificial neural networks (Wang et al., 2020), support vector machine (Tehrany et al., 2015), decisión tree (Khosravi et al., 2018), random forest (Chen et al., 2020) and ensemble models (Costache et al., 2020).

While the application of machine learning algorithms for flood susceptibility mapping is increasing, the drawback of using these methods is that they require a large volume of data and have been implemented primarily in environments with abundant historical flood records, real-time monitoring data, and well-documented infrastructure inventories. For example, some studies have applied a machine learning and GIS framework to predict flood susceptibility in urban metro systems under extreme rainfall scenarios (Lyu et al., 2025). Similarly, improved analytical network process (ANP) approaches integrated with GIS have been used to assess sensitivity to typhoon-induced flooding in coastal cities (Yariyan et al., 2020). In contrast, the province of Utcubamba, like other places in Peru, presents an environment with a scarcity of data, with limited hydrometeorological records, scarce census data, and a predominantly rural and agricultural population exposed to river flooding caused by orographic precipitation and the morphology of the inter-Andean valleys.

Given this, GIS-based AHP with weighted linear combination (WLC) provides a suitable, transparent, and replicable framework, as it integrates expert judgment with available spatial data without requiring extensive historical flood records for model training. The analytic hierarchy process (AHP) is one of the most influential techniques for evaluating multiple criterial base don their relative performance and expert opinion (Rajasekhar et al., 2019; Vahidnia et al., 2009). The synergistic effect of GIS and AHP improves the efficiency and results of flood analysis (Tripathi et al., 2020). GIS can be useful in disseminating flood-related data and alerts using spatial data infraestructura and web GIS (Agrawal and Gupta, 2017).

That being the case, recent research has applied multi-criteria decision-making methods, such as the Analytic Hierarchy Process (AHP) and its variants, to combine physical, environmental, and social factors in different watersheds and territories (Mokhtari et al., 2023; Rana et al., 2024; Taoukidou et al., 2025). These approaches have made it possible to identify areas with different levels of susceptibility to flooding, generating zoning maps that facilitate territorial planning and preventive management (Burayu et al., 2023). Likewise, the incorporation of validation techniques, such as ROC-AUC curves, has reinforced the reliability of the results and demonstrates that the integration of spatial information and multi-criteria analysis techniques are tools that can be used to address the problem of flooding in different contexts and opinions (Abdo et al., 2024; 2025).

In Peru, research on flooding using advanced methodological and technological approaches is still limited; however, in recent years there has been growing interest in incorporating innovative tools, driven by the problems that these events cause in various regions of the country (Caramanica et al., 2020; Drenkhan et al., 2015; Zubieta et al., 2017). Within this framework, studies have been developed that integrate hydrological models, satellite images, and multi-criteria analysis in GIS, which are essential given the high recurrence of floods, closely related to intense rainfall and the occurrence of the coastal El Niño phenomenon, whose impacts significantly affect infrastructure, crops, and the safety of communities (Ali Al-Rawas et al., 2025; Badillo-Rivera et al., 2025; Lavado-Casimiro et al., 2023).

This research is being conducted in the province of Utcubamba, department of Amazonas, a territory that has experienced marked urban growth in recent years. This expansion process has been oriented toward areas with low soil capacity and areas close to the Utcubamba River, which increases the vulnerability of infrastructure and the population to flooding (Plan de Desarrollo Urbano de Bagua Grande, 2014–2025). The province’s main tributary, the Utcubamba River, is undergoing a process of progressive degradation of the ecological quality of its waters as a result of intensive agricultural and livestock activities, posing risks to both riparian ecosystems and the health of local communities (Gamarra et al., 2018). Added to this are the hydrogeomorphological alterations resulting from urban and anthropogenic pressure, especially on its banks, where the riverbed has undergone modifications and deterioration (Barboza et al., 2017).

In this regard, the novelty of this study lies in the fact that it is the first systematic mapping of flood risk for the province of Utcubamba at the municipal level using an AHP-WLC model, the integration of various variables that strengthen the model and leave no methodological gaps, as well as the rigorous statistical validation that makes the model reliable and replicable in other areas with flooding problems and limited data. Taking the above into account, the objective is to assess and map flood risk areas, integrating geospatial databases from various platforms using a semi-quantitative approach, using Geographic Information System (GIS) tools in conjunction with the Analytic Hierarchy Process (AHP), which will allow for the weighting and classification of factors that influence the occurrence of floods, thus contributing to the identification of the most vulnerable areas.

2 Study area characteristics

The province of Utcubamba, located in the department of Amazonas, Peru, spans the coordinates 5°23′25“S to 6°10′53” S and 77°51′7“W to 78°42′12” W (Figure 1). The province covers approximately 3,850.93 km², with an estimated population of between 107,000 and 119,000 (INEI, 2018). Utcubamba is administratively divided into seven districts: Bagua Grande, El Milagro, Cajaruro, Cumba, Lonya Grande, Jamalca, and Yamón. The capital, Bagua Grande, situade at 440–446 m above sea level (INEI, 2014). Functions as the main urban and commercial center of the province and is located along the Utcubamba River, which drains northward into the Marañón River. This geographic configuration a narrow inter-Andean valley flanked by steep slopes makes the province particularly vulnerable to flood events, especially in the lower-lying floodplain áreas (Municipalidad Provincial de Utcubamba, 2013).

FIGURE 1

2.1 Characteristics climate and hydrology

The climate of the province of Utcubamba is warm and humid tropical, with variations depending on elevation. Lower-elevation areas experience a longer dry season, while higher-elevation areas tend to have more frequent rainfall. The heaviest rains occur between December and March, which corresponds to the period of rising river flows (Servicio Nacional de Meteorología e Hidrología del Perú, 2016).

The Utcubamba River is the province’s main hydrographic axis. Its banks, where towns such as Bagua Grande are located, have been exposed to continuous flooding caused by heavy rains and rising river levels, which have caused serious damage to homes, infrastructure, and agricultural areas (Instituto Nacional de Defensa Civil, 2021).

2.2 Geological context

Geologically, Utcubamba is located in the Andean zone of northern Peru, at the transition between the Cordillera Oriental and the Sub-Andean Belt. Outcrops of the Marañón Complex and the Mitu Group are found here, overlying Mesozoic sedimentary sequences such as the Pucará Group as well as units of the Oriente, Goyllarisquizga, Pulluicana, and Quilquiñan groups, as well as the Sarayaquillo, Chulec, Chonta, Celendín, Cajamarca, Cajaruro, Chota, Bellavista, and El Milagro formations, which are shown on the geological map of the study area (Figure 2A). Furthermore, the inter-Andean valleys are characterized by the presence of Pleistocene alluvial deposits and more recent fluvial deposits, whose permeability determines the soil’s hydrological behavior during extreme rainfall, surface runoff, and susceptibility to flooding (Medina Allcca et al., 2009).

FIGURE 2

2.3 Geomorphological characteristics

The geomorphology of the Utcubamba region is heterogeneous and is characterized by the Utcubamba-Bagua inter-Andean synclinal valley, which covers a large area, primarily around Bagua Grande (Castro Medina, 2010). At higher elevations, mountains and hills of structural and erosional control predominate, developed on sedimentary and metamorphic rocks; however, at the valley floor, colluvial slopes, foothill fans, and alluvial-torrential units also form, which favors surface runoff and water erosion (Figure 2B). In contrast, the valley floor consists of alluvial terraces, dry plains, alluvial plains, lagoons, and bodies of water. Similarly, alluvial plains, given their flat topography and low elevation relative to the river channel, combined with fine sediments, can be considered the area’s most susceptible to flooding during episodes of extreme high water. This difference between the rugged terrain and the alluvial plains is an important factor influencing river dynamics, runoff, and vulnerability within the basin (Castro Medina, 2010).

3 Metodology

3.1 Data acquisition and preparation

This study required extensive collection of spatial data from various sources to assess flood risk in the province of Utcubamba. After consulting with several experts and reviewing a considerable amount of relevant literature, 15 key indicators were selected to calculate flood risk. These variables are grouped into five factors according to their nature: topographic, geomorphological, hydrological, climatic, and anthropogenic (Amiri et al., 2024; Marth et al., 2016). When analyzed together, it is possible to determine the occurrence of floods, as they have a direct influence on events related to this phenomenon (Ashfaq et al., 2025; Khodaei et al., 2025; Shah and Ai, 2024). The variables were created and processed in QGIS (v. 3.40.8), where they were adjusted to the boundaries of the study area using the extraction tool. They were then resampled to 30 m using the bilinear resampling algorithm available in the software.

3.1.1 Topographical factors

Topographic conditions determine runoff direction, flow velocity, and water accumulation in low-lying or flat areas. In mountainous watersheds, elevation, slope, terrain curvature, topographic position, and surface roughness help identify valley bottoms and sectors where flow may concentrate during heavy rainfall events (Lee et al., 2018; Quesada-Román, 2021). Therefore, elevation, slope, profile curvature, plan curvature, TPI, and TRI were considered as topographic indicators of flood susceptibility.

Slope indicates the inclination of the terrain relative to the horizontal plane (Rimba et al., 2017), and controls surface water runoff, affecting both soil erosion and vertical percolation (Jahangir et al., 2019). It also influences the magnitude of the destructive forces of floods (Gacu et al., 2022), such that areas with lower slopes are more exposed (Liuzzo et al., 2019). Similarly, elevation is a key factor in flooding (Seejata et al., 2018), as it controls the direction and depth of water (Penki et al., 2022). ower-lying areas, especially flat surfaces, are at greater risk of flooding because water flows from higher to lower areas (Tang et al., 2018; Das and Pardeshi, 2018). Profile curvature measures the shape of the terrain in the direction of maximum slope (Krebs et al., 2015), where negative values indicate convex slopes, positive values indicate concave slopes, and values of 0 indicate uniform slopes (Ghahraman et al., 2023; Yeong Lee and Kim, 2020). Therefore, the greater the slope curvature, the easier it is for runoff water to accumulate, increasing vulnerability to flooding (Yeong Lee and Kim, 2020). Likewise, plane or surface curvature describes the divergence or convergence of water flow in an area (Jasiewicz and Stepinski, 2013) and is classified as concave (positive), convex (negative), and flat (value 0) (Alarifi et al., 2022). This information allows for the identification of flood-prone areas (Ghahraman et al., 2023), with flat areas being most susceptible to water accumulation and flash flooding (Cao et al., 2016). The Topographic Position Index (TPI) is a simple and repeatable technique for classifying terrain according to slope position and landscape shape (Alam et al., 2021), indicating the difference between the elevation of a focal cell and the average elevation of neighboring cells (Lei et al., 2021), and is calculated using Equations 1. 2, (Zhran et al., 2024).where: z0 is the center point, z is the average elevation surrounding the center point, and R is the predetermined radius. Positive TPI values indicate that the cell is higher than its surroundings, while negative values show that it is lower, like valleys (Reu et al., 2013). Terrain roughness index (TRI) quantifies the elevation variation between adjacent cells in a DEM (Ghahraman et al., 2023). Flat areas have values close to zero, while the highest positive values correspond to mountainous areas with steep ridges (Lei et al., 2021). This index is an important morphological indicator for the occurrence and behavior of floods (Bui et al., 2011; Luo and Liu, 2018) and is calculated using the equation:where: max and min are the highest and lowest values in the cells (Abdelkareem and Mansour, 2023). Topographic variables were obtained in QGIS (v.3.40.8) from a Digital Elevation Model (DEM) from the Shuttle Radar Topography Mission (SRTM) (https://earthexplorer.usgs.gov/) with 30 m of spatial resolutio.

3.1.2 Geomorphological factors

Geomorphological factors correspond to terrain characteristics and surface processes that influence flood magnitude, frequency, and spatial extent by controlling runoff pathways, flow concentration, channel overflow, lateral erosion, and flood propagation (Adesina et al., 2022). In tropical mountain watersheds, valley confinement, alluvial surfaces, depressions, piedmont fans, and floodplains help explain the spatial distribution of flood-prone areas (Quesada-Román, 2021; Quesada-Román and Peralta-Reyes, 2023).

Depressions are low-lying areas surrounded by high reliefs that store water during runoff, especially during rainfall in arid or (Alarifi et al., 2022; Zhu and Abdelkareem, 2021) emaining there until it infiltrates the underlying aquifers. Although they are important for water resources, they present a high risk of flash flooding in residential areas (Alarifi et al., 2022). Stream power index (SPI) characterizes the erosive capacity of water flow at a point on the topographic surface (Jebur et al., 2014), influencing the occurrence of floods (Khosravi et al., 2018) and controlling channel widening (Righini et al., 2017). It is calculated as:where: As is the upstream contributing area (m²) and β is the slope in a given cell (degrees) (Lei et al., 2021). Geomorphological factors were obtained from a Digital Elevation Model (DEM) from the Shuttle Radar Topography Mission (SRTM) (https://earthexplorer.usgs.gov/) and processed in QGIS (v.3.40.8) with a spatial resolution of 30 m.

3.1.3 Hydrological factors

Hydrological factors represent the connectivity between slopes, drainage networks, watercourses, and surface water accumulation areas. These conditions determine how runoff is generated, concentrated, and transferred within the watershed during rainfall events (Ahmed et al., 2023; Zhang et al., 2020). In this study, four hydrological indicators were considered: Topographic Wetness Index (TWI), distance to rivers, drainage density, and Modified Normalized Difference Water Index (MNDWI), which represent topographic wetness, proximity to watercourses, drainage organization, and the presence of surface water bodies, respectively (Rana et al., 2024).

Four variables were considered: Topographic Wetness Index (TWI) assesses topographic control in hydrological processes, such as runoff and flow accumulation (Ahmed et al., 2023). High TWI values indicate a higher probability of flooding, while low values indicate lower vulnerability (AlAli et al., 2023; Khosravi et al., 2019). The TWI was processed in QGIS (v.3.40.8) using the following equation:where: As is the upstream contribution area and β is the slope gradient (Ullah and Zhang, 2020). Distance to rivers influences soil and rock moisture on slopes (Miraki et al., 2019). Areas near rivers are at greater risk of flooding due to overflowing (Rana et al., 2024), while this risk decreases as the distance from the main drainage network increases (Ramkar and Yadav, 2021). This variable was calculated using a Digital Elevation Model (DEM) from the Shuttle Radar Topography Mission (SRTM) EarthExplorer portal of the USGS (https://earthexplorer.usgs.gov/) at a resolution of 30 m to derive the drainage networks, and then a Euclidean distance analysis was applied, processed in QGIS (v.3.40.8). Drainage density refers to the ratio between the total length of the main channel and the study area (Zhang et al., 2020and is directly linked to flooding. Higher density indicates greater surface runoff and a higher probability of flooding (Paul et al., 2019). It is calculated using the following equation:where: L represents the total length of the drainage channel in the area (km) and A describes the total area of the basin (km²) (Zhran et al., 2024). The variable was obtained from drainage networks using a Digital Elevation Model (DEM) from the Shuttle Radar Topography Mission (SRTM) with a resolution of 30 m, available on the USGS EarthExplorer portal (https://earthexplorer.usgs.gov/). The Modified Normalized Difference Water Index (MNDWI) allows open water bodies to be highlighted and the influence of built-up areas to be eliminated (Zhran et al., 2024). Water bodies have positive values (Ganjirad and Delavar, 2023) and are calculated using:where: MIR is a mid-infrared band (Xu, 2006). It was calculated from Landsat 8 satellite images, using the green band (Band 3) and mid-infrared band (Band 6), downloaded from the USGS Earth Explorer portal (https://earthexplorer.usgs.gov/) and processed in QGIS (v.3.40.8).

3.1.4 Climate factor

Precipitation directly influences river flow and the occurrence of floods (Das, 2018). High-intensity, long-duration rainfall increases runoff and therefore the likelihood of flooding (Ramkar and Yadav, 2021). The variable was obtained from WorldClim raster data (v.2.1) (www.worldclim.org/data/bioclim.html, accessed 23 September 2025) with an original resolution of 30 s (∼1 km²), and then resampled to 30 m using bilinear interpolation in QGIS (v.3.36.3).

3.1.5 Anthropogenic factors

Anthropogenic conditions can intensify flood risk by modifying infiltration capacity, surface runoff, drainage connectivity, and exposure patterns. Land-use change, urban expansion, agricultural pressure, and population concentration can alter floodplain dynamics and increase the vulnerability of settlements located near river channels or low-lying areas (Saha and Agrawal, 2020; Hoque et al., 2019). Therefore, land use and population density were considered as indicators of territorial modification and social exposure to flooding, especially in data-scarce contexts where spatial information is essential for risk assessment (Quesada-Román et al., 2025b).

Land use affects infiltration rate (Saha and Agrawal, 2020) and the frequency of floods (Das, 2018), showing a high correlation with areas prone to flooding, especially in urbanized areas where moisture insulation increases (Apollonio et al., 2016). The land use map was generated from a pre-existing map of the Amazonas department, available on the Geoportal of the Peruvian Ministry of Transport and Communications. Population density is a key factor in assessing social vulnerability, as areas with high populations are more susceptible to the impacts of flooding on physical and psychological health (Zhran et al., 2024; Hoque et al., 2019). This variable was generated from the database of the last national census and was subsequently processed in QGIS (v.3.36.3) using the Kernel Density tool to obtain the spatial distribution of the population.

3.2 Correlation analysis

Although the predictive variables were selected based on an exhaustive literature review and by a panel of experts, their spatial autocorrelation raised some concerns, as autocorrelation can lead to significant overfitting (Meyer et al., 2019) and inflate the model’s precision. To minimize this, in this study, Pearson’s correlation was calculated using the cor(method = “pearson”) function in RStudio (v. 4.5.1). The results did not show marked multicollinearity among the predictor variables, as all Pearson correlation coefficients were below the threshold of |r| = 0.8. Although several correlations were statistically significant, their magnitudes were not high enough to indicate problematic collinearity. In Table 1, the lower triangle shows the Pearson correlation coefficients (r), while the upper triangle shows the corresponding two-tailed significance levels obtained from the correlation analysis; furthermore, the diagonal represents autocorrelation, and significance levels are indicated as p < 0.001, p < 0.01, p < 0.05, and p ≥ 0.05.

3.3 Risk class assignment

The fifteen influencing variables selected for the flood risk model, which fall under the five factors mentioned above, are presented in Figure 3. Each variable was reclassified into five susceptibility categories ranging from very low to very high (Table 2). For this, the natural breaks method (Jenks) was used, which minimizes variance within each class while maximizing differences between classes; the “Reclassify” tool in QGIS (v. 3.36.3) was used for this purpose (Rivera et al., 2025). The spatial patterns reveal that the highest susceptibility values for the topographic and hydrological variables are concentrated along the floodplain of the Utcubamba River and its main tributaries, while the anthropogenic variables show maximum values in the urban center of Bagua Grande and the surrounding periurban areas.

FIGURE 3

3.4 Analytic hierarchy process (AHP)

The AHP, developed by Thomas Saaty (Saaty, 1980), is a decision-making tool that allows the establishment of a scale of preferences among a set of options, considering multiple perspectives and objectives (Ayalew et al., 2004). This method is based on pairwise comparisons between criteria, using a numerical scale from 1 to 9 to quantify the relative importance of each. The scale includes intermediate values ​​(2, 4, 6, 8) to reflect nuances in the preferences of decision-makers (Wind and Saaty, 1980).

In this study, the GIS-based AHP approach was chosen because it is well-suited for flood risk assessment in tropical mountain watersheds with limited data, where hydrometeorological records, flood inventories, and high-resolution hydraulic data are often scarce (Almouctar et al., 2024). Under these conditions, purely hydrological or machine learning approaches may be limited by a lack of calibration, validation, or training data (Yu et al., 2023); therefore, methods that integrate available spatial variables, expert judgment, and geomorphological information provide a practical framework for flood risk assessment at the municipal and watershed scales (Quesada-Román et al., 2025b). This approach is also consistent with integrated geomorphological and multi-criteria workflows for identifying flood-prone areas in tropical settings (Aichi et al., 2024; Quesada-Román and Montalván-Burbano, 2026).

To comparatively evaluate the model variables, a Pairwise Comparison Matrix (PCM) was constructed (Varma et al., 2025), of 15 × 15, where each pair of variables was analyzed according to its relevance in the flood risk assessment. From these comparisons, a normalized matrix was generated that illustrates the relative importance of each variable with respect to the others (Table 3).

To validate the method and the results, the consistency ratio (CR) was used, which allows evaluating the reliability of the AHP process and the robustness of the results obtained (Luu et al., 2018) If the CR values ​​are less than 10%, the matrix is ​​consistent (Wind and Saaty, 1980). The consistency of the relative importance weights assigned during the pairwise comparison can be verified using the following equation:where: CI is the consistency index, while RI is the random consistency index.

The IQ was calculated as follows:Where: λmáx is a largest eigenvalue y = order of the matrix.

The values for the random consistency index (RI) are proposed and standardized by Saaty and are determined based on the number of variables used in the AHP process (David Raj et al., 2022; Pradhan et al., 2021). In this study, we used 15 variables; therefore, the RI value is 1.583.

3.5 Weighted linear combination (WLC)

The flood risk map for the province of Utcubamba was generated using QGIS software (v3.36.3), employing the weighted linear combination (WLC) technique, a method commonly used in flood risk mapping research that applies the AHP approach (Hajibigloo et al., 2017; Karymbalis et al., 2021; Solaimani et al., 2023). Weighted linear combination (WLC), along with the AHP method, allows for the integration of multiple conditioning variables by assigning weights based on expert judgment, facilitating its application in GIS environments (Karymbalis et al., 2021; Solaimani et al., 2023). In this case, the QGIS Raster Calculator tool was used to generate the flood risk map. The weights of each variable were multiplied by their respective raster layer (Noori and Bonakdari, 2023) (Figure 4), applying the following formula.:where: H is the degree of risk, n is the number of factors, W i is the weight of factor i, and Xi is the rating of factor i (Karymbalis et al., 2021).

FIGURE 4

Subsequently, the flood risk map was segmented into uniform intervals using natural breaks, applying the Jenks method (Ahmed et al., 2023; Rana et al., 2024; Vojtek et al., 2021), generating five risk levels: very high, moderate high, low, and very low. The Jenks natural rupture classification method, proposed by Jenks and Caspall (1971) It is a data grouping method used to determine the optimal distribution of values across different classes, minimizing variance within classes and maximizing variance between classes. This can improve model robustness and moderate the nonlinear relationship between flood risk indicators (Zhang et al., 2020). Other classification methods, such as equal intervals, quantiles, or standard deviation, can distort the data or fail to reflect its true magnitude. In contrast, Jenks’ natural cutoff method adjusts the boundaries to the distribution of values, better representing the dispersion and clustering of flood risk (Moreira et al., 2021).

3.6 Validation the model

Given that the province of Utcubamba has limited georeferenced data on flooding, an indirect validation approach was adopted in this case (Dandapat and Panda, 2017), which is suitable for regions with limited data; Consequently, four statistical metrics were calculated to evaluate the model’s performance: the relative mean error (RME) quantifies the average relative deviation between observed and predicted risk classes, expressed as a percentage (Chakraborty and Mukhopadhyay, 2019). It is calculated as (insert formula), and the root mean square error (RMSE) measures the square root of the mean of the squared differences between observed and predicted values, penalizing larger errors more severely than smaller ones (Chai and Draxler, 2014). The mean absolute error (MAE) provides a direct measure of the mean prediction error in the same units as the risk classification scale (1–5) (Wang et al., 2023). Accuracy within ±1 risk class was calculated as the proportion of validation points for which the absolute difference between the observed and predicted classes did not exceed one category. Additionally, validation was performed using the AUC-ROC, a statistical and mathematical technique widely used to assess the accuracy and reliability of natural hazard modeling and mapping, particularly in flood susceptibility mapping, due to its reasonable, logical, and easily understandable method (Tehrany et al., 2015; Mitra et al., 2022; Shah and Ai, 2024). The ROC curve was constructed by plotting the true positive rate (sensitivity) on the Y-axis against the false positive rate (1 minus specificity) on the X-axis (Ashfaq et al., 2025) (Figure 4).

4 Results

4.1 AHP model performance evaluation

The AHP method was applied to map flood risk zones in the province of Utcubamba, Amazonas, Peru, yielding consistent and reliable results. The consistency ratio (CR) of the generated matrices was 0.076, below the 10% threshold recommended by Saaty, thus validating the coherence of the results. Among the variables analyzed, precipitation (Pp) had the highest prioritization factor at 14.29%, followed by slope (p) at 13.53% and elevation (E) at 13.16% (Table 4). These results reflect that rainfall intensity, along with terrain topography, are determining factors in the occurrence and magnitude of floods in the region and anywhere in the world.

4.2 Landslide risk mapping of the Utcubamba River basin, Peru

The flood risk map for the province of Utcubamba (Figure 5) reveals a clear spatial gradient from the valley floor toward the surrounding slopes. Approximately 401.39 km² (10.09%) of the provincial territory is classified as very high risk, concentrated predominantly along the Utcubamba River floodplain between Bagua Grande and El Milagro, where the convergence of high rainfall, low elevation, flat topography, and dense population creates conditions of extreme vulnerability (Figure 5B). Another 824.25 km² (22.47%) is classified as high risk, mainly in transition zones at the margins of the floodplain and along secondary tributaries. Together, the high- and very high-risk categories account for 32.56% of the provincial territory, corresponding to areas requiring priority attention for flood risk management and land-use planning. Moderate-risk zones (30.1%) constitute the largest category and correspond to areas of moderate slope with intermediate drainage connectivity, while the low- and very low-risk categories (37.3% combined) are associated with elevated, well-drained terrain in the eastern and western mountain ranges.

FIGURE 5

The results of the spatial classification by risk categories (Figure 6) indicate that the moderate-risk class covers the largest area (1,198.05 km², 30.1%), followed by low risk (990.81 km², 24.9%), high risk (824.25 km², 22.5%), very high risk (401.39 km², 10.1%), and very low risk (491.26 km², 12.4%). The concentration of high and very high risk in the valley floor (32.56% combined) reflects the disproportionate flood hazard faced by the most densely populated areas of the province, underscoring the need for priority risk management interventions in the Utcubamba River floodplain corridor.

FIGURE 6

4.3 Population exposure to flood risk

The spatial analysis of population exposure to flood risk reveals marked territorial heterogeneity across the seven districts of Utcubamba (Figure 7). l Milagro and Bagua Grande present the highest proportional exposure, with 23.79% and 19.29% of their respective territories classified as high or very high risk, driven by their location on the active floodplain of the Utcubamba River and their elevated population density (Figure 8). In absolute terms, Bagua Grande concentrates the largest number of inhabitants within high risk zones (Figure 7B), consistent with its role as the provincial capital and main urban center. At the other extreme, Yamón, Jamalca, and Lonya Grande show substantially lower territorial exposure (below 8% each), consistent with their higher mean elevation and more dissected topography, which limits floodplain extent. Notably, the populated centers of Puerto Naranjitos, Goncha, Misquiyacu Bajo, and Puerto Nuevo have 100% of their area within high or very high risk zones, representing the most critical localities for emergency planning and risk governance at the municipal level.

FIGURE 7

FIGURE 8

Figure 9 presents the flood risk exposure profile for each district in the province of Utcubamba using a spider (radar) chart (Figure 9A) and a mean composite risk score ranked by district (Figure 9B). The spider chart displays relative risk scores across five conditioning factor groups—topographic, geomorphological, hydrological, climatic, and anthropogenic—with each axis scaled from 1 (very low) to 5 (very high) based on AHP-weighted variables. Bagua Grande and El Milagro exhibit the largest polygon areas, reflecting consistently high scores across all five factor groups, with mean composite scores of 4.44 and 4.56 respectively, placing both districts in the very high risk category. Their elevated scores in the climatic and anthropogenic dimensions are particularly notable, reflecting the combined effect of high precipitation and intense land-use pressure on flood risk. In contrast, Yamón, Jamalca, and Lonya Grande present contracted polygons with mean scores below 2.3, indicative of low to moderate overall exposure. The intermediate districts—Cajaruro and Cumba—display asymmetric profiles with relatively higher hydrological and topographic scores compared to their anthropogenic scores, suggesting that natural terrain factors dominate their flood susceptibility rather than land-use-driven vulnerability.

FIGURE 9

4.4 Model validation

Indirect validation yielded an RME of 12.82% and an RMSE of 0.7845 (Table 5); an RME between 10% and 20% indicates good model performance, and the mean absolute error (MAE = 0.5128) and accuracy within ±1 risk class (94.9%) further confirm that the AHP-GIS model provides a reliable spatial representation of flood risk levels throughout the study area.

The ROC curve analysis yielded an AUC of 0.816 (95% CI: 0.721–0.895), indicating good discriminatory performance of the AHP-GIS flood risk model (Figure 10). According to standard classification criteria (Swets, 1988), AUC values between 0.80 and 0.90 reflect good model accuracy. The optimal classification threshold, determined by Youden’s J-index (threshold = 0.750), achieved a sensitivity of 94.9% and a specificity of 64.1%, demonstrating that the model is particularly effective at correctly identifying flood-prone areas while maintaining a reasonable false-positive rate. The combination of high sensitivity (94.9%) and acceptable specificity (64.1%) reflects an asymmetric but justified performance profile for flood risk mapping.

FIGURE 10

5 Discussions

Floods are considered one of the most frequent and devastating natural disasters worldwide, characterized by their high destructive potential and their ability to significantly alter the physical and social environment (Tanoue et al., 2016). In recent years, the constant occurrence of these events has generated serious economic impacts, affecting infrastructure, productive activities and above all causing irreversible damage to human resources (Ishiwatari and Sasaki, 2024). Therefore, knowing and mapping the magnitude and risks allows us to analyze the influence of environmental, social, and physical variables and at the same time understand how these affect the occurrence of this type of disaster (Riedel et al., 2024). To develop this type of research, multiple approaches have been used, such as spatial modeling techniques based on machine learning algorithms or spatial statistical analysis methods (Sharma et al., 2024).

To identify areas susceptible to flooding in the basin, the multi-criteria decision-making technique was used through the Analytic Hierarchy Process (AHP) method integrated with Geographic Information Systems (GIS) (Kader et al., 2024). This methodology made it possible to consider topographic, hydrological, climatic and anthropogenic factors in an interrelated way when mapping risk zones (Ocak et al., 2024). Through a pairwise comparison matrix, the relative weighting of the factors and their variables was obtained, evaluating their influence on the occurrence of floods (Ashfaq et al., 2025). In the present study, the model results show that climatic and topographic factors, especially precipitation and slope, are the most influential, which agrees with recent research demonstrating that the intensity and distribution of precipitation, along with the morphometric conditions of the terrain, control flooding processes in basins of different regions (Dey et al., 2023; Majeed et al., 2023). Furthermore, current studies indicate that drainage density and land use modulate the hydrological behavior of a territory and can increase risk exposure in areas that have been impacted (Ullah et al., 2024).

From a geomorphological perspective, the significant slope and drainage density suggest that runoff is generated rapidly on the slopes and flows through the tributary channels toward the main valley, in the lower-lying areas, alluvial surfaces and floodplains favor water accumulation, channel overflow, sediment deposition, and lateral flood spread (Arana-Ruedas and Del Piero, 2025). This interpretation is consistent with geomorphological approaches that emphasize the role of terrain morphology, drainage patterns, slope dynamics, and the configuration of alluvial plains in flood susceptibility within tropical mountain basins (Quesada Roman, 2021; Quesada-Román and Peralta-Reyes, 2023). Therefore, the high- and very-high-risk zones identified in Utcubamba should be understood as the result of upstream runoff concentration, valley confinement, drainage connectivity, and the occupation of areas naturally prone to flooding.

The high influence of precipitation in the model should also be interpreted in the context of climate variability and climate change. In tropical Andean basins, flood frequency and intensity are influenced by changes in precipitation patterns, variability associated with the El Niño–Southern Oscillation (ENSO), and hydroclimatic extremes capable of modifying basin response. Recent studies based on CMIP6 climate projections indicate that flood-prone areas may expand or intensify under future warming scenarios, particularly where extreme rainfall interacts with steep slopes, occupied floodplains, and land-use change (Hidalgo et al., 2024; Petiangma et al., 2025). Likewise, hydroclimatic extremes can generate cascading socioeconomic impacts on infrastructure, agriculture, transport, and livelihoods, reinforcing the need to interpret flood risk as both a physical and territorial problem (Quesada-Román et al., 2024). Therefore, the high and very high-risk zones identified in Utcubamba should be considered priority areas for adaptive planning under future hydroclimatic intensification scenarios.

The model developed in this research presented a consistency index (CR = 0.076), which strengthens the relationship between the comparison judgments and the reliability of the applied weighting process. This level of consistency falls within the optimal parameters proposed for multi-criteria decision models, thus reinforcing the methodological validity of the study (Morales and de Vries, 2021).

The validation results obtained in this study deserve particular emphasis given the data-scarce context in which the model was developed. An AUC of 0.816 classified as “good” discriminatory performance according to Swets (1988) places this AHP-GIS model among the better-performing multi-criteria flood risk assessments reported in the recent literature for tropical and Andean environments, where AUC values typically range between 0.70 and 0.85 (Ashfaq et al., 2025; Abdo et al., 2024). The high sensitivity of 94.9% is particularly relevant from an operational standpoint: it indicates that the model correctly identifies nearly all flood-prone locations in the validation inventory, minimizing the risk of under-mapping critical areas a property that is essential when flood risk maps are used to support land-use planning, infrastructure siting, and early warning system design (Calvo-Solano and Quesada-Román, 2024).

The moderate specificity (64.1%) reflects the well-documented asymmetry between flood occurrence records and non-flood locations in data-scarce inventories, where underreporting of flood events tends to inflate false-positive rates at the optimal threshold (Sahrane et al., 2023). Taken together, the RME of 12.82%, RMSE of 0.7845, MAE of 0.5128, and accuracy within ±1 risk class of 94.4% confirm that the AHP-WLC-GIS framework produces spatially reliable flood risk classifications that can be operationalized for disaster risk governance in tropical mountain provinces with limited hydrometeorological infrastructure. These findings support the replicability of the methodology in comparable data-scarce contexts across Peru and the broader Andean region, where conventional hydrological modeling approaches are constrained by the absence of stream gauge networks, high-resolution DEMs, and long-term flood inventories.

The flood risk mapping carried out for the province of Utcubamba identified that approximately 401.39 km² (10.09%) of the territory is under a very high risk level, while 824.25 km² (22.47%) correspond to high risk; therefore, it is inferred that about a third of the provincial territory (32.56%) presents a significant susceptibility to flooding. These results justify a high exposure of the territory to extreme hydrometeorological events, a scenario that coincides with the hydrological behavior of tropical basins with intense rainfall and irregular relief (Monteiro et al., 2022). Furthermore, the combination of orographic processes and changes in land use intensifies territorial vulnerability, and where the accumulation of flows is aggravated by alteration of land use and unplanned urban expansion (Kaya, 2025). Where recent studies support and have demonstrated that the reduction of vegetation cover and the waterproofing of the land modify the infiltration capacity and the water balance, amplifying the magnitude and frequency of flooding events in tropical regions (Islam, 2024; Rahman et al., 2023).

The findings of this study confirm that the population’s exposure to the risk of flooding in the province of Utcubamba is not homogeneous, but shows a marked territorial heterogeneity, concentrating in certain districts and populated centers (Dhungana, 2024). In particular, the districts of El Milagro and Bagua Grande have a large portion of their territory exposed (23.79% and 19.29%, respectively), while several populated centers such as Puerto Naranjitos, Goncha, Misquiyacu Bajo, and Puerto Nuevo are entirely under very high-risk conditions. This spatial distribution of risk coincides with studies conducted in Peru (Badillo-Rivera et al., 2025; Thouret et al., 2013), where plains areas near rivers and with disordered urban expansion processes present a high vulnerability, especially when the population settles in flood-prone areas located in the immediate vicinity of watercourses (Oteng-Ababio et al., 2024). Given that the province of Utcubamba requires particular urgency in the face of extreme hydrological events such as the recently documented overflows of the Utcubamba River, which have affected residential and agricultural areas, impacting homes, infrastructure, and the livelihoods of the inhabitants (Instituto Nacional de Defensa Civil, 2023).

The concentration of high and very high flood risk in districts such as Bagua Grande and El Milagro, as well as in settlements near the Utcubamba River, highlights the need to strengthen urban growth control, riverbank management, municipal preparedness, and land-use regulations in flood-prone areas. This is particularly relevant in Latin American municipalities, where social vulnerability, informal urban expansion, and limited institutional capacities can exacerbate the effects of disasters (Quesada-Román et al., 2025a; Quesada-Román et al., 2026b). In light of this, flood risk maps are not only diagnostic tools but also decision-support tools for land-use planning, infrastructure siting, and disaster risk reduction (Zhang et al., 2020; Quesada-Román and Peralta-Reyes, 2023).

In these types of environments, extreme events are not only related to the total amount of rainfall, but also to its intensity and duration, since these factors can also determine the rate of concentration of river flows (Fowler et al., 2021). Thus, even when short-duration rains can trigger unexpected increases in the flow of tributaries and cause them to overflow, or when the soils are already saturated, this explains the high proportion of areas classified as having high and very high risk (Li et al., 2025).

The results are consistent with studies conducted in tropical basins in Asia, Africa, and Latin America, which demonstrate a high sensitivity of river systems to hydrometeorological disturbances (Bursteinas et al., 2023; Rahayu et al., 2024). In such regions, the hydrological response is influenced by the slope and soil type, which determine the infiltration capacity and the concentration time (Vegad et al., 2024). In clay or compacted soils, infiltration is drastically reduced, favoring surface runoff and thus the recurrence of flooding in low-lying or confluence areas (Kuok et al., 2023). This pattern is consistent with Horton’s theory, which establishes a direct relationship between infiltration capacity and runoff generation, particularly in tropical climates with intense rainfall and short return periods (Hidayat et al., 2023).

The flood risk map for the province of Utcubamba (Figure 5) indica que approximately 401.39 km² (10.09%) of the provincial territory is classified as very high risk, predominantly concentrated in Bagua Grande and along the Utcubamba River floodplain toward El Milagro. This spatial pattern responds to the convergence of physical and anthropogenic factors in the valley bottom: low elevation, gentle slopes (<10°), high drainage density (2.57–4.08 km/km²), proximity to the main channel (<223 m), and elevated TWI values collectively favor water accumulation, slow runoff drainage, and rapid channel overflow during extreme rainfall events. High flood susceptibility zones are consistently characterized by low elevations, gentle slopes, close proximity to rivers, and high drainage density a pattern documented across AHP-GIS flood studies in comparable mountain basin contexts (Ashfaq et al., 2025; Teku et al., 2026).

The inverse relationship between slope and flood hazard is well established, with flatter areas where water stagnation is more likely consistently reporting elevated flood risk (Teku et al., 2026). These physical conditions are compounded by sustained unplanned urban expansion in Bagua Grande, which has reduced infiltration capacity, increased impervious surface coverage, and encroached upon natural floodplain storage areas, amplifying flood risk beyond what terrain factors alone would produce. Rapid urban expansion reduces natural landscape features and increases impervious surfaces, increasing stormwater runoff and river discharge rates and ultimately exacerbating flood risk (Esraz-Ul-Zannat et al., 2024; Madadi and Sadeghi, 2026). An additional 824.25 km² (22.47%) is classified as high risk in transitional zones along the floodplain margins and secondary tributaries, where drainage connectivity and river proximity remain high despite moderate increases in elevation. Together, high and very high risk categories account for 32.56% of the provincial territory, underscoring the need for priority flood risk management and land-use planning in the Utcubamba River.

On the other hand, territorial vulnerability is marked by land-use change processes, which critically alter the natural water balance. Deforestation of hillsides, the removal of vegetation cover, and unplanned urban expansion reduce the capacity of ecosystems to retain and regulate water flow (Lyu et al., 2025). The loss of vegetation cover implies less rain interception, a reduction in temporary storage in the canopy, and an increase in direct runoff into waterways (Sugianto et al., 2022). This process, combined with the sealing of the land by urban infrastructure, contributes to a greater accumulation of water flows and therefore to an intensification of risk (Kaiser and Akter, 2025).

Similarly, human pressure on floodplains represents a determining factor in the exposure of the population (Farshid and Null, 2025). These areas, formerly occupied by agricultural activities due to their fertility and accessibility, coincide with areas of greater susceptibility to flooding (Devitt et al., 2023). This spatial pattern is also reflected in studies of agricultural producing countries, where settlements on riverbanks have increased the recurrence of damage from extreme hydrometeorological events (Ogundolie et al., 2024). In this sense, land occupation without considering hydrological dynamics is one of the main risk factors, as it transforms natural processes of water dissipation and storage into mechanisms of accumulation and overflow (Devitt et al., 2023; Sun et al., 2022).

The high and very high flood risk zones identified in this study concentrated along the Utcubamba River floodplain and in the districts of Bagua Grande and El Milagro provide a spatial basis for prioritizing nature-based solutions (NbS) and ecosystem-based adaptation (EbA) as complementary approaches to structural flood mitigation (Quesada-Román and Montalván-Burbano, 2026; Chausson et al., 2020). Riparian vegetation restoration along the Utcubamba River and its tributaries would attenuate flood peaks, stabilize channel margins, and reduce surface runoff velocity (Christopher et al., 2024; Wohl et al., 2019), while watershed conservation in higher-elevation districts such as Yamón, Jamalca, and Lonya Grande would preserve infiltration capacity and regulate baseflow (Ellison et al., 2017; Griscom et al., 2017). Floodplain protection policies restricting urban expansion into risk zones identified in Figure 5 would further prevent the loss of natural flood storage areas (Temmerman et al., 2013; Devitt et al., 2023). Integrated watershed management frameworks that coordinate land-use regulation, reforestation, riverbank management, and community-based early warning systems represent the most effective pathway for flood risk reduction under current and future hydroclimatic conditions in tropical Andean basins (Quesada-Román and Montalván-Burbano, 2026; Acosta-Quesada and Quesada-Román, 2024; Petiangma et al., 2025), and are increasingly recognized within the Sendai Framework and the Kunming-Montreal Global Biodiversity Framework as essential complements to conventional engineering responses in data-scarce tropical regions (Cohen-Shacham et al., 2016; Seddon et al., 2020).

It is also important to acknowledge that in steep tropical mountain basins such as Utcubamba, flood risk does not operate in isolation but is embedded within a multi-hazard and cascading process context. Floods frequently interact with landslides, debris flows, sediment accumulation, and riverbank instability where a storm event triggers flooding, which in turn triggers landslides that may further increase flooding through river blockage and elevated sediment loads within the fluvial system (Gill and Malamud, 2014). These processes often occur in cascading sequences and have been particularly destructive in tropical and humid subtropical mountains, where sediment availability and antecedent rainfall conditions can initiate hazard cascades even under moderate rainfall intensities (Arango-Carmona et al., 2025). In the Andean context, intense rainfall events can simultaneously trigger slope failures in upper catchment areas while generating flash floods and sediment-laden flows in the valley floor, as reflected in the geological and geomorphological setting of Utcubamba (Acosta-Quesada and Quesada-Román, 2024; Quesada-Román, 2026).

On the other hand, the use of the AHP-GIS multi-criteria model allowed for the coherent integration of the main physical, climatic, and anthropogenic factors that influence flood risk (Sahrane et al., 2023). This approach, widely validated in recent studies, offers a more accurate spatial representation of susceptible areas and constitutes an effective tool for land-use planning (Asiri et al., 2024; Okafor and Oriakhi, 2024). Its application in tropical basins has proven to be important due to its ability to weigh variables of different natures such as slope, land use, drainage, precipitation and altitude, thus providing a complete characterization of risk (Sahraei et al., 2023).

This study demonstrates that mapping based on multi-criteria methodologies integrated with GIS not only allows the identification of critical areas, but also serves as a strategic tool for understanding vulnerability, anticipating risk scenarios, and laying the groundwork for preventive actions. It also provides systematic scientific evidence for future decision-making, taking an integrated approach to the main variables that trigger these types of natural phenomena.

Among the main limitations we have identified in our study, we note that the 30-m spatial resolution (SRTM DEM) introduces uncertainties in the topographic and hydrological indices. With this resolution, the TWI and SPI may excessively smooth small-scale drainage features in steep inter-Andean valleys, underestimating convergence zones in narrow ravines and misrepresenting catchment areas by 15%–30% compared to higher-resolution DEMs (≤12.5 m), as reported in comparable studies of tropical mountains (Habtezion et al., 2016; Saksena and Merwade, 2015). Similarly, slope and curvature derivatives are sensitive to DEM resolution, where coarser grids tend to reduce local slope variability and may attenuate TRI and TPI values in highly dissected terrain such as that of Utcubamba.

It is well known that the inventory data used for statistical analysis is subject to inventory bias due to incomplete data collection; this directly affects model performance by overestimating specificity in the AUC-ROC analysis (Sahrane et al., 2023). However, it is also recognized that differentiated assessments (i.e., quantitative and expert-based) are necessary to understand the limitations and reliability of landslide susceptibility analyses (Steger et al., 2017), as applied in this study, which explains the observed asymmetry between the model’s high sensitivity (94.9%) and its moderate specificity (64.1%) at the optimal Youden threshold.

Nevertheless, we believe that future studies should address these limitations by incorporating higher-resolution digital elevation models (DEMs) (12.5 m ALOS PALSAR or DEMs derived from unmanned aerial vehicles), expanding georeferenced flood inventories through community-based participatory mapping, and applying Monte Carlo or resampling sensitivity analyses to propagate the uncertainty of AHP weights into the confidence intervals of the risk maps.

6 Conclusion

This study assessed flood risk in the province of Utcubamba, in the Amazonas region (Peru), by integrating geospatial databases, Geographic Information Systems (GIS), the Hierarchical Analytic Process (HAP), and weighted linear combination, since this province, like others in Peru, has limited hydrometeorological data. Thanks to the methodological approach used, it was determined that variables such as precipitation, slope, elevation, drainage density, and terrain roughness have a significant influence on the flood risk model. In this regard, flood risk in Utcubamba should be understood as the result of the interaction between hydroclimatic forces, the morphology of the watershed, and human occupation of susceptible areas, thereby confirming the importance of considering not only climatic triggers but also geomorphological and territorial factors, such as terrain configuration, runoff concentration, valley floor areas, drainage patterns, land use, and population exposure.

Based on our study, we have mapped the districts and population centers at risk of flooding, identifying that 32.56% of the province of Utcubamba requires priority attention from local authorities, as our report classifies 10.09% of the province (401.39 km²) as being at very high risk, and 22.47% (824.25 km²) as high risk. Therefore, it is necessary to implement strategies for disaster risk reduction and adaptive spatial planning, directly addressing land use planning, control of urban sprawl, riverbank management, infrastructure improvements, and municipal disaster preparedness.

Conducting studies in tropical and Andean watersheds where detailed flood inventories, hydrometric stations, and high-resolution hydrological data are limited can pose a challenge; however, by using the AHP-GIS method proposed in this study, these limitations are overcome, as it integrates spatial data with expert criteria on flooding, which ensures that the semi-quantitative methodology developed in this research has produced highly reliable statistical data and accurate reports. It can be replicated in municipalities and regional governments that require operational tools for flood risk assessment but lack high-resolution temporal and spatial data, thereby closing a significant gap in disaster mitigation for such risks in remote areas of our country.

Key limitations include the spatial resolution of geospatial data and the lack of detailed historical records on flooding. Therefore, it is suggested that future studies incorporate complementary methods, such as machine learning algorithms, multi-temporal analysis, and advanced multi-criteria techniques, to improve the accuracy and scope of flood risk assessments in the region.

References

• 1

  AbdelkareemM.MansourA. M. (2023). Risk assessment and management of vulnerable areas to flash flood hazards in arid regions using remote sensing and GIS-Based knowledge-driven techniques. Nat. Hazards117 (3), 2269–2295. 10.1007/S11069-023-05942-X/FIGURES/15

  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  AbdoH. G.DarwishK. S.BindajamA. A.NiknamA.YoussefY. M.AhmedM. F. M.et al (2024). Detection of flood vulnerable areas in urban basins using multi-criteria analysis and geospatial tools: a case study from eastern mediterranean. Environ. Earth Sci.83 (17), 1–20. 10.1007/S12665-024-11801-0/METRICS

  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  AbdoH. G.ZengT.AlshayebM. J.PrasadP.AhmedM. F. M.AlbanaiJ. A.et al (2025). Multi-criteria analysis and geospatial applications-based mapping flood vulnerable areas: a case study from the eastern mediterranean. Nat. Hazards121 (1), 1003–1031. 10.1007/S11069-024-06864-Y/METRICS

  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  Acosta-QuesadaM.Quesada-RománA. (2024). Landslides and flood hazard mapping using geomorphological methods in Santa Ana, Costa Rica. Int. J. Disaster Risk Reduct.113, 104882. 10.1016/J.IJDRR.2024.104882

  • CrossRef
  • Google Scholar

• 5

  AdesinaE.AdewuyiA.AbdumalikO.MorenikejiG.NjokuD. (2022). Geomorphic assessment of flood hazard within the urban area of chanchaga local government area, minna, Nigeria. Int. J. Environ. Geoinformatics9 (1), 102–115. 10.30897/IJEGEO.877629

  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  AdgerW. N.BrooksN.BenthamG.AgnewM. (2004). New indicators of vulnerability and adaptive capacity. Final Project Report. Tyndal Center Climate Change, Research University of East Anglia, Norwich., January.

  • Google Scholar
  • View reference in article

• 7

  AhmedI.DasN.DebnathJ.BhowmikM.ChandraI.CollegeV.et al (2023). Flood Hazard Zonation Using Gis-based multi-parametric Analytical Hierarchy Process. 10.21203/RS.3.RS-3047215/V1

  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  AichiA.IkirriM.Ait HaddouM.Quesada-RománA.SahooS.SinghaC.et al (2024). Integrated GIS and analytic hierarchy process for flood risk assessment in the Dades Wadi watershed (Central High Atlas, Morocco). Results Earth Sci.2, 100019. 10.1016/J.RINES.2024.100019

  • CrossRef
  • Google Scholar
  • View reference in article

• 9

  AlAliA. M.SalihA.HassaballaA. (2023). Geospatial-based Analytical Hierarchy Process (AHP) and Weighted Product Model (WPM) techniques for mapping and assessing flood susceptibility in the Wadi Hanifah Drainage Basin, Riyadh Region, Saudi Arabia. Water15(10), 1943. 10.3390/W15101943

  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  AlamA.AhmedB.SammondsP. (2021). Flash flood susceptibility assessment using the parameters of drainage basin morphometry in SE Bangladesh. Quat. Int.575–576, 295–307. 10.1016/J.QUAINT.2020.04.047

  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  AlarifiS. S.AbdelkareemM.AbdallaF.AlotaibiM. (2022). Flash flood hazard mapping using remote sensing and GIS techniques in Southwestern Saudi Arabia. Sustainability14(21), 14145. 10.3390/SU142114145

  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  Ali Al-RawasG.Piero Arana-RuedasD. R.Pino-VargasE.del Águila-RíosS.HuaynaG. (2025). Mapping flood-prone areas using GIS and morphometric analysis in the mantaro watershed, Peru: approach to susceptibility assessment and management. Sustainability17(17), 7809. 10.3390/SU17177809

  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  AlmouctarM. A. S.WuY.AnS.YinX.QinC.ZhaoF.et al (2024). Flood risk assessment in arid and semi-arid regions using multi-criteria approaches and remote sensing in a data-scarce region. J. Hydrology Regional Stud.54, 101862. 10.1016/J.EJRH.2024.101862

  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  AmiriA.SoltaniK.EbtehajI.BonakdariH. (2024). A novel machine learning tool for current and future flood susceptibility mapping by integrating remote sensing and geographic information systems. J. Hydrology632, 130936. 10.1016/J.JHYDROL.2024.130936

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  ApollonioC.BalaccoG.NovelliA.TarantinoE.PiccinniA. F. (2016). Land use change impact on flooding areas: the case Study of Cervaro Basin (Italy). Sustainability8(10), 996. 10.3390/SU8100996

  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  Arana-RuedasD. P. R.Del PieroR. (2025). Mapping flood-prone areas using GIS and morphometric analysis in the Mantaro Watershed: approach to susceptibility assessment and management. Sustainability17 (17), 7809. 10.3390/su17177809

  • CrossRef
  • Google Scholar
  • View reference in article

• 17

  Arango-CarmonaM. I.VoitP.HürlimannM.AristizábalE.KorupO. (2025). Hillslope torrential hazard cascades in tropical mountains. Nat. Hazards Earth Syst. Sci.25 (9), 3641–3663. 10.5194/nhess-25-3641-2025

  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  AshfaqS.TufailM.NiazA.MuhammadS.AlzahraniH.TariqA. (2025). Flood susceptibility assessment and mapping using GIS-based analytical hierarchy process and frequency ratio models. Glob. Planet. Change251, 104831. 10.1016/J.GLOPLACHA.2025.104831

  • CrossRef
  • Google Scholar
  • View reference in article

• 19

  AsiriM. M.AldehimG.AlruwaisN.AllafiR.AlzahraniI.NouriA. M.et al (2024). Coastal Flood risk assessment using ensemble multi-criteria decision-making with machine learning approaches. Environ. Res.245 (December 2023), 118042. 10.1016/j.envres.2023.118042

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  AyalewL.YamagishiH.UgawaN. (2004). Landslide susceptibility mapping using GIS-based weighted linear combination, the case in Tsugawa area of Agano River, Niigata prefecture, Japan. N. Ugawal. 2004•Springer 1 (1), 73–81. 10.1007/S10346-003-0006-9

  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  Badillo-RiveraE.SantiagoR.PomaI.ChávezT.Arroyo-PazA.Aucahuasi-AlmidonA.et al (2025). Flood susceptibility mapping in El Niño Phenomenom integrating multitemporal radar analysis, GIS and machine learning techniques, Piura River basin, Peru. Front. Environ. Sci.13, 1672107. 10.3389/FENVS.2025.1672107

  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  BarbozaE.CorrotoF.SalasR.GamarraÓ.BallarínD.OlleroA. (2017). Hidrogeomorfología en áreas tropicales: aplicación del índice hidrogeomorfológico (IHG) en el río Utcubamba (Perú). Ecol. Apl.16 (1), 39–47. 10.21704/rea.v16i1.902

  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  BuiD. T.LofmanO.RevhaugI.DickO. (2011). Landslide susceptibility analysis in the Hoa Binh province of Vietnam using statistical index and logistic regression. Natural Hazards, 59 (3), 1413–1444. 10.1007/S11069-011-9844-2

  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  BurayuD. G.KaruppannanS.ShuniyeG. (2023). Identifying flood vulnerable and risk areas using the integration of analytical hierarchy process (AHP), GIS, and remote sensing: a case study of southern Oromia region. Urban Clim.51, 101640. 10.1016/J.UCLIM.2023.101640

  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  BursteinasI.LopesK.OliveiraC. G.DeFagundesM. R. (2023). Megacity using different high resolution DTMs. Water.

  • Google Scholar
  • View reference in article

• 26

  CabreraJ. S.LeeH. S. (2019). Flood-Prone area assessment using GIS-based multi-criteria analysis: a case study in Davao Oriental, Philippines. Water11(11), 2203. 10.3390/W11112203

  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  Calvo-SolanoO.Quesada-RománA. (2024). Worldwide research trends and networks on flood early warning systems. GeoHazards5 (3), 582–595. 10.3390/GEOHAZARDS5030030/S1

  • CrossRef
  • Google Scholar

• 28

  CaoC.XuP.WangY.ChenJ.ZhengL.NiuC. (2016). Flash flood hazard susceptibility mapping using frequency ratio and statistical index methods in coalmine subsidence areas. Sustain. (Switzerland)8 (9), 948. 10.3390/SU8090948

  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  CaramanicaA.MesiaL. H.MoralesC. R.HuckleberryG.Luis Jaime CastilloB.QuilterJ. (2020). El Niño resilience farming on the north coast of Peru. Proc. Natl. Acad. Sci. U. S. A.117 (39), 24127–24137. 10.1073/PNAS.2006519117/SUPPL_FILE/PNAS.2006519117.SAPP.PDF

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  Castro MedinaW. F. (2010). Geología: informe temático. Proyecto Zonificación Ecológica y Económica del departamento de Amazonas. Inst. de Investigaciones de la Amazonía Peruana; Gobierno Regional de Amazonas. Available online at: https://app.ingemmet.gob.pe/biblioteca/pdf/LIB-207.pdf (Accessed October 12, 2025).

  • Google Scholar
  • View reference in article

• 31

  ChaiT.DraxlerR. R. (2014). Root mean square error (RMSE) or mean absolute error (MAE)? -Arguments against avoiding RMSE in the literature. Geosci. Model Dev.7 (3), 1247–1250. 10.5194/GMD-7-1247-2014

  • CrossRef
  • Google Scholar
  • View reference in article

• 32

  ChakrabortyS.MukhopadhyayS. (2019). Assessing flood risk using analytical hierarchy process (AHP) and geographical information system (GIS): application in Coochbehar district of West Bengal, India. Nat. Hazards99 (1), 247–274. 10.1007/S11069-019-03737-7

  • CrossRef
  • Google Scholar
  • View reference in article

• 33

  ChaudharyM. T.PirachaA.AgarwalR.RutaG. (2021). Natural Disasters—Origins, impacts, Manag. Encycl.1(4), 1101–1131. 10.3390/ENCYCLOPEDIA1040084

  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  ChaussonA.TurnerB.SeddonD.ChabaneixN.GirardinC. A. J.KaposV.et al (2020). Mapping the effectiveness of nature-based solutions for climate change adaptation. Glob. Change Biol.26 (11), 6134–6155. 10.1111/gcb.15310

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  ChenW.LiY.XueW.ShahabiH.LiS.HongH.et al (2020). Modeling flood susceptibility using data-driven approaches of naïve Bayes tree, alternating decision tree, and random forest methods. Sci. Total Environ.701, 134979. 10.1016/J.SCITOTENV.2019.134979

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  ChristopherA. S.Leal FilhoW.WalshC. L.KosovacA. (2024). Enhancing river floodplain management with nature-based solutions: overcoming barriers and harnessing enablers. WIREs Water11 (4), e1723. 10.1002/wat2.1723

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  Cobian AlvarezJ.ResosudarmoB. P. (2025). Under the Water: Flood Impacts and Economic Dynamics in Northern Peru. 10.2139/SSRN.5346804

  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  Cohen-ShachamE.WaltersG.MaginnisS.JanzenC. (2016). Nature-Based Solutions to Address Global Societal Challenges. IUCN. 10.2305/IUCN.CH.2016.13

  • CrossRef
  • Google Scholar
  • View reference in article

• 39

  CostacheR.HongH.PhamQ. B. (2020). Comparative assessment of the flash-flood potential within small mountain catchments using bivariate statistics and their novel hybrid integration with machine learning models. Sci. Total Environ.711, 134514. 10.1016/J.SCITOTENV.2019.134514

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 40

  DandapatK.PandaG. K. (2017). Flood vulnerability analysis and risk assessment using analytical hierarchy process. Model. Earth Syst. Environ.3 (4), 1627–1646. 10.1007/S40808-017-0388-7/METRICS

  • CrossRef
  • Google Scholar
  • View reference in article

• 41

  DanumahJ. H.OdaiS. N.SaleyB. M.SzarzynskiJ.ThielM.KwakuA.et al (2016). Flood risk assessment and mapping in Abidjan district using multi-criteria analysis (AHP) model and geoinformation techniques, (cote d’ivoire). Geoenvironmental Disasters3 (1), 10. 10.1186/S40677-016-0044-Y

  • CrossRef
  • Google Scholar
  • View reference in article

• 42

  DasS. (2018). Geographic information system and AHP-based flood hazard zonation of Vaitarna basin, Maharashtra, India. Arabian J. Geosciences11 (19), 1–13. 10.1007/S12517-018-3933-4/METRICS

  • CrossRef
  • Google Scholar
  • View reference in article

• 43

  DasS.PardeshiS. D. (2018). Morphometric analysis of Vaitarna and Ulhas river basins, Maharashtra, India: using geospatial techniques. Appl. Water Sci.8 (6), 1–11. 10.1007/S13201-018-0801-Z/FIGURES/7

  • CrossRef
  • Google Scholar
  • View reference in article

• 44

  DasS.GuptaA.GhoshS. (2017). Exploring groundwater potential zones using MIF technique in semi-arid region: a case study of Hingoli district, Maharashtra. Spatial Inf. Res.25 (6), 749–756. 10.1007/S41324-017-0144-0/METRICS

  • CrossRef
  • Google Scholar
  • View reference in article

• 45

  DasS.PardeshiS. D.KulkarniP. P.DokeA. (2018). Extraction of lineaments from different azimuth angles using geospatial techniques: a case study of Pravara basin, Maharashtra, India. Arabian J. Geosciences11 (8), 1–13. 10.1007/S12517-018-3522-6/METRICS

  • CrossRef
  • Google Scholar
  • View reference in article

• 46

  David RajA.KumarS.SooryamolK. R. (2022). Modelling climate change impact on soil loss and erosion vulnerability in a watershed of Shiwalik Himalayas. Catena214, 106279. 10.1016/J.CATENA.2022.106279

  • CrossRef
  • Google Scholar
  • View reference in article

• 47

  DevittL.NealJ.CoxonG.SavageJ.WagenerT. (2023). Flood hazard potential reveals global floodplain settlement patterns. Nat. Commun.14 (1), 1–11. 10.1038/s41467-023-38297-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 48

  DeyB.AbirK. A. M.AhmedR.SalamM. A.RedowanM.MiahM. D.et al (2023). Monitoring groundwater potential dynamics of north-eastern Bengal Basin in Bangladesh using AHP-Machine learning approaches. Ecol. Indic.154, 110886. 10.1016/J.ECOLIND.2023.110886

  • CrossRef
  • Google Scholar
  • View reference in article

• 49

  DhunganaB. (2024). Urban – Rural Exposure to Flood Hazard and Social Vulnerability in the Conterminous United States.

  • Google Scholar
  • View reference in article

• 50

  DrenkhanF.CareyM.HuggelC.SeidelJ.OréM. T. (2015). The changing water cycle: climatic and socioeconomic drivers of water-related changes in the Andes of Peru. Wiley Interdiscip. Rev. Water2 (6), 715–733. 10.1002/WAT2.1105

  • CrossRef
  • Google Scholar
  • View reference in article

• 51

  EllisonD.MorrisC. E.LocatelliB.SheilD.CohenJ.MurdiyarsoD.et al (2017). Trees, forests and water: cool insights for a hot world. Glob. Environ. Change43, 51–61. 10.1016/j.gloenvcha.2017.01.002

  • CrossRef
  • Google Scholar
  • View reference in article

• 52

  Esraz-Ul-ZannatM.Dedekorkut‐HowesA.MorganE. A. (2024). A review of nature-based infrastructures and their effectiveness for urban flood risk mitigation. WIREs Clim. Change15 (3), e0889. 10.1002/wcc.889

  • CrossRef
  • Google Scholar
  • View reference in article

• 53

  FarshidA.NullS. E. (2025). Flood injustice in 500-year floodplains. Front. Hum. Dyn.7 (April), 1–15. 10.3389/fhumd.2025.1496686

  • CrossRef
  • Google Scholar
  • View reference in article

• 54

  FowlerH. J.LenderinkG.PreinA. F.WestraS.AllanR. P.BanN.et al (2021). Anthropogenic intensification of short-duration rainfall extremes. Nat. Rev. Earth & Environ. (2, 2), 107–122. 10.1038/s43017-020-00128-6

  • CrossRef
  • Google Scholar
  • View reference in article

• 55

  GacuJ. G.MonjardinC. E. F.SenoroD. B.TanF. J. (2022). Flood risk assessment using GIS-Based analytical hierarchy process in the municipality of Odiongan, Romblon, Philippines. Appl. Sci.12(19), 9456. 10.3390/APP12199456

  • CrossRef
  • Google Scholar
  • View reference in article

• 56

  GainA. K.MojtahedV.BiscaroC.BalbiS.GiupponiC. (2015). An integrated approach of flood risk assessment in the eastern part of Dhaka City. Nat. Hazards79 (3), 1499–1530. 10.1007/S11069-015-1911-7/METRICS

  • CrossRef
  • Google Scholar
  • View reference in article

• 57

  GamarraO.CorrotoF.BarrenaM.RascónJ.ChávezJ. (2018). Calidad ecológica del agua en la cuenca del río Utcubamba, Amazonas, Perú. Available online at: www.innovateperu.gob.pe (Accessed October 8, 2025).

  • Google Scholar
  • View reference in article

• 58

  GhahramanK.NagyB.Nooshin NokhandanF. (2023). Flood-Prone zones of meandering Rivers: machine learning approach and considering the role of morphology (Kashkan River, Western Iran). Geosciences13 (9), 267. 10.3390/GEOSCIENCES13090267

  • CrossRef
  • Google Scholar
  • View reference in article

• 59

  GigovićL.PamučarD.BajićZ.DrobnjakS. (2017). Application of GIS-interval rough AHP methodology for flood hazard mapping in Urban areas. Water (Switzerland)9 (6), 360. 10.3390/W9060360

  • CrossRef
  • Google Scholar
  • View reference in article

• 60

  GillJ. C.MalamudB. D. (2014). Reviewing and visualizing the interactions of natural hazards. Rev. Geophys.52 (4), 680–722. 10.1002/2013RG000445

  • CrossRef
  • Google Scholar
  • View reference in article

• 61

  GriscomB. W.AdamsJ.EllisP. W.HoughtonR. A.LomaxG.MitevaD. A.et al (2017). Natural climate solutions. Proc. Natl. Acad. Sci.114 (44), 11645–11650. 10.1073/pnas.1710465114

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 62

  GuoE.ZhangJ.RenX.ZhangQ.SunZ. (2014). Integrated risk assessment of flood disaster based on improved set pair analysis and the variable fuzzy set theory in central Liaoning Province, China. Nat. Hazards74 (2), 947–965. 10.1007/S11069-014-1238-9/METRICS

  • CrossRef
  • Google Scholar
  • View reference in article

• 63

  HabtezionN.NasabM. T.ChuX. (2016). How does DEM resolution affect microtopographic characteristics, hydrologic connectivity, and modelling of hydrologic processes?Hydrol. Process.30 (24), 4870–4892. 10.1002/hyp.10967

  • CrossRef
  • Google Scholar
  • View reference in article

• 64

  HadipourV.VafaieF.DeilamiK. (2020). Coastal flooding risk assessment using a GIS-Based spatial multi-criteria decision analysis approach. Water12(9), 2379. 10.3390/W12092379

  • CrossRef
  • Google Scholar
  • View reference in article

• 65

  HajibiglooM.GhezelsoflooA. A.MemarianH.Berdi SheikhV. (2017). Mapping the flood-prone areas for developing a flood risk management system in the northeast of Iran. Water Harvest. Res.2 (2), 1–12. 10.22077/JWHR.2018.825

  • CrossRef
  • Google Scholar
  • View reference in article

• 66

  HidalgoH. G.AlfaroE. J.Quesada-RománA. (2024). Flood projections for selected Costa Rican main basins using CMIP6 climate models downscaled output in the HBV hydrological model for scenario SSP5-8.5. Hydrological Res. Lett.18 (1), 35–42. 10.3178/HRL.18.35

  • CrossRef
  • Google Scholar
  • View reference in article

• 67

  HidayatD. P. A.DarsonoS. L. W.FaridM. (2023). Evaluation of infiltration modeling in the cisadane watershed in Indonesia: existing and new approach equation. Water (Switzerland)15 (23), 4149. 10.3390/w15234149

  • CrossRef
  • Google Scholar
  • View reference in article

• 68

  HoqueM. A. A.TasfiaS.AhmedN.PradhanB. (2019). Assessing spatial flood vulnerability at Kalapara Upazila in Bangladesh using an analytic hierarchy process. Sensors19 (6), 1302. 10.3390/S19061302

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 69

  Instituto Nacional de Defensa Civil [INDECI] (2021). Informe de emergencia N° 1834: Inundación en la provincia de Utcubamba. Amazon. Recuperado de. Available online at: https://portal.indeci.gob.pe/wp-content/uploads/2021/12/INFORME-DE-EMERGENCIA-N%C2%BA-1834-22DIC2021-INUNDACION-EN-LA-PROVINCIA-DE-UTCUBAMBA-AMAZONAS-2.pdf (Accessed October 3, 2025).

  • Google Scholar
  • View reference in article

• 70

  Instituto Nacional de Defensa Civil [INDECI] (2023). Informe de Emergencia de las Lluvias intensas en el departamento de Amazonas. Available online at: https://portal.indeci.gob.pe/wp-content/uploads/2025/02/INFORME-DE-EMERGENCIA-N.o-0160-12FEB2025-LLUVIAS-INTENSAS-EN-EL-DEPARTAMENTO-DE-SAN-MARTÍN-16.pdf (Accessed November 21, 2025).

  • Google Scholar
  • View reference in article

• 71

  Instituto Nacional de Estadística e Informática (INEI) (2014). Departamento Amazonas: población total proyectada y ubicación geográfica de la capital legal, según provincia y distrito, 2014 [Archivo Excel]. Publicaciones digitales. Available online at: https://www.inei.gob.pe/media/MenuRecursivo/publicaciones_digitales/Est/Lib1159/cuadros/amazonas/amazonas_1_3.xls (Accessed November 27, 2025).

  • Google Scholar
  • View reference in article

• 72

  Instituto Nacional de Estadística e Informática (INEI) (2018). “Amazonas: resultados definitivos,” in Tomo I: Aspectos generales, análisis de los principales resultados y cuadros estadísticos de población, vivienda y hogar [PDF]. Lima: INEI. Available online at: https://www.inei.gob.pe/media/MenuRecursivo/publicaciones_digitales/Est/Lib1567/01TOMO_01.pdf (Accessed December 2, 2025).

  • Google Scholar
  • View reference in article

• 73

  IshiwatariM.SasakiD. (2024). Investing in resilience: estimating financial needs and benefits of flood protection in developing Asian countries. Proc. Int. Assoc. Hydrological Sci.386, 87–93. 10.5194/piahs-386-87-2024

  • CrossRef
  • Google Scholar
  • View reference in article

• 74

  IslamK. (2024). GIS based flood susceptibility mapping in the Keleghai river basin , India : a comparative assessment of bivariate statistical models. Discov. Water4, 129. 10.1007/s43832-024-00186-7

  • CrossRef
  • Google Scholar
  • View reference in article

• 75

  JahangirM. H.Mousavi ReinehS. M.AbolghasemiM. (2019). Spatial predication of flood zonation mapping in Kan River Basin, Iran, using artificial neural network algorithm. Weather Clim. Extrem.25, 100215. 10.1016/J.WACE.2019.100215

  • CrossRef
  • Google Scholar
  • View reference in article

• 76

  JasiewiczJ.StepinskiT. F. (2013). Geomorphons-a pattern recognition approach to classification and mapping of landforms. Geomorphology182, 147–156. 10.1016/J.GEOMORPH.2012.11.005

  • CrossRef
  • Google Scholar
  • View reference in article

• 77

  JeburM. N.PradhanB.TehranyM. S. (2014). Optimization of landslide conditioning factors using very high-resolution airborne laser scanning (LiDAR) data at catchment scale. Remote Sens. Environ.152, 150–165. 10.1016/J.RSE.2014.05.013

  • CrossRef
  • Google Scholar
  • View reference in article

• 78

  JenksG. F.CaspallF. C. (1971). Error on choroplethic maps: definition, measurement, reduction. Ann. Assoc. Am. Geogr.61 (2), 217–244. 10.1111/j.1467-8306.1971.tb00779.x

  • CrossRef
  • Google Scholar
  • View reference in article

• 79

  KaderZ.IslamM. R.AzizM. T.HossainM. M.IslamM. R.MiahM.et al (2024). GIS and AHP-based flood susceptibility mapping: a case study of Bangladesh. Sustain. Water Resour. Manag.10 (5), 1–18. 10.1007/S40899-024-01150-Y

  • CrossRef
  • Google Scholar
  • View reference in article

• 80

  KaiserZ. R. M. A.AkterF. (2025). From risk to resilience and sustainability: addressing urban flash floods and waterlogging. Risk Sci.1, 100011. 10.1016/J.RISK.2025.100011

  • CrossRef
  • Google Scholar
  • View reference in article

• 81

  KarymbalisE.AndreouM.BatzakisD. V.TsanakasK.KaralisS. (2021). Integration of GIS-Based multicriteria decision analysis and analytic hierarchy process for flood-hazard assessment in the Megalo Rema River catchment (East Attica, Greece). Sustainability13 (18), 10232. 10.3390/SU131810232

  • CrossRef
  • Google Scholar

• 82

  KayaB. (2025). Extreme precipitation and flood hazard assessment for sustainable climate adaptation: a case study of. 1–25.

  • Google Scholar

• 83

  KazakisN.KougiasI.PatsialisT. (2015). Assessment of flood hazard areas at a regional scale using an index-based approach and Analytical Hierarchy process: application in Rhodope-Evros region, Greece. Sci. Total Environ.538, 555–563. 10.1016/J.SCITOTENV.2015.08.055

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 84

  KhodaeiH.Nasiri SalehF.Nobakht DalirA.ZareiE. (2025). Future flood susceptibility mapping under climate and land use change. Sci. Rep.15 (1), 1–18. 10.1038/S41598-025-97008-0

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 85

  KhosraviK.PhamB. T.ChapiK.ShirzadiA.ShahabiH.RevhaugI.et al (2018). A comparative assessment of decision trees algorithms for flash flood susceptibility modeling at Haraz watershed, northern Iran. Sci. Total Environ., 627, 744–755. 10.1016/J.SCITOTENV.2018.01.266

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 86

  KhosraviK.MelesseA. M.ShahabiH.ShirzadiA.ChapiK.HongH. (2019). Flood susceptibility mapping at Ningdu catchment, China using bivariate and data mining techniques. Extreme Hydrology Clim. Var. Monit. Model. Adapt. Mitig., 419–434. 10.1016/B978-0-12-815998-9.00033-6

  • CrossRef
  • Google Scholar

• 87

  KrebsP.StockerM.PezzattiG. B.ConederaM. (2015). An alternative approach to transverse and profile terrain curvature. Int. J. Geogr. Inf. Sci.29 (4), 643–666. 10.1080/13658816.2014.995102;SUBPAGE:STRING:ACCESS

  • CrossRef
  • Google Scholar

• 88

  KuokK. K.ChiuP. C.RahmanM. R.SaidK. A.binM.ChinM. Y. (2023). Evaluation of total infiltration and storage capacities for different soil types in Sarawak using SWMM. Discov. Water3 (1), 18. 10.1007/s43832-023-00042-0

  • CrossRef
  • Google Scholar

• 89

  Lavado-CasimiroW.QquentaJ.MontesinosC. AsenciosH.FelipeO. (2023). Modelamiento espacio-temporal de inundaciones en la cuenca de los Andes Huallaga en Perú. EGU General Assembly 2023, Vienna, Austria, EGU23-2938. 10.5194/egusphere-egu23-2938

  • CrossRef
  • Google Scholar

• 90

  LeeK.JungD.LeeS.이대업 ㆍ정성호 ㆍ이기하이경상 (2018). Analysis of large-scale flood inundation area using optimal topographic factors. J. Korea Water Resour. Assoc.51(6), 481–490. 10.3741/JKWRA.2018.51.6.481

  • CrossRef
  • Google Scholar

• 91

  LeiX.ChenW.PanahiM.FalahF.RahmatiO.UuemaaE.et al (2021). Urban flood modeling using deep-learning approaches in Seoul, South Korea. J. Hydrology601, 126684. 10.1016/J.JHYDROL.2021.126684

  • CrossRef
  • Google Scholar

• 92

  LiH.HuK.LiuS.ChengH.WenZ.ZhangX.et al (2025). Abundant antecedent rainfall incubated a group-occurring debris flow event in the Dadu River Basin, Southwest China. Landslides22 (6), 1955–1971. 10.1007/S10346-025-02489-9

  • CrossRef
  • Google Scholar

• 93

  LiuG.ZhangY.ZhangJ.LangQ.ChenY.WanZ.et al (2023). Geographic-Information-System-Based risk assessment of flooding in Changchun Urban rail transit System. Remote Sens., Page 3533, 15(14), 3533. 10.3390/RS15143533

  • CrossRef
  • Google Scholar

• 94

  LiuzzoL.SammartanoV.FreniG. (2019). Comparison between different distributed methods for flood susceptibility mapping. Water Resour. Manag.33 (9), 3155–3173. 10.1007/S11269-019-02293-W/METRICS

  • CrossRef
  • Google Scholar

• 95

  LuoW.LiuC. C. (2018). Innovative landslide susceptibility mapping supported by geomorphon and geographical detector methods. Springer15 (3), 465–474. 10.1007/S10346-017-0893-9

  • CrossRef
  • Google Scholar

• 96

  LuuC.Von MedingJ.KanjanabootraS. (2018). Assessing flood hazard using flood marks and analytic hierarchy process approach: a case study for the 2013 flood event in Quang Nam, Vietnam. Nat. Hazards, 90(3), 1031–1050. 10.1007/S11069-017-3083-0

  • CrossRef
  • Google Scholar

• 97

  LyuH.YinZ.ShenS.ChenX.SuD. (2025). A novel framework incorporating machine learning into GIS for flood susceptibility prediction of urban metro systems. Sci. China Technol. Sci., 68(6), 1620701-. 10.1007/S11431-024-2925-8

  • CrossRef
  • Google Scholar

• 98

  MadadiP.SadeghiA. (2026). Integrating urban expansion and flood risk: a spatial assessment of impervious surface growth and floodplain exposure in Mecklenburg County (2011–2021). World Water Policy12 (1), e70052. 10.1002/wwp2.70052

  • CrossRef
  • Google Scholar

• 99

  MajeedM.LuL.AnwarM. M.TariqA.QinS.El-HefnawyM. E.et al (2023). Prediction of flash flood susceptibility using integrating analytic hierarchy process (AHP) and frequency ratio (FR) algorithms. Front. Environ. Sci.10 (January), 1–14. 10.3389/fenvs.2022.1037547

  • CrossRef
  • Google Scholar

• 100

  MarthJ. D.MouraN. S. V.KoesterE. (2016). Estudo da suscetibilidade à inundação com base em análise geomorfológica, bacia hidrográfica do Arroio Santa Isabel, região costeira do Rio Grande do Sul, Brasil. Rev. Do Inst. Geol. (Descontinuada)37 (1), 13–28. 10.5935/0100-929X.20160002

  • CrossRef
  • Google Scholar

• 101

  Medina AllccaL.Vílchez MataM. S.Dueñas BravoS. (2009). Riesgos geológicos en la región Amazonas (Boletín C-39). Instituto Geológico. Available online at: https://hdl.handle.net/20.500.12544/244 (Accessed December 18, 2025).

  • Google Scholar

• 102

  MeyerH.ReudenbachC.WöllauerS.ModellingT. (2019). Importance of Spatial Predictor Variable Selection in Machine Learning applications–Moving from Data Reproduction to Spatial Prediction. Elsevier. Available online at: https://www.sciencedirect.com/science/article/pii/S0304380019303230 (Accessed November 18, 2025).

  • Google Scholar

• 103

  MirakiS.ZanganehS. H.ChapiK.SinghV. P.ShirzadiA.ShahabiH.et al (2019). Mapping groundwater potential using a novel hybrid intelligence approach. Water Resour. Manag., 33(1), 281–302. 10.1007/S11269-018-2102-6

  • CrossRef
  • Google Scholar

• 104

  MokhtariE.MezaliF.AbdelkebirB.EngelB. (2023). Flood risk assessment using analytical hierarchy process: a case study from the Cheliff-Ghrib watershed, Algeria. J. Water Clim. Change14 (3), 694–711. 10.2166/WCC.2023.316

  • CrossRef
  • Google Scholar

• 105

  MonteiroS.Oliveira-jF.DeGhaffarB.TariqA.QinS.MumtazF.et al (2022). Rainfall in the urban area and Its impact on climatology and population growth. Atmosphere. 13, 1610. 10.3390/atmos13101610

  • CrossRef
  • Google Scholar

• 106

  MoralesF. F.de VriesW. T. (2021). Establishment of natural hazards mapping criteria using Analytic Hierarchy Process (AHP). Front. Sustain.2 (May), 1–13. 10.3389/frsus.2021.667105

  • CrossRef
  • Google Scholar

• 107

  MoreiraL. L.de BritoM. M.KobiyamaM. (2021). Effects of different normalization, aggregation, and classification methods on the construction of flood vulnerability indexes. Water, 13(1), 98. 10.3390/W13010098

  • CrossRef
  • Google Scholar

• 108

  Municipalidad Provincial de Utcubamba (2013). “Plan de desarrollo urbano de la ciudad de Bagua Grande (2014–2025): Expansión urbana y condiciones del suelo en Bagua Grande,” in Documento de planificación territorial municipal. Scribd. Available online at: https://es.scribd.com/document/601400876/Pdu-Bagua-Grande-propuesta (Accessed November 21, 2025).

  • Google Scholar

• 109

  NooriA.BonakdariH. (2023). A GIS-Based fuzzy hierarchical modeling for flood susceptibility mapping: a case Study in Ontario, Eastern Canada. Environ. Sci. Proc., Page 62, 25(1), 62. 10.3390/ECWS-7-14242

  • CrossRef
  • Google Scholar

• 110

  OcakF.BahadirB. A. M.ÜniversitesiS.KavakM.YüksekokuluM.VeŞ.et al (2024). Evaluation of potential flood areas in the Basin of Lake Ladik through AHP and GIS integration, (Samsun, Türkiye). Jeomorfol. Araştırmalar Derg.2024 (13), 71–93. 10.46453/JADER.1513212

  • CrossRef
  • Google Scholar

• 111

  OgundolieO. I.OlabiyisiS. O.GaniyuR. A.JeremiahY. S.OgundolieF. A. (2024). Assessment of flood vulnerability in Osun River Basin using AHP method. BMC Environ. Sci.1 (1), 1–21. 10.1186/s44329-024-00009-z

  • CrossRef
  • Google Scholar

• 112

  OkaforG. U.OriakhiO. (2024). Flood susceptibility modelling using GIS-based Analytical Hierarchy Process (AHP) in Benin City, Nigeria. J. Sci. Technol. Res.6 (3), 229–242. 10.5281/zenodo.13937655

  • CrossRef
  • Google Scholar

• 113

  Oteng-AbabioM.AgergaardJ.Møller-JensenL.AndreasenM. H. (2024). Flood risk reduction and resilient city growth in sub-Saharan Africa: searching for coherence in Accra’s urban planning. Front. Sustain. Cities6, 1118896. 10.3389/frsc.2024.1118896

  • CrossRef
  • Google Scholar

• 114

  PaulG. C.SahaS.HembramT. K. (2019). Application of the GIS-Based probabilistic models for mapping the flood susceptibility in Bansloi sub-basin of ganga-bhagirathi River and their comparison. Remote Sens. Earth Syst. Sci.2 (2–3), 120–146. 10.1007/S41976-019-00018-6/METRICS

  • CrossRef
  • Google Scholar

• 115

  PenkiR.BasinaS. S.TanniruS. R. (2022). Application of Geographical Information System-based Analytical Hierarchy Process Modeling for Flood Susceptibility Mapping of Krishna District in Andhra Pradesh. 10.21203/RS.3.RS-1399020/V1

  • CrossRef
  • Google Scholar

• 116

  PetiangmaD. M.SinghG. G.Quesada-RománA.HidalgoH.BlakeS.GonzalezA.et al (2025). Climate change and flood susceptibility in Bocas del Toro, Panama: a multi-criteria spatial analysis approach. J. Environ. Manag.395, 127741. 10.1016/J.JENVMAN.2025.127741

  • CrossRef
  • Google Scholar

• 117

  PradhanR. M.GuruB.PradhanB.BiswalT. K. (2021). Análisis multicriterio integrado para el mapeo del potencial de aguas subterráneas en terrenos de roca dura precámbricos (norte de Gujarat), IndiaTaylor & Francis66 (6), 961–978. 10.1080/02626667.2021.1906427

  • CrossRef
  • Google Scholar

• 118

  Quesada RománA. (2021). Zonificación de deslizamientos e inundaciones usando análisis geomorfológicos en una cuenca dinámica de Costa Rica. Rev. Cartográfica (102), 125–138. 10.35424/rcarto.i102.901

  • CrossRef
  • Google Scholar

• 119

  Quesada-RománA.Montalván-BurbanoN. (2026). Global trends in scaling nature-based solutions for disaster risk reduction. Phys. Chem. Earth142, 104302. 10.1016/J.PCE.2026.104302

  • CrossRef
  • Google Scholar

• 120

  Quesada-RománA.Peralta-ReyesM. (2023). Geomorphological mapping global trends and applications. Geographies3(3), 610–621. 10.3390/GEOGRAPHIES3030032

  • CrossRef
  • Google Scholar

• 121

  Quesada-RománA.HidalgoH. G.AlfaroE. J. (2024). Assessing the impact of tropical cyclones on economic sectors in Costa Rica, Central America. Trop. Cyclone Res. Rev.13 (3), 196–207. 10.1016/J.TCRR.2024.08.001

  • CrossRef
  • Google Scholar

• 122

  Quesada-RománA.Picado-MongeA.Rivera-SolísJ.HernándezM.RuizM.Torres-BerhardL.et al (2025a). Método para el cálculo del riesgo de inundaciones a escala de cuencas y municipios en América Central. Rev. Geográfica de América Cent.2 (75), 285–322. 10.15359/RGAC.75-2.10

  • CrossRef
  • Google Scholar

• 123

  Quesada-RománA.Rivera-SolísJ.Picado-MongeA. (2025b). Occurrence, impacts, and future challenges of disaster risk in Panama. Georisk Assess. Manag. Risk Eng. Syst. Geohazards19 (1), 45–61. 10.1080/17499518.2024.2337389

  • CrossRef
  • Google Scholar

• 124

  Quesada-RománA.Becerra-SandovalG.Ruiz-ÁlvarezM.VelásquezG. (2026a). Socioeconomic Determinants and Spatial Trends of Disaster Risk in Nicaragua. 10.6084/M9.FIGSHARE.32043952.V1

  • CrossRef
  • Google Scholar

• 125

  Quesada-RománA.Calvo-SolanoO. D.Cruz-GarcíaD.CastellanosE. J.LeyD.VásquezO.-J. (2026b). Unveiling historical data and socioeconomic drivers of disaster risk at national and municipal levels in Guatemala. Discov. Hazards2(1), 7-. 10.1007/S44475-026-00015-6

  • CrossRef
  • Google Scholar

• 126

  RahayuH. P.ZulfaK. I.NurhasanahD.HaighR.AmaratungaD.WahdinyI. I. (2024). Unveiling transboundary challenges in river flood risk management: learning from the Ciliwung River basin. Nat. Hazards Earth Syst. Sci.24 (6), 2045–2064. 10.5194/NHESS-24-2045-2024

  • CrossRef
  • Google Scholar

• 127

  RahmanZ. U.UllahW.BaiS.UllahS.JanM. A.KhanM.et al (2023). GIS-based fl ood susceptibility mapping using bivariate statistical model in Swat River Basin, Eastern Hindukush region, Pakistan. Environ. Inf. Remote Sens. July11, 1–17. 10.3389/fenvs.2023.1178540

  • CrossRef
  • Google Scholar

• 128

  RajasekharM.Sudarsana RajuG.SreenivasuluY.Siddi RajuR. (2019). Delineation of groundwater potential zones in semi-arid region of Jilledubanderu river basin, Anantapur District, Andhra Pradesh, India using fuzzy logic, AHP and integrated fuzzy-AHP approaches. HydroResearch2, 97–108. 10.1016/J.HYDRES.2019.11.006

  • CrossRef
  • Google Scholar

• 129

  RamkarP.YadavS. M. (2021). Flood risk index in data-scarce river basins using the AHP and GIS approach. Nat. Hazards109 (1), 1119–1140. 10.1007/S11069-021-04871-X/METRICS

  • CrossRef
  • Google Scholar

• 130

  RanaS. M. S.HabibS. A.SharifeeM. N. H.SultanaN.RahmanS. H. (2024). Flood risk mapping of the flood-prone Rangpur division of Bangladesh using remote sensing and multi-criteria analysis. Nat. Hazards Res.4 (1), 20–31. 10.1016/j.nhres.2023.09.012

  • CrossRef
  • Google Scholar

• 131

  ReuJ.DeBourgeoisJ.BatsM.GeomorphologyA. Z. (2013). Application of the Topographic Position Index to Heterogeneous Landscapes, Elsevier. Available online at: https://www.sciencedirect.com/science/article/pii/S0169555X12005739 (Accessed November 9, 2025).

  • Google Scholar

• 132

  RiedelL.RöösliT.VogtT.BreschD. N. (2024). Fluvial flood inundation and socio-economic impact model based on open data. Geosci. Model Dev.17 (13), 5291–5308. 10.5194/gmd-17-5291-2024

  • CrossRef
  • Google Scholar

• 133

  RighiniM.SurianN.WohlE.MarchiL.ComitiF.AmponsahW.et al (2017). Geomorphic response to an extreme flood in two Mediterranean rivers (northeastern Sardinia, Italy): analysis of controlling factors. Elsevier290, 184–199. 10.1016/J.GEOMORPH.2017.04.014

  • CrossRef
  • Google Scholar

• 134

  RimbaA.SetiawatiM.SambahA.ScienceF. M.-U. (2017). Physical Flood Vulnerability Mapping Applying Geospatial Techniques in Okazaki City, Aichi Prefecture, Japan. mdpi.Com.Available online at: https://www.mdpi.com/2413-8851/1/1/7 (Accessed November 22, 2025).

  • Google Scholar

• 135

  RiveraC. A.Valqui-ReinaS. V.García-NaranjoL. F.Ocaña-ZúñigaC. L.Auquiñivin-SilvaE. A.Chapa-GonzaS. R.et al (2025). Geospatial landslide risk mapping using AHP and GIS: a case Study of the utcubamba River Basin, Peru. Appl. Sci., 15(17), 9423. 10.3390/APP15179423

  • CrossRef
  • Google Scholar

• 136

  SahaA. K.AgrawalS. (2020). Mapping and assessment of flood risk in Prayagraj district, India: a GIS and remote sensing study. Nanotechnol. Environ. Eng.5 (2), 1–18. 10.1007/S41204-020-00073-1/METRICS

  • CrossRef
  • Google Scholar

• 137

  SahraeiR.Kanani-SadatY.SafariA.HomayouniS. (2023). Flood susceptibility modelling using geospatial-based Multicriteria decision making in large Scale areas. ISPRS annals of the photogrammetry. Remote Sens. Spatial Inf. Sci.10 (4/W1-2022), 677–683. 10.5194/isprs-annals-X-4-W1-2022-677-2023

  • CrossRef
  • Google Scholar

• 138

  SahraneR.BounabA.EL KharimY. (2023). Investigating the effects of landslides inventory completeness on susceptibility mapping and frequency-area distributions: case of Taounate province, Northern Morocco. CATENA220, 106737. 10.1016/J.CATENA.2022.106737

  • CrossRef
  • Google Scholar

• 139

  SahraneR.BounabA.KharimY. E. (2023). Investigating the effects of landslides inventory completeness on susceptibility mapping and frequency–area distributions: case of Taounate province, Northern Morocco. Catena220, 106737. 10.1016/j.catena.2022.106737

  • CrossRef
  • Google Scholar

• 140

  SardonH.Lavado-CasimiroW.FelipeO. (2022). “Inventario de datos de eventos de inundaciones del Peru,” in Estudio Final, Servicio Nacional de Meteorologia e Hidrologia del Peru - SENAMHI.

  • Google Scholar

• 141

  SeddonN.ChaussonA.BerryP.GirardinC. A. J.SmithA.TurnerB. (2020). Understanding the value and limits of nature-based solutions to climate change and other global challenges. Philosophical Trans. R. Soc. B Biol. Sci.375 (1794), 20190120. 10.1098/rstb.2019.0120

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 142

  SeejataK.YodyingA.WongthadamT.MahavikN.TantaneeS. (2018). Assessment of flood hazard areas using Analytical Hierarchy Process over the Lower Yom Basin, Sukhothai Province. Procedia Eng.212, 340–347. 10.1016/J.PROENG.2018.01.044

  • CrossRef
  • Google Scholar

• 143

  Servicio Nacional de Meteorología e Hidrología del Perú (SENAMHI) (2016). Ciclos horarios de precipitación en el Perú utilizando información satelital. Available online at: https://www.senamhi.gob.pe/load/file/01401SENA-34.pdf (Accessed October 7, 2025).

  • Google Scholar

• 144

  ShahS. A.AiS. (2024). Flood susceptibility mapping contributes to disaster risk reduction: a case study in Sindh, Pakistan. Int. J. Disaster Risk Reduct.108, 104503. 10.1016/J.IJDRR.2024.104503

  • CrossRef
  • Google Scholar

• 145

  SharmaA.PooniaM.RaiA.BiniwaleR. B.TügelF.HolzbecherE.et al (2024). Flood susceptibility mapping using GIS-based frequency ratio and Shannon’s Entropy index bivariate statistical models: a case Study of Chandrapur District, India. ISPRS Int. J. Geo-Information13 (8), 297. 10.3390/ijgi13080297

  • CrossRef
  • Google Scholar

• 146

  SmithM.RickerM.PrützR.AnandM.LehnerB.FlörkeM.et al (2021). Global assessment of Current and future River flooding and the role of nature-based solutions for risk management. Nat. Conservancy, 1–16.

  • Google Scholar

• 147

  SolaimaniK.ShokrianF.DarvishiS. (2023). An assessment of the integrated multi-criteria and new models efficiency in watershed flood mapping. Water Resour. Manag.37 (1), 403–425. 10.1007/S11269-022-03380-1/METRICS

  • CrossRef
  • Google Scholar

• 148

  StegerS.BrenningA.BellR.GladeT. (2017). The influence of systematically incomplete shallow landslide inventories on statistical susceptibility models and suggestions for improvements. Landslides14, 1767–1781. 10.1007/s10346-017-0820-0

  • CrossRef
  • Google Scholar

• 149

  SugiantoS.DeliA.MiswarE.RusdiM.IrhamM. (2022). The effect of land use and land cover changes on flood occurrence in Teunom watershed. Aceh Jaya. Land11 (8), 1271. 10.3390/land11081271

  • CrossRef
  • Google Scholar

• 150

  SunQ.FangJ.DangX.XuK.FangY.LiX.et al (2022). Multi-scenario urban flood risk assessment by integrating future land use change models and hydrodynamic models. Nat. Hazards Earth Syst. Sci.22 (11), 3815–3829. 10.5194/nhess-22-3815-2022

  • CrossRef
  • Google Scholar

• 151

  SwainK. C.SinghaC.NayakL. (2020). Flood susceptibility mapping through the GIS-AHP technique using the cloud. ISPRS Int. J. Geo-Information, 9(12), 720. 10.3390/IJGI9120720

  • CrossRef
  • Google Scholar

• 152

  TangZ.ZhangH.YiS.XiaoY. (2018). Assessment of flood susceptible areas using spatially explicit, probabilistic multi-criteria decision analysis. J. Hydrology558, 144–158. 10.1016/J.JHYDROL.2018.01.033

  • CrossRef
  • Google Scholar

• 153

  TanoueM.HirabayashiY.IkeuchiH. (2016). Global-scale river flood vulnerability in the last 50 years. Sci. Rep.6, 1–9. 10.1038/srep36021

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 154

  TaoukidouN.KarpouzosD.GeorgiouP. (2025). Flood hazard assessment through AHP, fuzzy AHP, and frequency ratio methods: a comparative analysis. Water5, 17(14), 2155. 10.3390/W17142155

  • CrossRef
  • Google Scholar

• 155

  TehranyM. S.PradhanB.JeburM. N. (2014). Flood susceptibility mapping using a novel ensemble weights-of-evidence and support vector machine models in GIS. J. Hydrology512, 332–343. 10.1016/J.JHYDROL.2014.03.008

  • CrossRef
  • Google Scholar

• 156

  TehranyM. S.PradhanB.MansorS.AhmadN. (2015). Flood susceptibility assessment using GIS-based support vector machine model with different kernel types. Catena125, 91–101. 10.1016/J.CATENA.2014.10.017

  • CrossRef
  • Google Scholar

• 157

  TekuA.SisayT.AliA.AyalewA. (2026). “Optimizing flood hazard zonation and planning landscape-based mitigation measures in Gimba Sub Watersheds, Northeastern Ethiopia,” J. Flood Risk Manag.19, e70172. 10.1111/jfr3.70172

  • CrossRef
  • Google Scholar

• 158

  TemmermanS.MeireP.BoumaT. J.HermanP. M. J.YsebaertT.De VriendH. J. (2013). Ecosystem-based coastal defence in the face of global change. Nature504 (7478), 79–83. 10.1038/nature12859

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 159

  ThouretJ. C.EnjolrasG.MartelliK.SantoniO.LuqueJ. A.NagataM.et al (2013). Combining criteria for delineating lahar-and flash-flood-prone hazard and risk zones for the city of Arequipa, Peru. Nat. Hazards Earth Syst. Sci.13, 339–360. 10.5194/nhess-13-339-2013

  • CrossRef
  • Google Scholar

• 160

  TripathiA. K.AgrawalS.GuptaR. D. (2020). Cloud enabled SDI architecture: a review. Earth Sci. Inf.13 (2), 211–231. 10.1007/S12145-020-00446-9/METRICS

  • CrossRef
  • Google Scholar

• 161

  UllahK.ZhangJ. (2020). GIS-based flood hazard mapping using relative frequency ratio method: a case study of Panjkora River Basin, eastern Hindu Kush, Pakistan. PLOS ONE15 (3), e0229153. 10.1371/JOURNAL.PONE.0229153

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 162

  UllahN.TariqA.QasimS.PanezaiS.UddinM. G.Abdullah-Al-WadudM.et al (2024). Geospatial analysis and AHP for flood risk mapping in Quetta, Pakistan: a tool for disaster management and mitigation. Appl. Water Sci.14 (11), 1–18. 10.1007/s13201-024-02293-1

  • CrossRef
  • Google Scholar

• 163

  VahidniaM. H.AlesheikhA. A.AlimohammadiA. (2009). Hospital site selection using fuzzy AHP and its derivatives. J. Environ. Manag.90 (10), 3048–3056. 10.1016/J.JENVMAN.2009.04.010

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 164

  VarmaA. K.DhoteA.MathewA.NareshC.ShekarP. R. (2025). Flood hazard zonation using remote sensing, geographic information system, and analytic hierarchy process in the Bhagirathi River Basin, Uttarakhand, India. Results Earth Sci.3, 100105. 10.1016/J.RINES.2025.100105

  • CrossRef
  • Google Scholar

• 165

  VegadU.PokhrelY.MishraV. (2024). Flood risk assessment for Indian sub-continental river basins. Hydrology Earth Syst. Sci.28 (5), 1107–1126. 10.5194/HESS-28-1107-2024

  • CrossRef
  • Google Scholar

• 166

  VojtekM.VojtekováJ.PhamQ. B. (2021). Gis-based spatial and multi-criteria assessment of riverine flood potential: a case study of the nitra river basin, slovakia. ISPRS Int. J. Geo-Information10 (9), 578. 10.3390/IJGI10090578/S1

  • CrossRef
  • Google Scholar

• 167

  WangY.FangZ.HongH.PengL. (2020). Flood susceptibility mapping using convolutional neural network frameworks. J. Hydrology582, 124482. 10.1016/J.JHYDROL.2019.124482

  • CrossRef
  • Google Scholar

• 168

  WindY.SaatyT. L. (1980). Marketing applications of the analytic hierarchy process. Pubsonline.Informs.OrgY Wind, TL SaatyManagement Sci. 1980. pubsonline.Informs.Org26 (7), 641–658. 10.1287/MNSC.26.7.641

  • CrossRef
  • Google Scholar

• 169

  WohlE.KramerN.Ruiz-VillanuevaV.ScottD. N.ComitiF.GurnellA. M.et al (2019). The natural wood regime in rivers. BioScience69 (4), 259–273. 10.1093/biosci/biz013

  • CrossRef
  • Google Scholar

• 170

  XuH.TianZ.SunL.YeQ.RagnoE.BrickerJ.et al (2022). Compound flood impact of water level and rainfall during tropical cyclone periods in a coastal city: the case of Shanghai. Nat. Hazards Earth Syst. Sci.22 (7), 2347–2358. 10.5194/NHESS-22-2347-2022

  • CrossRef
  • Google Scholar

• 171

  YariyanP.AvandM.AbbaspourR. A.Torabi HaghighiA.CostacheR.GhorbanzadehO.et al (2020). Flood susceptibility mapping using an improved analytic network process with statistical models. Geomatics, Nat. Hazards Risk11 (1), 2282–2314. 10.1080/19475705.2020.1830181

  • CrossRef
  • Google Scholar

• 172

  Yeong LeeJ.KimJ.-S. (2020). ROC analysis of topographic factors in flood vulnerable area considering Surface runoff characteristics. Ecol. Resilient Infrastructure7 (4), 327–335. 10.17820/ERI.2020.7.4.327

  • CrossRef
  • Google Scholar

• 173

  YoussefA. M.PradhanB.SefryS. A. (2016). “Flash flood susceptibility assessment in Jeddah city (Kingdom of Saudi Arabia) using bivariate and multivariate statistical models,” Environ. Earth Sci.7511–16. 10.1007/S12665-015-4830-8

  • CrossRef
  • Google Scholar

• 174

  YuH.LuoZ.WangL.DingX.WangS. (2023). Improving the accuracy of flood susceptibility prediction by combining machine learning models and the expanded flood inventory data. Remote Sens.15 (14), 3601. 10.3390/RS15143601

  • CrossRef
  • Google Scholar

• 175

  ZhangD.ShiX.XuH.JingQ.PanX.LiuT.et al (2020). A GIS-based spatial multi-index model for flood risk assessment in the Yangtze River Basin, China. Environ. Impact Assess. Rev.83, 106397. 10.1016/J.EIAR.2020.106397

  • CrossRef
  • Google Scholar

• 176

  ZhranM.GhanemK.TariqA.AlshehriF.JinS.DasJ.et al (2024). Exploring a GIS-based analytic hierarchy process for spatial flood risk assessment in Egypt: a case study of the Damietta branch. Environ. Sci. Eur.36 (1), 1–25. 10.1186/S12302-024-01001-9/TABLES/2

  • CrossRef
  • Google Scholar

• 177

  ZhuQ.AbdelkareemM. (2021). Mapping groundwater potential zones using a knowledge-driven approach and GIS analysis. Water13 (5), 579. 10.3390/W13050579

  • CrossRef
  • Google Scholar

• 178

  ZubietaR.GetiranaA.EspinozaJ. C.Lavado-CasimiroW.AragonL. (2017). Hydrological modeling of the peruvian-ecuadorian Amazon Basin using GPM-IMERG satellite-based precipitation dataset. Hydrology Earth Syst. Sci.21 (7), 3543–3555. 10.5194/HESS-21-3543-2017

  • Pubmed Abstract
  • CrossRef
  • Google Scholar