An Example of Vulnerability Analysis According to Disasters: Neighborhoods in the Southern Region of Izmir

Abstract

This study analyzes disaster vulnerability in rural settlements in İzmir’s southern region in Türkiye. Recognizing that vulnerability significantly contributes to disaster risk, the research investigates social, economic, natural, and artificial vulnerability factors. The primary goal is to assess vulnerability levels and propose disaster-sensitive planning strategies, especially for rural settlements. This research focuses on six districts—Selçuk, Bayındır, Tire, Beydağ, Ödemiş, and Kiraz—which include 341 neighborhoods, 75% of which are rural. It aims to measure vulnerability using specific social, economic, natural, and artificial indicators, addressing risks from earthquakes, floods, and landslides. This study intends to inform future planning strategies to enhance disaster resilience at the neighborhood and district levels. The analysis employed a scoring system for vulnerability indicators, assigning weights between 0 and 1 based on risk levels. Social and economic factors were uniformly assessed across disaster types, while natural and artificial factors were evaluated separately for earthquakes, floods, and landslides. Geographic Information System (GIS) tools were used to map and visualize vulnerability scores, with statistical analyses conducted at neighborhood and district scales. The results highlight significant variability in vulnerability levels across districts and neighborhoods. Differentiated strategies are essential for addressing disaster risks in rural areas. This study recommends prioritizing districts based on their vulnerability profiles and integrating disaster-sensitive planning into regional development strategies. These findings contribute to the ongoing discourse on vulnerability analysis and rural disaster resilience planning.

1. Introduction

Disasters can occur as natural or human-induced events and cause significant physical, economic, and social damage in the regions where they occur [1]. However, the transformation of an event into a disaster is not solely dependent on the presence of a physical hazard; it is also directly related to how vulnerable society is to this hazard due to its social, economic, and demographic characteristics [2,3,4], as well as how prepared the community is for disasters [5,6,7].

In this context, vulnerability or potential for damage indicates how defenseless a community, individual, or system is against disasters, and it plays a significant role in determining disaster risk [8]. Therefore, hazard and vulnerability factors shape a region’s disaster risk [9].

Consequently, focusing solely on the hazards that occur is insufficient to understand and mitigate the impacts of disasters. Attention must also be paid to the vulnerability factors that render society, individuals, and systems defenseless against these hazards [10].

Studies have shown that vulnerability levels are measurable, and they should be examined using varying indicators depending on different types of disasters [11,12,13,14,15,16]. Therefore, the results of these measurements will provide important data for developing strategies at both planning and implementation scales.

When examining the national and international literature, it is observed that studies related to vulnerable areas in disasters focus on physical, social, economic, and environmental factors. Many studies have been conducted to identify which groups or areas are more sensitive to the effects of disasters [17,18,19,20,21,22,23]. The prominent findings in this context are as follows:

• Disasters affect all individuals in society; however, they have much more devastating effects on vulnerable groups such as children, the elderly, women, and people with disabilities.
• The difficulties caused by conditions specific to rural areas and other challenges faced by disadvantaged groups make these individuals more vulnerable to disaster risks compared to those living in urban areas.
• The socio-economic and spatial characteristics of rural settlements play a significant role in determining resilience levels.
• Physical and structural factors (such as high-density construction, infrastructure deficiencies, and the impacts of building quality) influence vulnerability levels.

These listed factors indicate that social sustainability should be addressed alongside spatial sustainability. When it comes to rural areas, due to their higher vulnerability compared to urban areas, vulnerability assessments have become increasingly important in rural planning strategies and practices.

Türkiye is a country that faces a high level of risk from both earthquakes and other natural hazards. These risks are exacerbated not only by natural conditions but also by existing issues related to urban development and infrastructure [24,25,26,27,28]. The two earthquakes that occurred in Kahramanmaraş in 2023 resulted in the loss of more than 50,000 lives and the destruction of 37,984 buildings, thereby providing a stark illustration of the country’s vulnerability to disasters [29,30]. Therefore, disaster-related strategies in Türkiye should be developed separately for urban and rural areas. In the case of rural areas, it is essential to recognize their heightened vulnerability stemming from their socio-spatial characteristics. Although the indicators used to assess rural vulnerability resemble those applied in urban contexts, the socio-spatial attributes highlight the necessity of incorporating place-specific considerations into strategic frameworks. Common challenges in rural areas include limited transportation connectivity, restricted access to resources, and deficiencies in technical infrastructure such as internet services. Nevertheless, beyond these shared fundamental issues, rural areas also exhibit distinctive place-specific problems and characteristics [31,32,33,34].

In addition to being a country with high disaster risk, Türkiye also prioritizes urban areas in its planning legislation and practices, making studies on the vulnerability of rural areas significant. This study aims to contribute to the existing literature by discussing how and with which indicators vulnerability should be addressed throughout rural regions. Given the hierarchical planning structure, all settlements facing earthquake and disaster risks must be planned and designed to ensure the sustainability of their social, cultural, and economic structures, from higher to lower scales. In other words, as a field of science and a profession, urban planning should develop disaster-sensitive implementation examples from the regional scale to the urban and neighborhood scale. This study aims to contribute to efforts to identify the vulnerability status of rural settlements at the regional scale within urban planning.

Studies conducted to measure vulnerability have used various indicators to create resilient settlements. These indicators are generally categorized into different groups such as physical, social, economic, ecological, environmental, cultural, attitudinal, and institutional [35,36,37,38,39,40,41]. This study also aims to conduct a vulnerability analysis at the district and neighborhood scale according to the relevant categories and indicators required by the study area, to contribute to the vulnerability measurements of similar settlements and to open discussions on the planning strategies to be developed for such settlements. Thus, this study emphasizes that vulnerability should be addressed with different datasets and indicators at different scales, and that only in this way can the planning efforts at different scales gain the quality to guide future studies.

The study area includes the districts of Selçuk, Bayındır, Tire, Beydağ, Ödemiş, and Kiraz, which are located in the southern region of İzmir, known for its agricultural activities and rural settlements. In addition to these characteristics, the area has high disaster risk due to the ground characteristics shaped by the Küçük Menderes River. A significant earthquake measuring 6.9 occurred on 30 October 2020, and was felt intensely in the rural area. Below are visuals showing the location of the study area in Türkiye and İzmir (Figure 1), as well as urban and rural settlements within the study area (Figure 2).

There are a total of 341 neighborhoods within the study area borders covering the districts of Selçuk, Bayındır, Tire, Beydağ, Ödemiş, and Kiraz. A total of 255 of these neighborhoods are rural, and 86 are urban. The table showing the number and characteristics of neighborhoods by district is given below (

Table 1. Number and characteristics of neighborhoods by districts.

District	Number of Urban Neighborhoods	Number of Rural Neighborhoods	Total
Selçuk	6	8	14
Bayındır	23	36	59
Tire	24	64	88
Beydağ	4	21	25
Ödemiş	25	74	99
Kiraz	4	52	56
Total	86	255	341

). The figure below illustrates the locations, characteristics, and populations of the neighborhoods within each district (Figure 3).

2. Methodology

The selection of key indicators in this study is grounded in the body of literature reviewed in the introduction. While internationally recognized frameworks—such as those developed by the Intergovernmental Panel on Climate Change (IPCC) and the United Nations Office for Disaster Risk Reduction (UNDRR)—employ well-established indicator sets, their direct application in the present research is constrained by two primary factors: the limited availability of relevant datasets in Türkiye and the incompatibility of these indicators with the spatial scale of the analysis (i.e., the neighborhood level). For example, the IPCC’s indicators are predominantly climate-oriented, encompassing variables such as greenhouse gas emissions and concentrations, radiative forcing, surface temperature change, anthropogenic contributions to warming (attribution), the global carbon budget, energy imbalance, sea-level and precipitation changes, and the frequency and intensity of extreme temperature events. By contrast, the indicators articulated within the UNDRR’s Sendai Framework for Disaster Risk Reduction (2015–2030) are aligned with the principles of sustainable development and oriented towards disaster risk reduction, including mortality and loss rates, the number of affected individuals, direct economic losses, damage to infrastructure and disruption of essential services, the adoption rate of national and local disaster risk reduction strategies, international cooperation and support, the availability of multi-hazard early warning systems, and access to risk information [42,43,44]. Consequently, given that the primary objective of this study is to conduct a vulnerability assessment using location-specific datasets, the indicators have been selected on the basis of their relevance to the local context. Nevertheless, certain overlapping indicators—such as those relating to the affected population and infrastructure characteristics—have also been incorporated.

This study first established a dataset specific to rural settlements to identify the area’s unique characteristics. In the second phase, these indicators were categorized under social, economic, natural, and artificial factors. A scoring system was developed, considering the impact levels of each indicator on the identified disaster types. The main elements of the method used in this study and the scoring structure are illustrated below (Figure 4).

Indicators pertaining to social, economic, natural, and artificial vulnerabilities were derived from the relevant literature. The underlying assumptions regarding these indicators, the rationale for their significance, and the supporting scholarly evidence are presented in

Table 2. Social vulnerability indicators.

Indicators	Assumptions	Rationales	References
Total Population (2023)	As the population increases, social vulnerability tends to rise.	An increase in loss of life and property A growing demand for infrastructure Exceeding the capacity of existing infrastructure Difficulties in evacuation and emergency response during and after disasters	[45,46,47,48,49,50,51,52,53]
Gender	Women exhibit higher levels of social vulnerability.	The responsibility of women for dependents such as children, the elderly, and persons with disabilities The attribution of family care responsibilities to women due to gender roles Higher levels of psychological impact observed among women An increased risk of gender-based violence against women during disasters Disaster management policies being gender-blind The neglect of women’s poverty status in disaster management policies More limited access of women to education, health, and social services	[47,54,55,56,57,58,59,60,61,62,63,64,65]
Child Population (0–14) Elderly Population (65+)	Children and the elderly exhibit heightened levels of social vulnerability.	Lower physiological, psychological, and social adaptability of children and the elderly to disasters and post-disaster interventions Higher levels of psychological impact among children and the elderly Specific needs of children and the elderly during disasters Increased risk of violence and exploitation targeting children and the elderly in disaster contexts Deterioration and inadequacy of social infrastructure (e.g., education, health, rehabilitation) serving children and the elderly Increased risk of disease during disasters Prevalence of chronic illnesses, mobility limitations, and disruptions in care chains Greater sensitivity of the elderly to climatic conditions	[47,60,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86]
Number of Household Members	As household size increases, social vulnerability also rises.	Increased costs of evacuation and disaster adaptation Greater challenges in meeting disaster preparedness requirements Heightened demand for designated assembly areas Strain on resources, food supplies, and shelter capacity Increased risk of loss of life and property	[47,87,88,89,90,91,92,93,94,95]

,

Table 3. Economic vulnerability indicators.

Indicators	Assumptions	Rationales	References
Child Dependency Ratio Elder Dependency Ratio	As the dependency ratio of children and the elderly increases, economic vulnerability correspondingly intensifies.	Increased caregiving burden Reduction in productive capacity Lower adaptive capacity of elderly and children to disasters and post-disaster interventions from physiological, psychological, social, and economic perspectives Weakening of economic recovery processes Requirement for psycho-social interventions Strain on logistical infrastructure Pressure on financial and human resources	[47,88,89,95,96,97,98,99]
Primary Economic Activity Type	The primary type of economic activity influences the level of economic vulnerability.	Dependence on a single sector weakens economic resilience A multi-sectoral economic structure strengthens resilience to disasters and reduces external dependency Tourism and agriculture are more severely affected by disasters The dominant sector determines production and trade chains Labor-intensive sectors require longer recovery periods following disasters	[100,101,102,103,104,105,106,107,108,109]
Road Connection Railway Connection	The absence and/or limitation of road and railway accessibility increases economic vulnerability.	Delays in the economic recovery of the disaster-affected region Suspension of port operations (if applicable) Increases in costs within the emergency response supply chain Disruption or halting of production and trade activities Undermining of the continuity of economic activities, including the flow of goods, services, and labor	[110,111,112,113,114,115,116,117]

,

Table 4. Artificial vulnerability indicators.

Indicators	Assumptions	Rationales	References
Settlement Area Type	The proximity of settlements to coastal zones, rivers, and lakes increases disaster vulnerability, particularly in relation to earthquakes, floods, and landslides.	The inundation or overflow of water during disasters can cause significant damage to residential areas. The inundation or overflow of water can adversely affect production zones and economic activities. The inundation or overflow of water can impair infrastructure, disrupt transportation systems, and damage critical support and logistics facilities.	[60,118,119,120,121,122,123,124,125,126,127]
Settlement/Building Block Form	The spatial form of a settlement (linear, dispersed, or grid) influences its disaster vulnerability.	Dispersed and unplanned settlements exhibit heightened seismic vulnerability. Grid settlement forms facilitate emergency response and intervention. Dispersed settlements are associated with higher levels of human and material losses. Compact and dense settlements demonstrate elevated disaster risk. Compact settlements are particularly susceptible to flood hazards. Emergency response and intervention are more challenging in dispersed settlements.	[128,129,130,131,132,133,134,135,136,137,138,139,140,141]
Structure Relationships	The spatial relationship between buildings (whether contiguous or detached) influences disaster vulnerability.	In attached (adjacent) building configurations, the impact of the pounding effect is significantly higher. Block-type structures are particularly more vulnerable to seismic hazards. Detached building configurations exhibit greater resistance, especially against earthquakes. Adjacent structures with varying floor heights are more prone to vulnerability. Detached and dispersed building configurations are associated with lower levels of human and material losses. Emergency response is more challenging in densely built-up areas with attached building patterns.	[142,143,144,145,146,147,148,149]
Number of Floors	Building height significantly influences disaster vulnerability.	Mid-rise structures are generally more vulnerable to earthquakes. Low-rise structures are more susceptible to flood-related vulnerabilities. In high-rise buildings, the duration of emergency response and evacuation is prolonged. High-rise developments increase impervious surface cover, exacerbating hazard impacts. High-rise structures tend to exhibit greater overall vulnerability to disasters.	[139,150,151,152,153,154,155,156,157,158,159,160]
Drinking Water Infrastructure Sewage Infrastructure	The condition of potable water and sewerage infrastructure is a critical factor influencing disaster vulnerability.	The presence of infrastructure and potable water systems mitigates vulnerability across all types of disasters. The availability of infrastructure and potable water reduces the risks of epidemics and population displacement.	[161,162,163,164,165,166]

and

Table 5. Natural vulnerability indicators.

Indicators		Assumptions	Rationales	References
Height Above Sea Level (m)		Elevation above sea level influences vulnerability to disasters.	Low-lying settlements exhibit higher vulnerability to earthquakes and floods. The high proportion of populations residing in low-lying settlements increases the likelihood of significant human and material losses.	[167,168,169,170,171,172,173,174,175]
Slope Condition (%)		Slope conditions influence vulnerability to disasters.	Settlements located on low slopes exhibit higher vulnerability to floods. Settlements located on steep slopes are more vulnerable to earthquakes and landslides.	[176,177,178,179,180,181,182,183,184,185]
Geological Formation Type		Geological formation types influence vulnerability to earthquakes.	The presence of alluvial and/or loose fill soils increases seismic risk.	[186,187,188,189,190,191,192,193,194]
Fault Line Nature Proximity to Fault Line (m)		The presence of fault lines and proximity to them significantly influence earthquake vulnerability.	The proximity of a settlement to a fault line increases vulnerability.	[195,196,197,198,199]
	Proximity to the Shore (m) Proximity to Streams (m)	The proximity of a settlement to coastal areas, rivers, or lakes increases disaster vulnerability, particularly in relation to earthquakes, floods, and landslides.	Damage to residential areas due to flooding and water overflow in all disaster types.	[60,118,119,120,121,122,123,124,125,126,127]
	Soil Class	The quality of agricultural land influences disaster vulnerability, particularly with respect to floods and landslides.	Damage to production areas resulting from flooding and water overflow. Damage to infrastructure, transportation systems, and support–logistics nodes caused by flooding and water overflow. Increased risk of floods and landslides in low-quality or unproductive agricultural lands.	[180,192,200,201,202]
Natural Land Use		The land use patterns of agricultural areas (e.g., olive groves, vineyards, orchards) influence disaster vulnerability in terms of landslides.	Cultivated agricultural lands and forested areas exhibit lower vulnerability to landslides.	[203,204,205,206,207,208,209]

.

In this study, vulnerability indicators were rescaled to a 0–1 range using min–max normalization to ensure cross-indicator comparability [210,211,212,213]. This method mitigates weighting imbalances arising from scale differences among indicators, thereby enabling all indicators to contribute on an equal basis. Normalization is a statistical transformation technique that allows indicators measured in different units and ranges to be made comparable. Following normalization, all indicators were assigned equal weights, with each weight calculated as the reciprocal of the total number of indicators [214,215,216,217].

Consistent with the literature, this study applied the 0–1 range weighting, which is widely adopted in vulnerability assessments. In the scoring process, indicators considered no-risk were given a score of 0, while those posing very high risk were assigned a score of 1. However, since the number of sub-indicators varied for each indicator, intermediate values differed accordingly. For instance, in the case of artificial vulnerability, building height was assessed in three sub-groups, leading to scores of 0, 0.33, and 0.67 for intermediate levels, while settlement type was assessed in four sub-groups, resulting in scores of 0.25, 0.50, and 0.75. In addition, elements not covered by the classification were scored on a binary (0–1) basis. For instance, with respect to drinking water infrastructure, a value of 0 was assigned if available and 1 if absent.

Since the indicators and weights determining artificial and natural vulnerabilities differed according to disaster types, separate calculations were made for earthquakes, floods, and landslides. However, as shown in

Table 6. Social and economic vulnerability scores.

Social Vulnerability			Economic Vulnerability
Indicators	Sub-Indicators	Weighting	Indicators	Sub-Indicators	Weighting
Total Population *	32–672	0.2	Child Dependency Ratio *	0–15	0.33
	673–1626	0.4		15.01–30	0.67
	1627–3758	0.6		30.01+	1
	3759–7187	0.8	Elder Dependency Ratio *	0–20	0.33
	7188–11,315	1		20.01–40	0.67
Gender *	Women	1		40.01+	1
	Men	0	Primary Economic Activity Type *	Private (Central Neighborhood)	0.25
Child Population (0–14) *	0–830	0.33
	831–1665	0.67		Service Activities	0.50
	1666+	1
Elderly Population (65+) *	0–480	0.33		Agricultural Labor	0.75
	481–960	0.67		Agricultural Production/ Livestock Activities	1
	961+	1
Number of Household Members *	<5	0.33
	5–10	0.67	Road Connection ▲	First Degree	0.33
	10+	1		Second Degree	0.67
				Third Degree	1
			Railway Connection ▲	Available	0
				Not Available	1

* TSI, 2023. ▲ Checked/measured using current Google Earth Maps and Street Views.

, social and economic vulnerability indicators were evaluated collectively for all disaster types. The scoring structures for social, economic, natural, and artificial vulnerability indicators are presented below (Table 6,

Table 7. Artificial and natural vulnerability scores for earthquakes.

Artificial Vulnerability			Natural Vulnerability
Indicators	Sub-Indicators	Weighting	Indicators	Sub-Indicators	Weighting
Settlement Area Type ▲	Coastal Area	1	Height Above Sea Level (m) ▼	0–50	1
	River/Lake Shore	0.75		51–100	0.8
	Other (Plain, etc.)	0.50		101–500	0.6
	Valley/Slope	0.25		501–1000	0.4
Settlement Form ▲	Linear	0.33		1001+	0.2
	Compact	0.67	Slope Condition (%) ▼	Low Slope (0–10)	0.25
	Dispersed	1		Slight Slope (11–30)	0.5
Building Block Form ▲	Grid	0.33		Mixed	0.75
	Mixed	0.67		Highly Sloped (≥31)	1
	Organic	1	Geological Formation Type ●	Alluvial Fan, Terrace, Unconsolidated Quaternary Deposits, Pliocene Quaternary Deposits	1
Structure Relationships ▲	Adjacent Order	1		İzmir Flysch (sandstone), Schist, Gneiss, Paleozoic Schist, Marble, Quartzite, Phyllite, Siltstone, Loras Formation (limestone), Kırkağaç Formation, etc.	0
	Mixed	0.67	Fault Line Nature ●	Active/Strike-slip/Normal/Vertical Fault	1
	Separate Order	0.33		Approximate Fault Location	0.5
Number of Floors ▲	Low-Rise (1–2)	0.33
	Mid-Rise (3–5)	0.67
	High-Rise (6+)	1
Drinking Water Infrastructure ●	Available	0	Proximity to Fault Line (m) ●	0–100	1
	Not Available	1		101–500	0.8
Sewage Infrastructure ●	Available	0		501–1000	0.6
	Not Available	1		1001–1500	0.4
				1501+	0.2

▲ Checked/measured using current Google Earth Maps and Street Views. ▼ Measured using GIS Applications. ● [218].

,

Table 8. Artificial and natural vulnerability scores for floods.

Artificial Vulnerability			Natural Vulnerability
Indicators	Sub-Indicators	Weighting	Indicators	Sub-Indicators	Weighting
Settlement Area Type ▲	Coastal Area	1	Slope Condition (%) ▼	Low Slope (0–10)	1
	River/Lake Shore	0.75		Slight Slope (11–30)	0.75
	Other (Plain, etc.)	0.5		Mixed	0.50
	Valley/Slope	0.25		Highly Sloped (≥31)	0.25
Settlement Form ▲	Compact	1	Proximity to the Shore (m) ▼	0–5000	1
	Linear	0.67		5001–10,000	0.8
	Dispersed	0.33		10,001–50,000	0.6
Building Block Form ▲	Organic	1		50,001–100,000	0.4
	Grid	0.67		100,001+	0.2
	Mixed	0.33	Proximity to Stream (m) ▼	0–50	1
Structure Relationships ▲	Adjacent Order	1		51–100	0.8
	Mixed	0.67		101–500	0.6
	Separate Order	0.33		501–1000	0.4
Number of Floors ▲	Low-Rise (1–2)	1		1001+	0.2
	Mid-Rise (3–5)	0.67	Soil Class ●	I and II Class	0.25
	High-Rise (6+)	0.33		III and IV Class	0.50
Drinking Water Infrastructure ●	Available	0		V-VI-VII Class	0.75
	Not Available	1		VIII Class	1
Sewage Infrastructure ●	Available	0
	Not Available	1

▲ Checked/measured using current Google Earth Maps and Street Views. ▼ Measured using GIS Applications. ● [218].

and

Table 9. Artificial and natural vulnerability scores for landslides.

Artificial Vulnerability			Natural Vulnerability
Indicators	Sub-Indicators	Weighting	Indicators	Sub-Indicators	Weighting
Settlement Area Type ▲	Sea/River/Lake Shore, Other (Plain, etc.)	0.5	Height Above Sea Level (m) ▲	0–100	0.2
				101–500	0.4
	Valley/Slope	1		501–750	0.6
				751–1000	0.8
Building Block Form ▲	Grid	1		1001+	1
	Organic	0.33	Slope Condition (%) ▲	Low Slope (0–10)	0.25
	Mixed	0.67		Slight Slope (11–30)	0.5
Drinking Water Infrastructure ●	Available	0		Mixed	0.75
	Not Available	1		Highly Sloped (≥31)	1
Sewage Infrastructure ●	Available	0	Soil Class ●	I and II Class	0.25
	Not Available	1		III and IV Class	0.5
				V-VI-VII Class	0.75
				VIII Class	1
			Natural Land Use ▲	Irrigated Agriculture	1
				Rainfed Agriculture
				Vineyard	0.67
				Garden
				Pasture
				Heathland
				Olive	0.33
				Forest

▲ Checked/measured using current Google Earth Maps and Street Views. ▼ Measured using GIS Applications. ● [218].

).

The Vulnerability Score by District shown in Figure 3 has been obtained by dividing the total score from the social, economic, artificial (total of earthquake, flood, and landslide), and natural (total of earthquake, flood, and landslide) vulnerability categories of the districts by the product of the number of neighborhoods in the district and the number of variables. The vulnerability scores at both neighborhood and district levels were digitized and subsequently analyzed using the ArcGIS 10.4 software. Detailed explanations regarding the data obtained at the neighborhood and district levels are provided below.

3. Vulnerability Analyses at Neighborhood Scale

This section evaluates each neighborhood’s social, economic, natural, and artificial vulnerability values according to the method provided under the relevant headings. Each neighborhood has been separately addressed for natural and artificial vulnerability according to three different types of disasters (earthquake, flood, landslide).

3.1. Social and Economic Vulnerability Analyses

Social and economic vulnerability scores, calculated based on identified indicators at the neighborhood scale, were categorized and visualized using the ARCGIS 10.4 software. Neighborhoods with populations below 100 were excluded from the calculations due to the unavailability of data from TSI [219].

The results of the social vulnerability analysis indicate that neighborhoods with high vulnerability scores exist in all districts except Beydağ. Similarly, economic vulnerability analysis results show high vulnerability scores in specific neighborhoods in every district. Regarding social vulnerability, Ödemiş (10 neighborhoods) and Tire (7 neighborhoods) districts have the highest number of highly vulnerable neighborhoods. In terms of economic vulnerability, Kiraz (39 neighborhoods), Tire (38 neighborhoods), and Ödemiş (33 neighborhoods) districts stand out (Figure 5 and Figure 6).

3.2. Vulnerability Analyses by Disasters at Neighborhood Scale

Based on identified indicators, vulnerability scores calculated for earthquakes, floods, and landslides at the neighborhood scale were categorized and visualized as artificial and natural vulnerabilities using the ARCGIS 10.4 software.

Artificial vulnerability analysis results for earthquakes indicate neighborhoods with high district vulnerability scores. The results show that Ödemiş (73 neighborhoods), Kiraz (50 neighborhoods), and Tire (33 neighborhoods) districts have the highest number of highly vulnerable neighborhoods in terms of artificial vulnerability. In contrast, natural vulnerability results indicate that Bayındır (32 neighborhoods) and Tire (30 neighborhoods) districts have the highest number of highly vulnerable neighborhoods (Figure 7 and Figure 8).

Flood artificial vulnerability analysis also reveals neighborhoods with high scores in all districts. Specifically, Ödemiş (69 neighborhoods) and Kiraz (38 neighborhoods) districts stand out regarding artificial vulnerability. Natural vulnerability analysis indicates Tire (43 neighborhoods) and Ödemiş (41 neighborhoods) districts have the most vulnerable neighborhoods (Figure 9 and Figure 10).

In terms of landslides, artificial vulnerability analysis again shows high vulnerability scores in neighborhoods across all districts. Ödemiş (45 neighborhoods), Kiraz (27 neighborhoods), and Tire (27 neighborhoods) districts have the highest number of highly vulnerable neighborhoods. Natural vulnerability analysis reveals that Kiraz (38 neighborhoods) and Ödemiş (31 neighborhoods) districts are the most affected (Figure 11 and Figure 12).

4. Social, Economic, Natural, and Artificial Vulnerability Levels by Districts

Social and economic vulnerability scores calculated at the district scale based on identified indicators (Table 6) and artificial and natural vulnerability scores calculated for earthquakes, floods, and landslides (Table 7, Table 8 and Table 9) were categorized and visualized using the ARCGIS 10.4 software. Due to the unavailability of data from TSI for neighborhoods with populations below 100, these neighborhoods were excluded from the calculations. The values calculated for social and economic vulnerability for the districts are presented in

Table 10. Social and economic vulnerability by districts.

District	Neighborhood Number	Social Vulnerability			Economic Vulnerability
		Score	Indicator Number	SC/(NN × IN)	Score	Indicator Number	SC/(NN × IN)
Selçuk	12	25.53	5	0.43	35.95	5	0.60
Bayındır	48	87.2	5	0.36	176.59	5	0.74
Tire	59	118.62	5	0.40	225.81	5	0.77
Beydağ	16	24.84	5	0.31	61.36	5	0.77
Kiraz	52	82.64	5	0.32	207.85	5	0.80
Ödemiş	68	149.63	5	0.44	246.98	5	0.73

and Figure 13, while the values calculated for artificial and natural vulnerability are presented in

Table 11. Artificial and natural vulnerability by districts.

District	Disaster Type	Neighborhood Number	Artificial Vulnerability			Natural Vulnerability
			Score	Indicator Number	SC/(NN × IN)	Score	Indicator Number	SC/(NN × IN)
Selçuk	Earthquake	14	48.3	7	0.49	43.35	5	0.62
	Flood	14	47.94	7	0.49	40.3	4	0.72
	Landslide	14	21.54	4	0.38	23.3	4	0.42
	Total	14	117.78	18	0.47	106.95	13	0.59
Bayındır	Earthquake	59	207.94	7	0.50	175.75	5	0.60
	Flood	59	211.56	7	0.51	151.2	4	0.64
	Landslide	59	97.24	4	0.41	104.85	4	0.44
	Total	59	516.74	18	0.49	431.8	13	0.56
Tire	Earthquake	88	314.13	7	0.51	256.05	5	0.58
	Flood	88	317.4	7	0.52	218.15	4	0.62
	Landslide	88	150.79	4	0.43	174.39	4	0.50
	Total	88	782.32	18	0.49	648.59	13	0.57
Beydağ	Earthquake	25	90.88	7	0.52	62.25	5	0.50
	Flood	25	89.2	7	0.51	55.65	4	0.56
	Landslide	25	46.61	4	0.47	50.16	4	0.50
	Total	25	226.69	18	0.50	168.06	13	0.52
Kiraz	Earthquake	56	228.78	7	0.58	130.3	5	0.47
	Flood	56	234.08	7	0.60	108.35	4	0.48
	Landslide	56	125.24	4	0.56	151	4	0.67
	Total	56	588.1	18	0.58	389.65	13	0.54
Ödemiş	Earthquake	99	390.01	7	0.56	289.5	5	0.58
	Flood	99	411.3	7	0.59	218.15	4	0.55
	Landslide	99	212.6	4	0.54	214.6	4	0.54
	Total	99	1013.91	18	0.57	722.25	13	0.56

and Figure 14.

As can be seen, the districts show differences in terms of social and economic vulnerability. Therefore, evaluating these differences within the framework of social and economic strategies that will be developed at different planning scales is necessary.

5. Evaluation and Conclusions

The analysis demonstrates that despite being part of the same geographic region (Southern Izmir) and watershed (Küçük Menderes Basin), neighboring districts and their rural neighborhoods exhibit different levels of vulnerability across social, economic, natural, and artificial dimensions, and for different disasters (earthquakes, floods, and landslides). These findings emphasize the necessity of addressing disaster-sensitive planning at the settlement level. While national and regional planning strategies are developed at higher scales, disaster-sensitive planning strategies need to be differentiated at the district and neighborhood levels.

This study confirms this need through its analysis of 341 neighborhoods—75% of which are rural—in Bayındır, Beydağ, Kiraz, Ödemiş, and Selçuk districts.

This study proposes a strategic map divided into three regions (Figure 15):

• Region 1: Selçuk district.
• Region 2: Tire and Bayındır districts.
• Region 3: Ödemiş, Kiraz, and Beydağ districts.

Specific findings include the following:

• Social vulnerability: Ödemiş district is the most vulnerable (0.44), while Beydağ district, located immediately to the southeast and neighboring it, is the least vulnerable (0.31).
• Economic vulnerability: Kiraz district, located in the far east of the study area, is the most vulnerable (0.80), while Selçuk district, located in the far west, is the least vulnerable (0.60).
• Artificial vulnerability: Kiraz district is the most vulnerable (0.58) in the far east of the study area, while Selçuk district is the least vulnerable (0.47) in the far west. In addition, Kiraz district is the most vulnerable to earthquakes (0.58), floods (0.60), and landslides (0.56).
• Natural vulnerability: Selçuk district, located in the far west of the study area, is the most vulnerable (0.59), while Beydağ district, located in the far east, is the least vulnerable (0.52). In addition, Selçuk district is the most vulnerable to earthquakes (0.62) and floods (0.72), while Kiraz district is the most vulnerable to landslides (0.67).
• It has been determined that Ödemiş district is the most vulnerable according to social vulnerability factors, Kiraz district is the most vulnerable according to economic and artificial vulnerability factors, and Selçuk district is the most vulnerable according to natural vulnerability factors.

As a result, the districts of Selçuk, Bayındır, Tire, Beydağ, Kiraz, and Ödemiş show differences in terms of social, economic, natural, and artificial vulnerabilities. A similar situation is valid for different rural settlements of the same district. Therefore, the differentiation of the results regarding the effect of different factors on vulnerability indicates that the spatial, social, and economic strategies to be developed must also differ. In other words, these results highlight the need for developing “Rural Area Design Strategies” in creating disaster-sensitive planning decisions and consequently identifying disaster-sensitive “Design Regions.” The table below shows the priority status of the districts according to their vulnerabilities as defined in the scope (

Table 12. Proposal for prioritization of districts by disaster types.

Social Vulnerability	Economic Vulnerability	Artificial Vulnerability	Natural Vulnerability	Type of Disaster	Priority
Ödemiş and Selçuk	Kiraz, Tire, and Beydağ	Kiraz and Ödemiş	Selçuk and Bayındır	Earthquake	First
		Kiraz and Ödemiş	Selçuk and Bayındır	Flood
		Kiraz and Ödemiş	Kiraz	Landslide
Bayındır and Tire	Bayındır	Beydağ and Tire	Tire and Ödemiş	Earthquake	Second
		Tire, Bayındır, and Beydağ	Tire	Flood
		Beydağ	Ödemiş	Landslide

).

• Although Bayındır and Selçuk are less vulnerable compared to other districts for all disasters, they should be addressed together.
• Tire district has an equal level of vulnerability to Ödemiş in terms of earthquakes and to Beydağ in terms of landslides, making it the second priority district to be addressed.

The results obtained from this study show similarities with research addressing disaster-sensitive strategies in two fundamental aspects. The first similarity is the variability of vulnerability levels at the rural settlement level, even within the same district, as observed in academic studies that focus on location-specific vulnerability criteria [220,221,222,223,224]. The second similarity is the identification of vulnerabilities, as seen in national and international vulnerability studies [225,226,227,228,229], which highlights the necessity of establishing strategic areas and developing strategic documents based on these vulnerabilities. In line with these studies, it emphasizes the integration of multi-source datasets, careful consideration of spatial resolution, and the incorporation of hazard-related environmental parameters as fundamental components for advancing disaster-sensitive planning and vulnerability research.

As a result, the differentiation of vulnerability levels in terms of social, economic, natural, and artificial characteristics for settlements located in the same geographical area indicates the need to define areas where different planning strategies will be developed for these settlements. Moreover, vulnerability indicators for each disaster are part of an evolving process that develops alongside studies from various countries. This study aims to contribute to this process by providing examples of settlements from Türkiye’s Aegean Region regarding vulnerability indicators and their availability.

The comparative assessment of different vulnerability analysis methods highlights the rationale for adopting the equal-weighted min–max normalization approach in this study. While more sophisticated techniques such as the Analytic Hierarchy Process (AHP) and Principal Component Analysis (PCA) offer specific strengths—namely the incorporation of expert judgment (AHP) and the reduction in dimensionality through data-driven correlations (PCA)—both methods also present certain drawbacks. AHP is prone to subjectivity and consistency issues, whereas PCA, despite its statistical robustness, often yields results that are difficult to interpret for policy-making purposes.

In contrast, the equal-weighted normalization method employed here provides a transparent and straightforward framework, ensuring that all indicators contribute equally and enabling comparability at the neighborhood scale despite data limitations. This choice not only aligns with common practices in the relevant literature but also enhances the interpretability of the findings for both academic and policy audiences. Ultimately, the approach adopted in this study balances methodological rigor with practical applicability, thereby contributing to advancing disaster-sensitive planning and vulnerability research.

In the literature, numerous other indicators can be employed to assess vulnerability at different scales (national, regional, urban, and neighborhood levels). However, this study was unable to incorporate all of these indicators, which may be considered a limitation of the research. Nevertheless, the inability to access data for all indicators referenced in the literature does not stem from the authors. The primary reason for this limitation lies in the fact that institutional datasets in the country are not sufficiently detailed at the scale of rural settlements, which constitute the focus of this study. Therefore, beyond making a significant contribution to disaster-sensitive planning efforts in rural areas of Türkiye, this study also highlights a critical gap in the availability of datasets at this scale.