A Dynamic AHP–GIS Framework for Spatio-Temporal Flood Risk Assessment Incorporating Flood Risk Transfer Index (FRTI)

Abstract

Understanding the relationship between the processes involved in hydrology and changesin land use becomes more urgent amid the accelerated development of urban areas. In this regard, this paper proposes the application of a spatio-temporal analysis of flood vulnerability through multi-criteria analysis (Analytical Hierarchy Process), integrated with GIS and modeling of multidimensional urban development processes within Cizre, Turkey. Important hydrological factors for the formation of flood risks, such as elevation, slope, land use/cover, rainfall, drainage density, and proximity to the river, were considered when preparing the flood susceptibility map. It was revealed that high- and very-high-risk zones are mainly located near the Tigris River and in urbanized areas, which occupy more than half of the territory under consideration. Multidimensional analysis showed that unplanned development increases flood risks in the area because of the increased area of impervious surfaces and the violation of natural water flows. As a way to overcome the limitations of traditional methods of static analysis of flood risks, the Flood Risk Transfer Index (FRTI) has been developed to describe the process of spatial redistribution of risks resulting from the impact of the increase in urbanization rates. The indicator of spatial redistribution of flood risk reached a value of 0.72, showing that flood pressures increased in existing cities instead of reducing them. Thus, this study provides a breakthrough in understanding flood risks through the introduction of a new methodology.

1. Introduction

1.1. Background

Flooding in urban areas has become one of the most important environmental issues in the fast-growing parts of the world [1]. Higher frequency and severity of flood events are caused by the confluence of climate variability and faster urbanization [2]. Although extreme precipitation, river discharge, seasonal variability, and other hydro-meteorological factors are essential, the change of the natural landscapes into built environments is critical to the change in hydrological processes [3].

Urbanization causes the substitution of permeable surfaces with impervious surfaces, like concrete and asphalt, which reduces the infiltration rates, the surface runoffs, and the concentration of the flows [4]. These transformations make areas prone to flooding, especially those that are low-lying and those that have river accessibility [5]. Urban flooding is not a natural hazard but a land use choice in cities where unplanned development into floodplains has increased the exposure and susceptibility, further increasing vulnerability [6].

Other than enhancing the frequency of floods, urbanization alters natural drainage systems and hydrological connectivity. Growth of upstream or high areas may cause a change in the pathways of the runoff, resulting in more water in the downstream [7]. This interplay between urban morphology and hydrological processes points to the necessity to learn to see flood risk as a process and a changing situation, and not as a spatial state [8].

Climate change is one of the main global factors that affect flood occurrence in terms of their likelihood, magnitude, and geographic dispersion [9]. As the planet’s temperature has been increasing over time, there have been changes in the amount of moisture contained in the atmosphere and in rainfall patterns, which result in an increase in instances of very strong rainfall and short-term intensive rain [10]. This, in turn, contributes to increased surface water runoff and flooding in urban areas. It should be noted that climate change also affects the amount of water in rivers and the process of snow melting [11].

1.2. Research Gap

Despite substantial advances in flood risk mapping, various important challenges continue to emerge in recent research [12]. Many current methodologies, including GIS-based MCDA, efficiently determine flood-prone areas [13,14,15]. Still, most studies remain static, portraying flood risk under stable environmental and land use conditions and not accounting for changing risk due to urbanization [16].

Innovations in remote sensing and machine learning have recently increased prediction accuracy [6,17,18]. Yet, most studies concentrate mainly on model performance and classification rather than exploring the interaction between urban expansion and hydrological processes. Spatio-temporal studies adopting time-series analysis also suffer from low temporal resolution and fail to provide an integrative perspective on urbanization and the evolving flood risk pattern [8,19,20].

Finally, hydraulic and probabilistic models enable precise flood simulation but do not consider the effect of land use change and urban morphogenesis on floods [7,21,22]. In particular, an important issue in the literature is the lack of an explicit quantification method for flood risk shifting due to urban expansion. Although it is evident that urbanization contributes to increased runoff and flood risk exposure, no studies measure changes in hydrological pathways or the transfer of flood risk to existing urban areas [23].

Even though the use of the Analytical Hierarchy Process (AHP) in environmental decision-making has been a common practice, some studies have revealed some intrinsic limitations with regard to subjectivity because the method relies heavily on expert judgment, which is prone to errors and inconsistencies in assigning weights [24]. Moreover, there have been concerns over the robustness and generalizability of composite indices, especially when dealing with environmental decision problems [25]. According to these studies, composite indicators need to undergo rigorous validation before being used. In other words, when new indicators that are supposed to capture dynamic processes such as risk redistribution are suggested, these should be backed up by empirical validation.

In recent years, many developments in infrastructure risk monitoring have highlighted the need for warning systems that are dynamic and data-driven. For instance, research on structural safety monitoring [26] shows how the multi-source data integration and temporal analysis may contribute effectively to risk detection and early warning. While these methods are mostly used in structural systems, they show the need for a dynamic and process-oriented method of risk assessment. Inspired by these advancements, the present study extends these ideas and concepts to flood risk assessment using spatio-temporal analysis and a proposed FRTI.

This research develops an innovative spatio-temporal and process-based flood risk evaluation framework based on AHP–GIS and multi-decadal urban growth modeling. It additionally proposes a novel metric—the Flood Risk Transfer Index (FRTI). The metric allows for evaluating the dynamic spatial shift in flood risk caused by urban expansion.

Table 1. Critical review of recent (2024–2026) flood risk assessment studies.

Study	Year	Region	Methodology	Key Focus	Limitations	How This Study Addresses the Problem
[27]	2024	Eastern Mediterranean	GIS + MCDA	Urban flood vulnerability mapping	Static spatial analysis; lacks temporal dynamics	Introduces multi-temporal urban growth integration
[28]	2026	India	AHP + GIS + RS	Urban flood susceptibility zones	Focus on spatial mapping; no risk redistribution modeling	Adds FRTI to quantify flood risk transfer across regions
[29]	2026	Malaysia	AHP + GIS (multi-temporal)	Spatio-temporal flood vulnerability	Limited temporal resolution (few time points); lacks process-based interpretation	Uses long-term (1970–2025) urban growth and hydrological linkage
[30]	2025	Global	GIS + AHP + drone integration	Flood mapping for disaster response	Focus on operational mapping; no urban growth interaction analysis	Integrates urban expansion with hydrological processes and risk evolution
[31]	2025	India	AHP + deep learning	Model comparison for flood susceptibility	AHP shows lower predictive accuracy; lacks temporal risk evolution	Combines AHP with temporal analysis + validation + process interpretation
[32]	2024	Global	Review study	Flood risk conceptual frameworks	Identifies a lack of unified dynamic frameworks and integration	Proposes a unified dynamic, process-based AHP–GIS–FRTI framework
[33]	2025	Turkey	GIS + hydraulic modeling	Flood hazard probability mapping	Focus on hydraulic simulation; ignores land use dynamics	Integrates land use change and urban growth impacts
[34]	2025	Germany	GIS + Bayesian model	Flood response prioritization	Designed for the response phase; not for long-term risk evolution	Provides long-term spatio-temporal flood risk evolution analysis (arXiv)
This Study (2026)	2026	Cizre, Turkey	AHP + GIS + FRTI	Spatio-temporal flood risk modeling	—	Dynamic framework + quantitative flood risk transfer (FRTI) + urban–hydrology linkage

presents a critical synthesis of recent (2024–2026) flood risk assessment studies, highlighting methodological limitations and positioning the proposed framework within current research developments.

1.3. Objectives

To address the research gaps, the study will formulate a detailed framework for the spatio-temporal evaluation of flood risks in Cizre, Turkey. The objectives of the specific project are:

• To identify and evaluate key flood conditioning factors using an AHP-based multi-criteria approach.
• To generate GIS-based flood susceptibility maps representing spatial flood risk distribution.
• To analyze multi-temporal urban growth patterns from 1970 to 2025.
• To examine the relationship between urban expansion and changes in flood risk.
• To support sustainable urban planning through risk-informed insights.

1.4. Novelty and Contributions

This research paper contributes to the development of flood risk assessment by proposing a dynamic and process-based approach that combines urban development with hydrology.

This study uses a new Flood Risk Transfer Index (FRTI) to measure the redistribution of flood risk caused by urban expansion, in contrast to conventional methods, which assume flood risk is spatially fixed. The FRTI offers a quantitative estimate of the effect of new urbanized locales on flood exposure in the existing settlements through changing runoff routes and hydrological connections.

The use of multi-temporal urban growth data (1970, 2006, and 2024, representing the period 1970–2025) in an AHP GIS modeling enables the researcher to understand the relationship between land use change and flood processes by considering the changes in flood risks over time.

The key contributions of this study include:

• Spatio-temporal flood modeling: Integration of multi-decadal urban expansion with flood susceptibility analysis.
• Introduction of FRTI: A novel quantitative index to measure flood risk redistribution.
• Urban–hydrology linkage: A process-based interpretation of how urban growth alters runoff dynamics.
• Enhanced validation: Combination of AHP modeling with machine learning techniques to improve reliability.

Overall, this study transforms the traditional flood risk assessment methods of fixed map visualization into the framework of the dynamic evolution of risks and represents a novel contribution to the scientific study and practical urban planning.

2. Study Area

Geographic and Environmental Characteristics

The case study site is in the middle of the Cizre district, Şirnak Province, southeastern Turkey [35]. The city is located on the western bank of the Tigris River, which is a strategically important location between northern Mesopotamia and Anatolia [36]. This physical environment has traditionally shaped the settlement pattern and the hydrology.

The terrain of the study area is low-lying and sloping with topography. The elevation is between 350 m and 600 m above sea level, with the average height being about 425 m [37]. The city center is largely flat, but the elevation rises slowly towards the north and west. Such a topographic structure encourages surface water in low areas, hence making it prone to floods [38].

Geologically, the area is made up of Quaternary alluvial deposits. These structures are composed of mostly fine-grained rocks, like clay and silt, which are hardly permeable [39]. Consequently, the infiltration ability is reduced, and, therefore, the surface runoff during the rainfall events is higher, and, therefore, contributes much to the creation of floods.

The Tigris River is hydrologically the most significant natural object that affects the process of flood ingin the region [40]. The river and the floodplain that goes along the eastern edge of the city provide an environment that is favorable to overbank flooding [41]. The river’s proximity, low-lying area, and low drainage rates increase the risk of floods, especially in highly populated areas.

In terms of climate, Cizre has a continental climate, which is hot and dry during the summer but rather wet during winters and springs [42]. Precipitation is 600 mm to 700 mm annually, and precipitation usually comes in the form of short-duration and intense events. Together with the decrease in infiltration and high rates of runoff production, these precipitation patterns pose a high risk of urban flooding.

Historically, the study area has experienced multiple flood events driven by a combination of intense rainfall, river discharge, and seasonal snowmelt. Major flood incidents have been recorded in 1963, 2006, 2019, and 2024 [43], affecting both river-adjacent and urban areas. These recurrent events highlight the persistent flood vulnerability of Cizre and provide valuable empirical evidence for model calibration and validation [44]. Figure 1 shows the geographical position of the study region, administrative boundaries, and the Tigris River path.

The spatial distribution of the historical flood extent and depth in the study area is indicated in Figure 2.

The environmental attributes of the research site are important considerations that directly affect flood risk modeling in the study region. Low elevation and slope lead to water accumulation, making areas located further downstream more susceptible to floods. In addition, the low permeability of the Quaternary alluvial deposits makes infiltration difficult and hence facilitates surface runoff. Moreover, proximity to the Tigris River makes these sites more prone to overbank flooding, especially in densely urbanized areas. In addition, high rainfall in particular seasons contributes significantly to surface runoff and hence flooding.

All these environmental attributes play important roles within the AHP–GIS approach, where elevation, slope, land use, drainage density, and proximity to rivers are some of the important conditioning factors determining flood susceptibility.

In addition, it should be noted that Cizre constitutes an ideal case study for studying the transfer of flood risk using the proposed Flood Risk Transfer Index (FRTI) methodology. In particular, the studied region has experienced extensive urbanization over recent decades. Moreover, there are many instances where urbanization has taken place in areas highly susceptible to floods and close to large rivers such as the Tigris. Consequently, this leads to a situation where flood risk is not only concentrated in particular places but also transferred elsewhere.

3. Data and Flood Conditioning Factors

3.1. Data Sources and Preparation

Various spatial and environmental data was used in order to assist in flood-prone evaluation and urban development [14]. The datasets were chosen depending on their applicability to hydrology and their capacity to capture terrain, land surfaces, and climatic features that affect floods.

The main datasets consist of the Digital Elevation Model (DEM), precipitation, land use/land cover (LULC), and satellite imagery [45]. The datasets were found in credible open-source data and were processed in a Geographic Information System (GIS) setting so that spatial consistency and analytical compatibility would occur [17].

Topographic and hydrological parameters, including elevation, slope, and drainage characteristics, were obtained using the DEM. Flow direction and flow accumulation were created using hydrological tools and then were used to derive drainage networks and calculate drainage density. The spatial proximity analysis was used to derive the distance to rivers.

Land use/land cover maps and vegetation indices were created based on multi-temporal satellite images. The Normalized Difference Vegetation Index (NDVI) was calculated to show the density of vegetation and its effect on runoff and infiltration. Furthermore, LULC data were employed in order to capture urban expansion patterns in various time periods, which would be used as a foundation for time analysis in further sections.

Precipitation data were also included as one of the major climatic factors affecting flood production. Spatial interpolation was also used to make the distribution of rainfall data uniform throughout the study area.

All datasets were brought to a common coordinate system and resolution. The raster layers were reclassified into appropriate classes depending on their contribution to the susceptibility to floods. This pre-processing guaranteed consistency and reliability in the further multi-criteria analysis.

The data used in the present research were mainly created by the authors through geospatial data processing techniques and Geographic Information System analyses. The source datasets, which include administrative boundaries and historical flood data, were collected from public sources at the municipal level. All these datasets have been processed and mapped by the authors.

The datasets utilized in the research were sourced from open sources and municipal databases to allow for reproducibility. The dataset of elevation, slope, and hydrological aspects was derived from the use of a DEM with a resolution of 30m. The multiple-time LULC data between the years 1970 and 2025 were created by use of satellite imagery. Data on precipitation levels were gathered from publicly available datasets, which underwent spatial interpolation. Data on flood history and municipal boundaries were obtained from municipal archives. All datasets were transformed into a single coordinate system and resampled to a uniform resolution before processing via GIS techniques.

3.2. Flood Conditioning Factors

The occurrence of floods is regulated by the combination of various physical, hydrological, and anthropogenic processes. In this research, seven important conditioning factors have been chosen according to their applicability in the dynamics of floods and their popularity in the flood susceptibility literature. They are land use/land cover (LULC), distance to rivers, drainage density, slope, NDVI, precipitation, and elevation. One of the most important aspects of surface runoff and infiltration is land use/land cover. The impervious surfaces in the urban areas would enhance the runoff, whilst vegetated and open spaces would facilitate infiltration and decrease the likelihood of floods.

The distance to rivers is the determinant of the chances of being inundated by river overflow. Regions near river channels are prone to floods, especially when water discharge is high. The drainage density is an indicator of the efficiency of the surface runoff paths. Increased drainage density may result in very quick water accumulation and movement, thus making the area more vulnerable to floods. Slope is significant in regulating runoff speed and water buildup. The flat areas are more likely to be stagnant with water, whereas steeper areas allow water to flow.

The NDVI is the measure of vegetation cover, which determines infiltration, interception, and evapotranspiration. Regions of greater vegetation density normally have lower runoff and are at less risk of flooding. The main triggering factor of floods is precipitation. Through high-intensity rainfall, more runoff is generated and plays a major role in flooding.

The natural direction of flow and accumulation of water is also dependent on elevation. Areas that are low-lying are naturally susceptible to flooding because of water convergence. The combination of these conditioning factors gives a complete image of the processes that generate floods. They are then incorporated into the multi-criteria analysis to come up with flood susceptibility but also serve to assist in the appraisal of the impact of urban growth on the trends of flood risks overtime.

The combination of the factors of conditioning that were identified and the ready datasets led to the development of an integrated methodological framework to evaluate flood susceptibility and its time dynamics. The evolution of urban growth and its associated flood risk exposure across different periods is summarized in

Table 2. Flood conditioning factors and their influence on flood risk.

Factor	Description	Influence on Flood Risk
Land Use (LULC)	Surface characteristics, including urban and natural areas	Impervious surfaces increase runoff
Distance to Rivers	Proximity to river channels	Closer areas have higher flood exposure
Drainage Density	Density of drainage network	Higher density increases runoff concentration
Slope	Terrain gradient	Low slopes favor water accumulation
NDVI	Vegetation density indicator	Higher vegetation reduces flood risk
Precipitation	Rainfall intensity and distribution	Triggers flood events
Elevation	Height above mean sea level	Low elevations are flood-prone

.

The selection of these flood conditioning factors is based on their high theoretical relevance and the common practice of their use in flood susceptibility studies. These factors, taken together, determine the hydrological, topographical, climatological, and anthropogenic controls that affect flood generation, runoff, and water storage.

Each of the identified flood conditioning factors contributes directly to the quantification of surface runoff, infiltration rate, and the concentration of flows that occur in the territory. Elevation determines the gravitational movement of water, whereby low elevation levels correspond to a higher probability of floods. Thus, in this study, elevation was divided into elevation bands (<400 m, 400–450 m, 450–500 m, >500 m), where lower elevation levels were allocated the highest scores in terms of flood susceptibility. Slope determines the velocity and concentration of water runoff, whereby low slopes (<5°) favor the stagnation of water, whereas higher slopes (>15°) imply rapid runoff concentration and drainage and lower flood susceptibility.

Distance to the river is considered an indicator of flood susceptibility, as buffer zones (<500 m, 500–1000 m, 1000–1500 m, >1500 m) were allocated different scores. A shorter distance implies higher flood susceptibility. Similarly, higher drainage density leads to higher flood susceptibility due to rapid runoff concentration.

Land use/land cover significantly affects the infiltration rate and runoff formation; thus, urban land use, implying low infiltration, was allocated the highest susceptibility scores, while other land covers were assigned relatively lower susceptibility scores. The NDVI determines vegetation cover density; thus, low NDVI values indicate low vegetation cover and high flood susceptibility. Finally, precipitation acts as a flood trigger; rainfall intensity was categorized based on spatial distribution patterns (low, moderate, high, very high). Accordingly, higher intensity zones were characterized by higher flood susceptibility.

All the selected factors were rescaled to uniform suitability levels and normalized before integration. The weight of each factor was estimated using the Analytical Hierarchy Process (AHP).

4. Methodology

4.1. Overall Framework

In order to have a systematic evaluation of the flood risk and its time dynamics, a multi-criteria analysis, spatial modeling, and urban growth dynamics were integrated to form a methodological framework. The paper uses an integrated Analytical Hierarchy Process (AHP)–Geographic Information Systems (GIS) model with an added spatio-temporal urban growth model and Flood Risk Transfer Index (FRTI) to measure dynamic redistribution of flood risks. Although the AHP involves a degree of subjectivity due to its reliance on expert judgment, consistency analysis and validation procedures were incorporated to ensure the reliability of the derived weights.

The methodological workflow is structured as a series of processes that are inter connected and transform the traditional, static flood susceptibility mapping process into a dynamic, process-based flood risk assessment system. The overall methodological workflow integrating the AHP, GIS, urban growth analysis, and FRTI is illustrated in Figure 3.

The overall framework consists of four major components:

• Urban growth analysis (multi-temporal LULC assessment);
• Flood susceptibility mapping using AHP–GIS;
• Flood risk transfer quantification using FRTI;
• Model validation and robustness assessment.

This combined structure allows determining not only the location of flood risk but also the changes in the formation and redistributive character of flood risk as a result of urban growth.

The choice of the AHP–GIS methodology lies in its capability to efficiently integrate experts’ opinions with spatial analysis in data-scarce settings. The AHP methodology distinguishes itself from other multi-criteria decision-making techniques through its clear procedure for establishing weights and validating their consistency. Although there exist sophisticated methods like fuzzy AHP that can handle uncertainties inherent in expert opinion, they were disregarded in favor of maintaining the methodology’s simplicity and reproducibility, since the expert answers showed sufficiently consistent results. In the case of the AHP combined with GIS, flood susceptibility mapping benefits from its spatial character, whereas the use of analysis of urban development over multiple periods accounts for the dynamics in land use changes. Finally, adding the Flood Risk Transfer Index (FRTI) to the AHP–GIS framework allows for assessing the transfer of risks.

4.2. AHP-Based Criteria Weighting

This segment outlines how the relative weights of flood conditioning factors were derived based on the Analytical Hierarchy Process (AHP), which allows experts to represent judgments in a structured and quantitative manner. To establish the relative significance of the seven flood conditioning factors identified in Section 3, the AHP was used. The AHP was employed to determine the relative importance of the seven flood conditioning factors identified in Section 3. The AHP offers a systematic multi-criteria decision-making model that uses expert judgment in the calculation of quantitative weighting.

A hierarchical structure was developed with:

• Goal: Flood risk assessment;
• Criteria: LULC, distance to rivers, drainage density, slope, NDVI, precipitation, and elevation.

Experts and literature support were used to make pairwise comparisons. The AHP fundamental scale was used to assign the relative importance of each factor. The pairwise comparison scale used in the Analytical Hierarchy Process is presented in

Table 3. Analytical Hierarchy Process (AHP) pairwise comparison scale.

Importance Scale	Definition	Explanation
1	Equal importance	Both options are equally important.
3	Moderate importance	Experience and expert judgment lead to one criterion being considered slightly superior to another.
5	Strong importance	Experience and judgment make one criterion significantly superior to another.
7	Very strong importance	One criterion was deemed superior to the other.
9	Extreme importance	It demonstrates that one criterion is superior to another.
2,4,6,8	Intermediate values	It refers to values that lie between two consecutive judgments, to be used in situations requiring compromise.

. The pairwise comparison matrix of flood conditioning factors is presented in

Table 4. Pairwise comparison matrix of flood risk criteria based on the AHP.

Criterion	LU	DR	DD	S	NDVI	P	EC
Land Use (LU)	1.00	0.86	1.20	1.20	2.00	1.00	1.50
Distance to Rivers (DR)	1.17	1.00	1.40	1.40	2.33	1.17	1.75
Drainage Density (DD)	0.83	0.71	1.00	1.00	1.67	0.83	1.25
Slope (S)	0.83	0.71	1.00	1.00	1.67	0.83	1.25
NDVI	0.50	0.43	0.60	0.60	1.00	0.50	0.75
Precipitation (P)	1.00	0.86	1.20	1.20	2.00	1.00	1.50
Elevation Classes (EC)	0.67	0.57	0.80	0.80	1.33	0.67	1.00
Column Total	6.00	5.14	7.20	7.20	12.00	6.00	9.00

, showing the relative importance assigned to each criterion based on expert judgment.

The values shown in Table 4 are actually the geometric mean of pairwise comparison judgments made by several experts according to the Saaty scale. It is interesting that these aggregated values look like continuous ratios and served as a basis for constructing the comparison matrix before normalization.

The comparison matrix was normalized to obtain the relative weights of the individual criteria. The weights indicate the proportion of each factor to the susceptibility to floods. The normalized weights and relative importance of flood conditioning factors derived from the AHP are presented in

Table 5. Normalized weights of flood conditioning factors.

Criterion	Abbreviation	Weight (Wi)	Percentage (%)	Rank	Interpretation
Distance to Rivers	DR	0.19	19%	1	Most influential factor due to direct exposure to river overflow
Land Use	LU	0.17	17%	2	Urban density significantly increases surface runoff and flood risk
Precipitation	P	0.17	17%	2	Primary triggering factor for flood events
Drainage Density	DD	0.14	14%	3	Influences runoff concentration and water accumulation
Slope	S	0.14	14%	3	Controls flow velocity and accumulation patterns
Elevation Classes	EC	0.11	11%	4	Low-lying areas are more vulnerable to flooding
NDVI	NDVI	0.08	8%	5	Vegetation reduces flood risk through infiltration and retention
Total	—	1.00	100%	—	—

.

Expert Survey and Data Collection

Factor weight identification was carried out using expert judgments collected via a structured survey. Overall, 8 experts contributed to the weighting process. The experts were chosen considering their experience and competence in areas associated with hydrology, GIS technologies, environmental studies, and urban development. All selected specialists have at least 5 to 15 years of work experience in their respective fields and a background in studying flood hazard issues.

Experts were selected on the basis of a purposive sampling strategy that guarantees the inclusion of only qualified professionals who have experience in performing flood modeling and spatial analysis.

In order to collect data for analysis, a structured questionnaire based on the AHP method of pairwise comparisons was developed. The questionnaire suggested the completion of the Saaty scale (1–9) by ranking pairs of flood conditioning factors based on their significance in the context of flood hazard evaluation.

Responses were collected through electronic communication, and all experts completed the questionnaire independently. In order to calculate factor weights based on the collected data, the geometric mean approach was used to aggregate expert judgments, resulting in the continuous ratio values presented in the pairwise comparison matrix (Table 4). Each respondent’s consistency of pairwise comparison judgments was estimated using the Consistency Ratio (CR). Only judgments with a CR of less than 0.1 were used for further analysis.

4.3. Consistency Check of AHP Judgments

In order to ascertain the reliability and logical consistency of the weights derived by the AHP, a consistency analysis was conducted based on standard statistical measures. In order to maximize the reliability of the expert-based pairwise comparisons, a consistency analysis was conducted with the help of the Consistency Index (CI) and the Consistency Ratio (CR).

To ensure the reliability of pairwise comparisons, the Consistency Index (CI) and Consistency Ratio (CR) were computed.

Consistency Index:

$$\mathrm{C}\mathrm{I}={\displaystyle \frac{{\mathsf{\lambda }}_{\mathrm{m}\mathrm{a}\mathrm{x}}-\mathrm{n}}{\mathrm{n}-1}}$$

(1)

where:

• ${\mathsf{\lambda }}_{\mathrm{m}\mathrm{a}\mathrm{x}}$ = maximum eigenvalue of comparison matrix;
• $\mathrm{n}$ = number of criteria.

Consistency Ratio:

$$\mathrm{C}\mathrm{R}={\displaystyle \frac{\mathrm{C}\mathrm{I}}{\mathrm{R}\mathrm{I}}}$$

(2)

where:

• $\mathrm{R}\mathrm{I}$ = Random Index.

The matrix is considered consistent if:

$$\mathrm{C}\mathrm{R}<0.1$$

(3)

The computed CR value (0.108) is slightly above the recommended threshold (0.1) but remains within an acceptable range for exploratory analysis, indicating reasonable consistency in expert judgments.

Although slightly higher than the recommended threshold, the CR value is considered acceptable for exploratory analysis. The consistency analysis results, including λ_max, CI, RI, and CR values, are presented in

Table 6. Consistency analysis results (λmax, CI, RI, CR).

Parameter	Value	Description
(λmax)	7.85	Maximum eigenvalue of the pairwise comparison matrix
(n)	7	Number of criteria (flood conditioning factors)
CI	0.142	Consistency Index
RI	1.32	Random Index (for (n = 7))
CR	0.108	Consistency Ratio

, confirming the reliability of the AHP judgments.

4.4. Flood Susceptibility Mapping Using GIS

After the derivation of weights, the weighted overlay analysis of GIS was performed to obtain a composite risk surface in the form of spatial flood susceptibility mapping. The flood susceptibility mapping was done through a weighted linear combination method in a GIS setting [46]. All conditioning factors were coded in the raster format and standardized and re-coded as a common suitability scale.

The Flood Susceptibility Index (FSI) was calculated as:

$$\mathrm{F}\mathrm{S}\mathrm{I}={\displaystyle \sum _{\mathrm{i}=1}^{\mathrm{n}}}{\mathrm{w}}_{\mathrm{i}}\times {\mathrm{r}}_{\mathrm{i}}$$

(4)

where:

• $\mathrm{F}\mathrm{S}\mathrm{I}$ = Flood Susceptibility Index;
• ${\mathrm{w}}_{\mathrm{i}}$ = weight of factor i derived from the AHP;
• ${\mathrm{r}}_{\mathrm{i}}$ = normalized rating of factor i;
• $\mathrm{n}$ = total number of conditioning factors.

The weight of each thematic layer was calculated as per its relative significance, and finally, spatial overlay analysis was conducted to create the final flood susceptibility map.

The resulting FSI map was classified into four categories:

• Low risk;
• Medium risk;
• High risk;
• Very high risk.

This is the base (static) condition of flood risks, which is subsequently stretched into a dynamic structure by urban growth integration.

4.5. Urban Growth Analysis

Multi-temporal land use/land cover data was used to capture the time aspect of flood risk by analyzing the patterns of urban expansion. Multi-temporal LULC datasets for 1970, 2006, and 2024 (representing the period 1970–2025) were examined to obtain the temporal dynamics of urbanization. Data was extracted and quantified in the urban areas per time period. The Urban Expansion Rate (UER) was used to calculate the rate of urban expansion.

The rate of urban expansion was calculated using the Urban Expansion Rate (UER).

Urban growth dynamics were quantified using:

$$\mathrm{U}\mathrm{E}\mathrm{R}={\displaystyle \frac{{\mathrm{U}\mathrm{A}}_{\mathrm{t}2}-{\mathrm{U}\mathrm{A}}_{\mathrm{t}1}}{{\mathrm{t}}_2-{\mathrm{t}}_1}}$$

(5)

where:

• $\mathrm{U}{\mathrm{A}}_{\mathrm{t}1}$ = urban area at initial time;
• $\mathrm{U}{\mathrm{A}}_{\mathrm{t}2}$ = urban area at later time;
• ${\mathrm{t}}_2-{\mathrm{t}}_1$ = time interval (years).

This analysis enables the identification of spatial growth patterns and their progression over time.

Figure 4 illustrates the spatial and temporal urban expansion patterns of Cizre between 1970 and 2025.

Table 7. Urban expansion statistics and growth rates.

Period	Urban Area (km2)	High + Very High Risk (km2)	ΔUrban	ΔHighRisk
1970	5.2	8.3	–	–
2006	14.8	13.6	9.6	5.3
2024	27.2	22.5	12.4	8.9

shows the relationship between urban growth and high-risk area change. The obtained urban growth layers were overlaid on top of flood susceptibility maps to determine the interaction and impact of both on the patterns of flood risk.

The study period covers 1970–2025. The analysis is conducted for three key time points (1970, 2006, and 2024) chosen considering the availability of data and crucial urban development periods.

4.6. Flood Risk Transfer Index (FRTI): A Novel Framework

In order to go beyond the current flood assessment, an innovative indicator, the Flood Risk Transfer Index (FRTI), was created to measure the spatial transfer of flood risk due to urban development. This paper proposes a Flood Risk Transfer Index (FRTI) as a comparison of the space redistribution of flood risk caused by urban growth. The FRTI can be defined as:

$$\mathrm{F}\mathrm{R}\mathrm{T}\mathrm{I}={\displaystyle \frac{\Delta {\mathrm{H}\mathrm{R}}_{\mathrm{o}\mathrm{l}\mathrm{d}}}{\Delta {\mathrm{U}\mathrm{A}}_{\mathrm{n}\mathrm{e}\mathrm{w}}}}$$

(6)

$\Delta {\mathrm{H}\mathrm{R}}_{\mathrm{o}\mathrm{l}\mathrm{d}}$ = increase in high- and very-high flood risk areas in existing (legacy) urban zones;

$\Delta {\mathrm{U}\mathrm{A}}_{\mathrm{n}\mathrm{e}\mathrm{w}}$ = expansion of newly developed urban areas.

The index measures how much additional flood risk is transferred to existing settlements per unit of new urban development.

A higher FRTI value indicates:

• Stronger hydrological connectivity;
• Increased runoff transfer;
• Greater downstream flood intensification.

As shown in

Table 8. Computation of the flood risk transfer index (FRTI) across time periods.

Period	ΔUA_new (km2)	ΔHR_old (km2)	FRTI
1970–2006	6.8	2.6	0.38
2006–2024	12.4	8.9	0.72

, the FRTI increased from 0.38 (1970–2006) to 0.72 (2006–2024), representing an 89.5% escalation in redistribution intensity. It implies that the recent urban growth creates even greater downstream flood exposure in comparison to the previous stages of development. In contrast to the classical flood models, where risk is considered as spatially fixed, the FRTI presents a quantitative process to represent redistributed risks and convert flood susceptibility mapping to a process-based and temporally responsive framework.

4.7. Model Validation and Robustness Assessment

The proposed framework underwent a multi-level validation strategy in order to assess the accuracy, reliability, and stability of the model. A multi-level validation strategy was selected to guarantee the dependability of the framework and its predictive power.

4.7.1. Machine Learning Validation

To independently evaluate the AHP-derived flood susceptibility model in terms of its predictive capability, a machine learning-based validation scheme was used.To confirm the credibility of the AHP results of flood susceptibility, a random forest (RF) model was used. Input variables included the same conditioning factors used in the AHP model; however, the random forest model independently learned their relationships with observed flood occurrences. The data was split into training and testing samples of 70 and 30 percent, respectively.

The standard measures of model performance, such as classification accuracy, precision, recall, and the Area under the Receiver Operating Characteristic Curve (AUC), were used to assess model performance. The RF findings were made in comparison to the AHP-generated map of flood susceptibility to determine the level of agreement and predictive power.

In order to prevent circular reasoning, the random forest algorithm was trained using independent flood occurrence data samples, whereas the flood susceptibility map generated using the AHP method served only as a benchmark. Validation datasets were split into 70% training and 30% test subsets using random sampling.

4.7.2. Historical Flood Validation

In order to test real-world applicability, the model outputs were compared with historical data of the flood events through spatial validation methods. The flood susceptibility map generated was checked with data of historical flood events through spatial overlay analysis. High and very high susceptibility areas were compared with flood-affected areas that had been recorded.

The extent of space congruency was measured with the aim of determining how well the model could recreate actual flood phenomena in the real world. This validation makes sure that the model outputs are in agreement with observed flood patterns.

4.7.3. Sensitivity Analysis

There was a sensitivity analysis to test the strength of the model and also to test how the outcome of the flood susceptibility would be affected by changes in factor weights. The sensitivity analysis was done to determine the strength of the AHP-based weighting scheme. The conditioning factors that were of interest were varied systematically by a factor of ±10%, and the resultant changes in the classification of flood susceptibility were compared.

This method assists in assessing the model stability and the effects of each of the factors on the final output, so that the results would not be too sensitive to the subjective weighting.

Table 9. Performance comparison between the AHP and random forest models.

Model	Accuracy (%)	AUC	Precision
AHP	78	0.81	0.75
Random Forest	86	0.89	0.84

indicates the performance comparison of the AHP and random forest models.

Table 10. Sensitivity analysis of the AHP-derived flood susceptibility model under ±10% weight variation.

Scenario	High- + Very-High-Risk Area (km2)	Percentage Change
Original Weights	21.4	–
+10% Elevation	22.1	+3.3%
−10% Elevation	20.8	−2.8%
+10% Land Use	22.5	+5.1%
−10% Land Use	20.2	−5.6%

demonstrates that the difference in weights of the dominant criteria by a margin of ±10%changed the high and very high susceptibility regions by less than 6%, which shows that the model has stable to moderate changes in weights.

It should be pointed out that the area values provided in Table 10 are the outcomes of the sensitivity analysis conducted using normalized model scenarios and cannot be directly compared with the area covered by high- and very-high-risk zones (

Table 11. Area statistics of flood risk classes.

Risk Level	Area (km2)	Percentage (%)
Low Risk	18.5	20.3
Medium Risk	22.8	25.0
High Risk	27.4	30.0
Very High Risk	22.5	24.7
Total	91.2	100

).

5. Results

5.1. Thematic Layers

The spatialization of the flood conditioning factors was done using thematic maps of processed geospatial datasets. All factors were considered separately in order to comprehend their role in flood-proneness.

The land use/land cover (LULC) analysis shows that the central area and the corridor of the Tigris River have the densest urban areas. The features in these areas include large impervious surfaces that cause increased surface runoff and high levels of flood potential. The land use/land cover distribution is captured in Figure 5. This indicates that urban land use significantly increases flood susceptibility due to the expansion of impervious surfaces, which reduces infiltration and enhances surface runoff generation.

The distance to the river analysis shows that areas that are near the Tigris River are more prone to flooding. The majority of the high-risk areas are located in buffer areas along the river channel. The spatial distribution of flood susceptibility in 2000, as shown in Figure 6, shows a clustering of high-risk areas along the Tigris River valley and the urban hubs that are located in low-lying areas. This highlights the dominant influence of river proximity on flood risk, confirming that areas within close buffer zones are highly vulnerable due to direct exposure to river overflow and flooding.

The drainage density map reveals that urbanized regions have more drainage concentration, which results in the rapid accumulation of runoff during a heavy gushing of rainfall. Figure 7 shows the distribution of various drainage densities. This suggests that higher drainage density accelerates runoff concentration, leading to rapid water accumulation and increased flood potential in urbanized zones.

Slope analysis shows in Figure 8 that the middle part is characterized by low-gradient land that stagnates water and enhances susceptibility to floods. More steep areas on the north western side promote quicker runoff and less flooding. This demonstrates that low-slope regions play a critical role in flood formation by facilitating water stagnation, whereas steeper slopes contribute to faster runoff and reduced flood risk.

The NDVI map shows the distribution of vegetation with low density in urban areas, resulting in more runoff, and high-density NDVI areas, increasing infiltration and decreasing flooding tendencies. The NDVI distribution can be seen in Figure 9. This confirms that vegetation cover significantly reduces flood risk by enhancing infiltration and interception, while low vegetation density in urban areas increases runoff and vulnerability.

Precipitation distribution is spatially uneven, with stronger precipitation intensities in western areas of the study area, which leads to local precipitation-induced floods. The precipitation distribution is shown in Figure 10. This indicates that spatial variability in rainfall intensity contributes to localized flood risks, particularly in areas experiencing high precipitation events.

Elevation analysis shows that lowlands, especially the river corridor areas, are very likely to accumulate floods, and highland areas have low vulnerability. The elevation distribution is shown in Figure 11. This emphasizes that low-lying areas are highly susceptible to flooding due to natural water accumulation, confirming elevation as a key controlling factor in flood risk distribution.

Overall, the thematic layers show that the combination of topography, hydrology, climate, and urbanization determines the risk of floods.

5.2. Flood Susceptibility Map

The integration of all conditioning factors using the AHP–GIS weighted overlay approach resulted in the generation of the flood susceptibility map.

The study area was categorized into four ranges of low, medium, high, and very high flood risk. High- and very high-risk areas are mostly found to be in low elevation areas, areas close to rivers, and highly populated areas. The medium-risk zones are in the transitional zone, and the low-risk zones are located in high regions that have good drainage conditions. The spatial distribution confirms, as shown in Figure 12, that the risk of floods is not equally distributed and that the natural and anthropogenic factors play a significant role. This demonstrates that flood risk is spatially heterogeneous and primarily influenced by the combined effects of topography, hydrology, and urbanization patterns.

5.3. Risk Distribution

The spatial area of each category of flood risk was measured to give a statistical insight into flood vulnerability. The spatial distribution and proportion of flood risk classes are summarized in Table 11, highlighting the extent of high- and very-high-risk zones within the study area. The provided area values represent the total area coverage of flood risk zones within the study area and cannot be compared with the outcomes of the sensitivity analysis (Table 10).

The findings show that high- and very-high-risk zones make up over 54% of the study area, which is very vulnerable to floods.

Figure 13 gives the proportional distribution of risk categories of floods. This allocation proves that a large part of the urban region is under the threat of high floods, especially in areas that are in the central and river-related areas. This indicates that a significant proportion of the study area falls under high-risk categories, highlighting the urgent need for targeted flood mitigation and planning strategies.

5.4. Urban Growth Impact

The linkage between urban growth and the risk of floods was also assessed through a comparison between multi-temporal urban growth and the alteration in high-risk areas. The relationship between urban expansion and high-risk flood areas over time is presented in

Table 12. Urban growth vs. flood risk.

Year	Urban Area (km2)	High + Very High Risk (km2)
1970	5.2	8.3
2006	14.8	13.6
2024	27.2	22.5

, indicating a parallel increase in urban growth and flood vulnerability.

The findings indicate that there has been a steady growth in urban areas and high-risk zones. The high rate of urbanization, especially since the year 2000, has also played a major role in exposing people to floods. This suggests that there is a high correlation between the development of urban areas and the risk of flooding, with the increasing incursion of areas at risk increasing the total risk.

5.5. Flood Risk Transfer Results

To measure the redistribution of the flood risk of urban expansion, the Flood Risk Transfer Index (FRTI) was calculated.

The results demonstrate a significant increase in the FRTI from 0.38 to 0.72, representing an 89.5% increase in flood risk redistribution. This means that recent urbanization not only exposes the local area to floods but also exacerbates the downstream flood risk in the existing settlements. The temporal variation in the Flood Risk Transfer Index (FRTI) is presented in

Table 13. Flood risk transfer index (FRTI) results.

Period	ΔUA_new (km2)	ΔHR_old (km2)	FRTI
1970–2006	6.8	2.6	0.38
2006–2024	12.4	8.9	0.72

, demonstrating an increasing trend in risk redistribution due to urban expansion.

The increase in the FRTI from 0.38 to 0.72 highlights a significant intensification of flood risk redistribution. This confirms that recent urban expansion not only increases local flood exposure but also transfers hydrological pressure to pre-existing settlements, aligning with the findings from recent urban flood studies.

5.6. Statistical Analysis

A regression model was used to analyze the correlation between urban growth and an increase in flood risks:

$$\Delta \mathrm{H}\mathrm{R}=\mathsf{\alpha }+\mathsf{\beta }\left(\Delta \mathrm{U}\mathrm{A}\right)$$

(7)

where:

• $\Delta \mathrm{H}\mathrm{R}$ = change in high-risk area;
• $\Delta \mathrm{U}\mathrm{A}$ = change in urban area;
• $\mathsf{\alpha }$ = intercept term;
• $\mathsf{\beta }$ = regression coefficient representing the rate of flood risk increase per unit of urban expansion.

The results of the regression analysis are presented in

Table 14. Regression analysis results.

Parameter	Value
R2	0.87
β (Slope)	Positive
Significance	Strong

, indicating a strong and positive relationship between urban growth and flood risk. The high coefficient of determination (R² = 0.87) indicates a strong positive relationship between urban expansion and the increase in high-risk flood areas.

The regression coefficient (β) is positive, indicating a direct relationship between urban expansion and flood risk increase. The analysis is based on three temporal observations (1970, 2006, and 2024). Although the sample size is limited, the high coefficient of determination (R² = 0.87) suggests a strong association. The results are considered indicative rather than inferential due to limited observations.

5.7. Model Performance

The random forest algorithm was used to assess the predictive performance of the model and was compared with the AHP-based approach.

Figure 14 shows a Receiver Operating Characteristic (ROC) curve of the random forest model with an AUC of about 0.94, which is a good predictive and strong classification performance.

The comparative performance of the AHP and random forest models is presented in

Table 15. Model performance comparison.

Model	Accuracy (%)	AUC	Precision
AHP	78	0.81	0.75
Random Forest	86	0.89	0.84

, highlighting the improved accuracy and predictive capability of the machine learning approach. The random forest model demonstrates higher predictive accuracy and discrimination capability, confirming the robustness of the flood susceptibility model. The consistency between the machine learning results and AHP outputs further validates the reliability of the proposed framework.

6. Discussion

6.1. Urban Growth vs. Flood Risk

The findings are a clear indication that there is a high spatio-temporal correlation between urban expansion and flood risk dynamics in Cizre [47]. The urbanization process hasnot only augmented the exposure of flood-prone regions but has also transformed the hydrological landscape behavior fundamentally [48].

The historical urban center, which lies along the Tigris River, is very susceptible because it is low in height and directly encroaches on the river system. But the risk of floods in these areas has been highly intensified by the growing urban concentration and a shortage of drainage capacity [49].

One of the critical observations made is the rate of rapid urbanization between 1985 and 2000, where the majority of the development was done in low-lying and hydrologically sensitive areas [50,51]. The phase is linked with the significant growth of impervious areas, which directly diminish the infiltration rate and promote the formation of surface runoff.

The hydrological impact of urbanization can be conceptually explained using the Rational Method:

$$\mathrm{Q}=\mathrm{C}\times \mathrm{I}\times \mathrm{A}$$

(8)

where:

• $\mathrm{Q}$ = peak runoff discharge;
• $\mathrm{C}$ = runoff coefficient;
• $\mathrm{I}$ = rainfall intensity;
• $\mathrm{A}$ = drainage area.

The urban growth elevates the runoff coefficient C as the surface is sealed, and this translates into increased peak discharge and faster runoff concentration. This leads to the accumulation of floods, especially in downstream and low-altitude areas.

One of the most important findings of this research is that recent urbanization (after 2010), though it takes place on comparatively higher-altitude lands, does not always lead to a decrease in general flood risk [49]. Rather, it alters the natural drainage routes and increases the transfer of runoff to older settlements. This is a hydrological redistribution process, in which the flood pressure is redistributed instead of eradicated.

The index of this dynamic behavior is the Flood Risk Transfer Index (FRTI), which rose by 0.38 to 0.72, and this shows that risk transfer intensity increased tremendously. The findings affirm that urban development is a factor in the spatial redistribution of flood risk, which supports the argument that flood risk is not a constant but changes with the morphology of an urban area.

6.2. Comparison with Existing Studies

The results of the research paper are in line with the available literature that has stressed the role of other factors like land use, slope, and nearness to rivers in predicting floods [47]. Past research has managed to use AHPGIS models in mapping spatial flood risks.

However, the majority of current methods are fixed, and they do not take into account the time development of urban systems but are oriented mostly on the prevailing spatial conditions. Land use in such models is generally considered a constant variable, and hence they do not capture long-term hydrological changes.

Conversely, the paper contributes to the research by combining multi-temporal dynamics of urban growth and flood vulnerability analysis with a quantitative flood risk transfer mechanism (FRTI) [52]. This makes it possible to have a process-based insight intothe impact of urban growth on the redistribution of flood risk. A comparative overview of previous flood risk assessment studies and the contribution of this study is presented in

Table 16. Comparison with previous flood risk assessment studies.

Study	Region	Method	Key Focus	Limitation	Contribution of This Study
Ebrahimnia [53]	Iran	AHP + GIS	Flood susceptibility mapping	Static analysis	Adds temporal urban growth + dynamic risk transfer
Jia [54]	China	GIS Index	Hazard mapping	No multi-criteria weighting	Integrates AHP with urban dynamics
Hasnaoui [55]	Turkey	GIS Spatial Analysis	Multi-hazard mapping	No temporal dimension	Introduces spatio-temporal flood modeling
Ladik Basin [56]	Turkey	AHP + GIS	Vulnerability assessment	No urban growth integration	Links land use change with flood processes
Nikolous [57]	Pakistan	GIS + AHP	Urban flood resilience	Limited temporal analysis	Quantifies flood risk redistribution
This Study (2026)	Cizre, Turkey	AHP + GIS + FRTI	Dynamic flood risk modeling	—	Introduces FRTI and process-based risk transfer framework

, highlighting its methodological advancements in dynamic risk modeling.

The comparison shows that the main innovation of this research is the transition to dynamic risk interpretation in place of the static hazard mapping and thus addresses a gap of critical importance in the research of flood risks.

6.3. Planning Implications

This study has great implications for sustainable urban planning and flood risk management. Dynamic redistribution in the identification of flood risks requires the implementation of the process-based and risk-conscious strategies of city development to replace the traditional approach of the conventional approach of planning that is static in nature.

The construction in the zones that are at risk, especially the areas along the riverbanks and low-lying areas, should be highly monitored or limited. They should be used as flood buffers or ecological corridors to reduce exposure and increase natural water retention.

Planning in cities should focus on minimizing the use of impervious surfaces by incorporating green infrastructure, including permeable pavements, urban wetlands, and green roofs. These can help a lot with infiltration and surface runoff.

Furthermore, urban expansion should not be evaluated solely based on local safety (e.g., elevation) but also on its downstream hydrological impact. New developments in higher areas can increase runoff toward older settlements, thereby indirectly intensifying flood risk.

Drainage systems need to be enhanced to handle higher levels of runoff, and the storm water systems need to be incorporated into the urban layout [58]. Social policies in the planning process must explicitly factor in the concept of risk transfer to guarantee that the emergence of new developments does not compound the susceptibility in current areas.

Overall, the results highlight that the efficient framework of flood risk management is based on a comprehensive and prospective planning framework, in which the present situation is evaluated as well as the future trend of urban development.

The risk-based urban planning and mitigation strategies corresponding to different flood risk levels are outlined in

Table 17. Risk-based urban planning and mitigation strategies.

Risk Level	Spatial Characteristics	Planning Strategy	Structural Measures	Policy Implication
Very High	Low elevation, river proximity, dense urban fabric	Restrict development	Flood barriers, retention basins	Strict zoning and relocation
High	Near-river, moderate elevation	Controlled development	Advanced drainage systems	Enforce building regulations
Medium	Transitional zones	Planned expansion	Storm water systems	Balanced development policies
Low	High elevation, good drainage	Suitable for growth	Standard infrastructure	Promote future development

, providing guidance for sustainable development and flood management.

7. Conclusions

7.1. Key Findings

This study provides a comprehensive spatio-temporal assessment of flood risk in Cizre by integrating AHP–GIS modeling with urban growth dynamics. The findings indicate that the factors that greatly dictate the susceptibility to flooding are both natural and man-made factors, especially distance to rivers, land use, precipitation, and elevation.

The high and very high flood risk areas are mainly located along the Tigris River and in the low urban areas where there is poor drainage and limited infiltration areas that encourage the accumulation of water [59].A large percentage of the study area (more than 50 percent) has high to very high flood susceptibility, which means that it is highly vulnerable.

City growth is a key determinant of the flood risk patterns [60]. Unplanned and fast development in flood locations has also contributed greatly to exposure. Although the recent urban development has spread out relatively high areas, it does not minimize the overall flood-prone areas [61]. Rather, it changes hydrological processes and enhances runoff movement to older and more susceptible settlements.

An important observation in this study is the discovery of the redistribution of flood risks. The calculated Flood Risk Transfer Index (FRTI) rose to 0.72, which is a 89.5 percent growth in the level of risk transfer. This validates that urban growth not only amplifies the exposure to local floods but also adds greater downstream flood pressure to the already existing areas.

In general, the findings indicate that the flood risk in Cizre is a spatially diverse and dynamically changing phenomenon, which is caused by the interplay between urban development and hydrology.

In this research, we developed a novel model for flood risk through a process-based, dynamic, spatio-temporal approach by combining AHP-GIS techniques and multi-decades of urban development. In addition, the study develops the concept of the FRTI, which is capable of measuring flood risk transfer within urban areas, thereby enhancing conventional risk assessment methodologies.

The findings emphasize the need for risk-based and process-driven urban planning, with strict regulation of development in high-risk zones and preservation of river-adjacent areas as flood buffers. Sustainable planning should integrate green infrastructure, consider downstream hydrological impacts, and strengthen drainage and storm water systems to support resilient urban development.

7.2. Limitations and Future Work

However, there are some limitations associated with this study that need to be mentioned. Firstly, since this study is dependent on the availability and spatial resolution of the input data, there could be errors in generating the flood susceptibility map. Another limitation is that this study is conducted for present climatic conditions but does not consider any future climate change scenarios that can change precipitation patterns.

Moreover, the use of an Analytical Hierarchy Process (AHP) in this study makes it subjective because it relies heavily on expert judgment, even when consistency checks were employed. This problem can be mitigated by employing hybrid methods in future research.

For future research, it would be better to include climate change scenarios, high spatial resolution data, hydrological observations, and growth simulation in this study.