ORIGINAL RESEARCH article

Front. Earth Sci., 05 June 2026

Sec. Geohazards and Georisks

Volume 14 - 2026 | https://doi.org/10.3389/feart.2026.1837277

GIS-based spatial multi-hazard vulnerability assessment of cultural heritage sites: the case of Antakya

  • Department of City and Regional Planning, Faculty of Architecture, Iskenderun Technical University, İskenderun, Türkiye

Abstract

Cultural heritage sites possess outstanding universal value, yet they remain fragile and highly vulnerable to both natural and anthropogenic hazards. These hazards pose significant risks to the sustainability of heritage, predominantly in developing countries. This study offers a comprehensive assessment framework by integrating multi-hazard risk analysis of cultural heritage sites with remote sensing techniques and geographic information systems (GIS). Antakya, a city with a multi-layered cultural heritage molded by numerous civilizations, suffered severe damage to its heritage assets following the 6 February 2023, Kahramanmaraş earthquake. Against this backdrop, the present research analyzes cultural heritage areas and assets in Antakya in relation to multiple hazards, with the aim of identifying vulnerable sites. Seven hazards were considered: erosion, landslides, earthquakes, floods, fires, urban sprawl, and modern road networks. These hazards were mapped in a GIS environment using remote sensing methods, satellite imagery, and diverse datasets. Hazard weights were determined through the Analytic Hierarchy Process (AHP), enabling the creation of risk maps. The findings reveal that 75.85% of the 141 cultural heritage sites and 89.38% of the 979 cultural heritage assets in Antakya are at high or very high risk. These findings highlight the critical need for a comprehensive risk management plan to protect cultural heritage and strengthen resilience in Antakya.

1 Introduction

UNESCO has defined cultural heritage sites as sites of outstanding universal value from a historical, artistic, or scientific perspective (). Heritage sites are worthy of protection not only as physical entities but also for the significance and value they hold for society (). These sites make significant contributions to countries’ sustainable development, the transmission of historical and cultural heritage, and the improvement of urban quality of life through tourism (; ; ; ). However, heritage sites face the threat of extinction due to exposure to natural hazards (landslides, floods, erosion, earthquakes, sea-level rise/tsunamis, avalanches, hurricanes, desertification, severe weather events, etc.), anthropogenic factors, and environmental changes (urban sprawl, climate change, distance from road networks, wildfires, war and armed conflict, terrorism, tourism pressure, air pollution, etc.) and face the threat of extinction (; ; ; ; ; ; ; ; ; ). In recent years, particularly, many cultural heritage sites around the world have been damaged by multiple disasters, resulting in significant losses. Consequently, the need for comprehensive risk assessment approaches for these sites is increasing (; ).

In recent academic studies, risk assessment for multiple hazards in cultural heritage sites has become increasingly important. For example, analyzed management plans for UNESCO World Heritage Sites and demonstrated that these sites are frequently exposed to multiple hazards, including fire, flood, earthquakes, and storms. Similarly, developed a risk index that combines hazard levels with potential damage to assess multi-hazard risk in World Heritage sites across Europe.

Remote sensing and geographic information systems (GIS) are important tools for protecting cultural heritage sites against multiple hazards and risks, mitigating their impacts, and developing conservation policies and models. In particular, GIS-based models are effectively used in hazard susceptibility analysis, risk assessment, scenario development, monitoring environmental changes, and decision-support processes (; ; ; ). However, the vast majority of existing studies focus on single hazards and fall short of addressing multiple hazards within a holistic framework (; ; ; ; ). This situation creates a significant research gap, particularly in developing countries where the impacts of disasters are felt more acutely (). In this context, this study aims to develop a comprehensive hazard-vulnerability analysis model for cultural heritage sites in the city of Antakya by integrating GIS and the Analytic Hierarchy Process (AHP). The study aims to provide a comprehensive approach to identifying and prioritizing multiple hazards in cultural heritage sites.

2 Materials and methods

2.1 Study area

Antakya, the central district of Hatay Province in southern Türkiye’s Eastern Mediterranean Region (; ) is located between 36°12′–36°14′N latitude and 36°9′–36°12′E longitude. It covers an area of 585.10 km2, lies approximately 22 km inland from the coast, and is situated at an elevation of about 80 m above sea level (Figure 1). The city is positioned between Mount Habib-i Neccar and the Asi River, which flows through its center (; ). The area between the Asi River and Mount Habib-i Neccar is referred to as “Old Antakya,” whereas the western side of the river constitutes “New Antakya” ().

FIGURE 1

Although archaeological remains from the Paleolithic, Epipaleolithic, Neolithic, Chalcolithic, and Bronze Ages have been found in the city, the settlement established by the Seleucid Empire in 300 BC during the Hellenistic period marked the city’s attainment of its initial urban identity as “Antiokheia” (Antakya) (; ). Antiokheia subsequently became the third-largest city in the Roman Empire after Rome and Alexandria (; ). Owing to its strategic location and its position along the Silk and Spice Routes, the city emerged as a major trading hub throughout history. As a result, Antakya has been home to numerous civilizations from antiquity to the present, including the Roman, Byzantine, Arab-Islamic, Seljuk, Mamluk, Ottoman, and the modern Turkish Republic (; ; ). This economic, social, and cultural diversity endowed the city with a multicultural character, diversified its spatial uses, and contributed to the emergence of rich, diverse cultural heritage landscapes.

Different faith systems have coexisted in Antakya. For instance, St. Pierre Church, listed on UNESCO’s Tentative List, is regarded as the world’s first cave church. It is also where St. Pierre, the first bishop of the early Christian community, delivered his first sermon. In 1963, the Pope declared it a Christian pilgrimage site (). Similarly, the Habib-i Neccar Mosque is significant to Muslims as the first mosque in Anatolia (; ). In addition to these sites, the Antakya Jewish Synagogue, St. Paul Orthodox Church, Antakya Protestant Church, Antakya Catholic Church, Ulu Mosque, Sarımiye Mosque, and Affan Mosque are also located in the Historic City Center ().

There are 979 immovable cultural assets in Antakya. These assets include protected streets, monuments and memorials, administrative buildings, cultural facilities, martyrdom sites, military structures, industrial and commercial buildings, religious buildings, cemeteries, examples of civil architecture, and ruins. The city contains 1 urban conservation area, 1 urban buffer zone (urban site interaction transition zone) 3 natural sites, 1 buffer zone, 1 site area, 13 conservation areas, two proposed conservation areas, and 119 archaeological sites (). The Antakya Historic City Center is located within an urban conservation area, an urban buffer zone, and 1st-, 2nd-, and 3rd-degree archaeological sites, covering approximately 30.9 km2. Following the earthquake, the number of registered cultural assets in the Historic City Center increased from 613 to 957 (; ). Notably, 98% of the city’s cultural heritage assets are concentrated in the Historic City Center. Thus, the Antakya Historic City Center represents a multi-layered cultural heritage area, encompassing civil architectural structures, traditional residential fabric, the historic commercial core, monumental buildings, and archaeological cultural assets ().

However, the city of Antakya lies along the East Anatolian Fault, one of Türkiye’s largest fault lines. On 6 February 2023, earthquakes measuring 7.7 (Mw) and 7.6 (Mw) struck the districts of Pazarcık and Elbistan in Kahramanmaraş. Antakya suffered serious damage in these earthquakes. The earthquakes were felt severely in a total of eleven cities, particularly Kahramanmaraş and Hatay, causing serious damage. As a result, 53,537 people lost their lives, and 313,000 buildings were destroyed. 15.6 million people in the region were affected (; ; ). It has also been stated that the cost of the earthquakes to the Turkish economy was approximately $103.6 billion (). There are 8,444 historical structures in the region across 11 cities, of which 3,752 were damaged. The restoration cost is estimated to exceed 2 billion US dollars (). Following the earthquakes, 37.54% of the 613 cultural assets in Antakya’s Historic City Center were destroyed. 35.41% were severely damaged, and 2.73% required demolition. Furthermore, approximately 75.5% of the monumental structures, religious buildings, commercial areas, historic bazaar district, and examples of civil architecture that constitute the city’s identity were seriously damaged in the earthquake (). As a result, these structures that form the city’s identity are being rebuilt through reconstruction and restoration efforts (Figures 2–4).

FIGURE 2

).

FIGURE 3

FIGURE 4

2.2 Data collection

The study sources are listed in Table 1. KLM data provided by the Hatay Cultural Heritage Preservation Regional Board was used to determine the locations of cultural heritage areas and cultural assets. Slope, aspect, elevation, and curvature maps were derived from ALOS PALSAR 12.5-m-resolution satellite imagery using a DEM. The soil map was obtained from the Hatay Provincial Directorate of Agriculture. Active fault lines and geological data were provided by the General Directorate of Mineral Research and Exploration, while earthquake data were obtained in CVS file format from the United States Geological Survey (USGS) website. Land use/Land cover (LULC) data was created from CORINE 2018, Google Earth, and the 1/5,000 Scale Land Use Plan. NDVI and TWI were created from 2014 Landsat 9 OLI/TIRS data at 30 m resolution. Finally, Sentinel-2A’s 20-m-resolution images from 2015 to 2025 were used to map urban sprawl.

TABLE 1

Data nameData typeResolution (m) or scale barSource
Cultural heritage sites and assetsKML (vektor)1:1,000
Earthquake magnitudeCVS (raster)-
Active fault line
Geology		Vektor (shp)1:100,000
Land use/Land cover (LULC)
Road		Vektor (shp)100 m
1:5,000		; ; 1/5,000 scaled land use plan
Slope, aspect
Elevation, curvature		Raster (tiff)12.5, ALOS PALSAR Satellite data (DEM)
SoilVektor (shp)1:100,000
NDVI - TWIRaster (tiff)30 m. Landsat 9 OLI/TIRS (2021) Band 4 (red) and Band 5 (NIR)
PrecipitationTable (xlsl)-
Urban sprawlRaster (tiff)20 m
Burned area (dNBR)Raster (tiff)10 m

Data used in the study.

2.3 Methodology

This study covers risks posed by events that directly generate hazards, as well as anthropogenic factors that amplify their impact. The potential hazards in the research area affecting cultural heritage sites were examined, and a risk analysis was conducted. The study identified the hazards posing the highest risk. The city of Antakya is located on the Amik Plain, an area characterized by agricultural features, and is influenced by the alluvial soil structure and the presence of the Asi River, which originates in Lebanon and flows through the Antakya and Samandağ districts of Hatay in Türkiye, its location between the Amanos Mountains and the Habib-i Neccar Mountains, and the Mediterranean climate conditions (hot and dry summers, mild and rainy winters), making it vulnerable to various natural risks. The average annual precipitation is 1,200 mm. In terms of earthquake risk, the region is located in a first-degree earthquake zone due to the intersection of the Dead Sea Fault, the Karasu Fault, and the Cyprus-Antakya Fault. The presence of alluvial deposits in the soil increases the risk of soil liquefaction; the area’s geologically young sedimentary units, on the other hand, increase the risk of flooding. The flow of the Asi River within a graben and the presence of steep topography also increase the risks of flooding and landslides. Additionally, the fact that approximately 27% of Antakya is forested increases the risk of wildfires (; ). Accordingly, a risk assessment was conducted for cultural heritage sites in the Antakya district of Hatay province, focusing on natural hazards (floods, erosion, earthquakes, landslides) and anthropogenic hazards (fires, proximity to modern road networks, urban sprawl). In this assessment, 7 risk factors were identified based on natural and anthropogenic hazards. A risk map was created for each risk factor. The flow diagram of the research method is shown in Figure 5.

FIGURE 5

2.3.1 The AHP method

In this study, the AHP method was used to identify floods, landslide, and multiple risks in the city of Antakya. The Analytic Hierarchy Process (AHP) is a multi-criteria decision-making method developed by Saaty (). In this method, decision criteria are determined hierarchically, and alternatives are ranked according to their importance levels. Ranking based on expert opinions increases the method’s accuracy. AHP is widely used in risk assessments of cultural heritage sites (; ; ; ; ). This method helps to form risk classes by evaluating natural and anthropogenic factors within a hierarchical structure.

The mathematical formulation of the AHP method is given below Equations 1–6. A pairwise comparison matrix is created to determine the relative importance of each criterion (Table 2). Values between 1 and 9 are assigned to the criteria in this comparison matrix based on their importance. These values and importance levels are defined as follows: 1 = equally important; 3 = moderately important relative to the others; 5 = fundamental or strongly important; 7 = very strongly important; 9 = extremely important; and 2, 4, 6, 8 = intermediate values. The matrix is normalized by dividing each element by its column total. Then, row averages are taken to create criterion weights. Finally, the maximum (λmax), consistency index (CI), and consistency ratio (CR) are calculated to test the accuracy of the comparison matrix. In the formula, n is the number of parameters, and λmax is the largest eigenvalue of the matrix of order n. The Consistency Ratio (CR) is used to check for and therefore avoid possible inconsistencies in the judgment matrix, and must be below 0.1 for the weighting coefficients to be suitable (; ). Finally, CR is calculated. The random index (RI) in this equation is chosen according to the matrix size, i.e., the value of n in Table 3.

TABLE 2

K1K2K3…Kn
K1a11a12a13…a1n
K2a21a22a23…a2n
K3a31a32a33…a3n
……………
Knan1an2an3…ann

K pair-wise comparison matrix.

TABLE 3

n123456789101112131415
RI000.580.91.121.241.321.411.451.491.511.481.561.571.59

RI values ().

A total of three experts from different disciplines were consulted to determine the criterion weights in the AHP analysis. The experts comprised three academics with experience in disaster management, urban geography, geology, and cultural heritage. The experts were asked to evaluate the identified hazard factors using the pairwise comparison method. The resulting comparison matrices were combined to calculate the final weight values. Saaty’s 1–9 scale was used to create the pairwise comparison matrix. The relative importance of each criterion was determined based on the experts’ evaluations, and the weight values were obtained by normalizing the comparison matrix. The consistency ratio (CR) was found to be less than 0.1 for all comparisons, and the matrices were deemed consistent.

2.3.2 RUSLE method

The RUSLE method used to determine soil erosion risk in the study area is an extension of the Universal Soil Loss Equation (USLE) and integrates spatial methods (; ; ). The RUSLE calculation determines erosion in tons/hectare/year (A) by multiplying the rainfall erosion (R), soil erosion (K), length-slope (LS), cover management (C), and protection practices (P) factors () (Equation 7). Fieldwork and interviews with institutions confirmed that no erosion control actions were taken. Since there were no erosion-preventing conservation practices in the study area, the P factor was taken as 1, as indicated by .

2.3.3 NBR index and NDVI

In this study, the Normalized Burn Ratio (NBR) index was used for fire risk analysis, with the Google Earth Engine (GEE) platform for processing remote sensing data and performing spatial analyses. The NBR index is a spectral indicator widely used in the literature to assess fire effects. This index enables the determination of burn severity in post-fire areas using near-infrared (NIR) and short-wave infrared (SWIR) bands in satellite imagery. The calculation is performed by evaluating the difference between the reflectance values obtained from pre- and post-fire satellite data (Equations 8, 9). When post-fire vegetation cover is damaged, reflectance in the NIR bands increases while values in the SWIR bands decrease; this change allows the Normalized Burn Ratio (NBR) index to quantitatively reveal the effects of fire (; ; ; ; ). The NBR index is calculated using the near-infrared (NIR) and shortwave infrared (SWIR) bands:

Low NBR values indicate post-fire areas, while high values signal healthy vegetation cover. To assess fire intensity in greater detail, use the Delta NBR (dNBR) index, calculated as follows:

dNBR enables the determination of burn severity by classifying fire effects into five categories (unburned, low, low-medium, medium-high, high) ():

The study also calculated the NDVI index using Sentinel-2 data to assess vegetation change. NDVI reveals pre- and post-fire biomass changes and spatial and ecological differences in vegetation cover (). The mathematical formula of NDVI is given below (Equations 10, 11).

dNDVI is calculated using pre- and post-fire NDVI differences, and burn severity is classified into six categories: unburned, very low, low, moderate, high, and very high ().

2.3.4 Risk analysis

For the multiple hazard risk map of cultural heritage sites in Antakya district, Hatay province, the region’s geographic data were processed and reclassified in ArcGIS Pro. This resulted in the creation of raster maps with risk values assigned to each pixel. The risk factors and their weights are presented in Table 4.

TABLE 4

HazardsFactorsSub-classScaleWeightQualitative scaleScaleWeights
FloodSlope (°)<1050.1444Very high50.226
10–204High4
20–303Medium3
30–402Low2
>401Very low1
AspectSouth50.025
Southeast, southwest4
West3
Northeast, east, northwest2
Flat, north1
Elevation (m)<40050.154
400–8004
800–1,2003
1,200–1,6002
>1,6001
Precipitation (mm)>1,55050.251
13,501,5504
1,150–1,3503
950–1,1502
<9501
SoilAlluvial50.026
Red-brown mediterranean soils4
Colluvial soils4
Forest soils3
Brown forest soils2
Red mediterranean soils1
Distance from streams (m)<10050.166
100–2004
200–3003
300–4002
>4001
LULCUrban areas50.068
Water surfaces5
Mining areas4
Blank area4
Agricultural areas3
Other2
Forest areas1
NDVI−0.085–0.09150.064
0.092–0.1464
0.147–0.2073
0.208–0.2792
0.280–0.6531
CurvatureConcave50.028
Convex3
Flat1
TWI<610.074
6–112
11–153
15–23.54
>23.55
LandslideSlope (°)<1010.281Very high50.088
10–202High4
20–303Medium3
30–404Low2
>405Very low1
AspectNorth50.034
Northeast, northwest4
East, west3
Southeast, south, southwest2
CurvatureFlat10.038
Convex3
Concave5
Distance from roads (m)<5050.024
50–1004
100–1503
150–2002
>2001
LULCUrban areas50.142
Mining areas5
Blank area4
Agricultural areas3
Other2
Forest areas1
GeologyQuaternary, alluvium, continental, sedimentary rock (Q-21-k)50.196
Pliocene, sandstone–mudstone–limestone, shelf, sedimentary rock (PL-20-S)4
Tortonian, sandstone–mudstone–limestone, shelf, sedimentary rock (me-20-s)4
Langhian–Serravallian, sandstone–mudstone–limestone, shelf, sedimentary rock (mcmd-20-s)4
Langhian–Serravallian, limestone, shelf, sedimentary rock (mcmd-8-s)3
Ypresian–Lutetian, limestone, shelf, sedimentary rock (eaeb-8-s)2
Maastrichtian, gabbro, ophiolitic rock (WMz-km)1
Distance from faults (m)<10050.062
100–2004
200–3003
300–4002
>4001
Distance from streams (m)<5050.053
50–1004
100–1503
150–2002
>2001
Precipitation (mm)>1,55050.121
1,350–1,5504
1,150–1,3503
950–1,1502
<9501
Elevation (m)<40010.049
400–8002
800–1,2003
1,200–1,6004
>1,6005
EarthquakeEarthquake magnitude (Mw)4.65–4.8750.50Very high50.363
4.44–4.654High4
4.22–4.443Medium3
>4.222Low2
Distance from the active fault lines (m)>5,00050.50
5,000–100004
10,000–150003
15,000–200002
ErosionVery high50.039
High4
Medium3
Low2
Very low1
FireVery high50.090
High4
Medium3
Low2
Very low1
Proximity to modern road networks (m)Main road10.039
The other0
Urban sprawlExpansion area10.155

Multi-hazard risk assessment.

Land use/land cover (LULC) data were obtained from CORINE 2018, Google Earth, and a 1/5000-scale Land Use Plan. Additionally, the land use map was generated from CORINE data. Values were assigned to each pixel. According to the value assignment process, urban areas were assigned a value of 5, mining areas were assigned a value of 4 for flood risk and 5 for landslide risk, vacant areas were assigned a value of 4, agricultural areas were assigned a value of 3, other areas were assigned a value of 2, forest areas were assigned a value of 1, water surfaces were assigned a value of 5 for flood risk and 1 for landslide risk (Table 4). As a result, the LULC map was used to create floods, landslide, and erosion risk maps.

The geological map was obtained from the General Directorate of Mineral Research and Exploration (MTA), and the soil map from the Provincial Directorate of Agriculture. Soil and geological data were compiled from 1:100,000 scale vector data for Hatay province and converted to raster format for analysis. There are six different soil types in the study area: calcareous brown forest soils, brown forest soils, colluvial soils, red Mediterranean soils, red-brown Mediterranean soils, and alluvial soils. Geologically, the Quaternary, Pliocene, Langhian, Ypresian, and Maastrichtian units are present; these units exhibit different lithological characteristics, including alluvium, sandstone, mudstone, limestone, claystone, shelf sedimentary rock, and gabbro. Soil and geological units were classified into five classes according to landslide and flood risk and scored as very high (5), high (4), medium (3), low (2), and very low (1) (Table 4).

The flood risk map was created using rainfall, slope, aspect, soil, land use (LULC), elevation, distance from the river, NDVI, curvature, and TWI parameters. DEM data were derived from a 12.5 m resolution ALOS PALSAR image, while the NDVI and TWI layers were derived from Landsat 9 OLI/TIRS (Band 4–Red, Band 5–NIR) data. In contrast, the landslide risk map utilized rainfall, DEM, geology, LULC, distance to roads, distance to faults, and distance to rivers as parameters. For both maps, each factor was subdivided into subclasses and graded; factor weights were determined using the AHP method based on expert opinion. Consequently, flood and landslide risk levels were scored as very high (5), high (4), medium (3), low (2), and very low (1).

Rainfall, soil, DEM, and LULC data were used for erosion risk. These data were transferred to the GIS environment and overlaid. The earthquake risk map was created based on earthquake magnitude and distance from active faults. Earthquake data for the study area were obtained in CVS file format and imported into the ArcGIS Pro software as points. Then, an inverse-distance-weighting interpolation was applied within the study area boundaries to produce an earthquake magnitude map. The raster map created was reclassified and scored. For the fault proximity map, active fault line data from the General Directorate of Mineral Research and Exploration were used in vector format, and a buffer analysis was performed to determine proximity to the fault line. The raster map obtained was classified and scored. In the final stage, the earthquake magnitude and distance-to-active-fault-line maps were weighted to produce an earthquake risk map. Earthquake risk levels were scored as very high (5), high (4), medium (3), and low (2).

The urban sprawl in the Antakya district was analyzed by examining 15 years of land-use changes, with the urban macroform taken into account. Sentinel 2 satellite images from 2015 to 2025 were used with the MLC method. Accordingly, a value of 1 was assigned to the expansion area and 0 to the others.

Antakya’s modern road network data were obtained from the 1/5,000 Master Plan, Google Earth, and OpenStreetMap. A buffer zone analysis was applied to the proximity of heritage areas to the road network. Accordingly, areas within a 250 m radius were assigned a value of 1, while other areas were assigned a value of 0.

To spatially assess the effects of the fire, Sentinel-2 Level-2A surface reflectance data for the Antakya district were used. The GEE platform enabled image analysis. Images from the pre-fire period (2016–2018) and post-fire period (2023–2025) were selected. Seasonal filtering (May–September) and advanced cloud masking methods were applied. In the next step, the Normalized Burn Ratio (NBR) was calculated to determine burned areas. Burn severity was then classified using the difference NBR (dNBR). Classification followed the USGS standard and included the following levels: unburned (<0.10), low (0.10–0.26), medium-low (0.27–0.43), medium-high (0.44–0.65), and high (>0.66). In addition, the difference in the Normalized Difference Vegetation Index (NDVI) (dNDVI) was calculated to assess vegetation loss. Fire severity was classified into six levels: unburned (<0.07), very low (0.08–0.13), low (0.13–0.20), medium (0.20–0.33), high (0.33–0.44), and very high (>0.45). Thus, the effects of fires on vegetation cover and the spatial distribution of burned areas were analyzed and visualized using GIS-based mapping. Finally, fire risk was classified into five categories and scored as very high (5), high (4), medium (3), low (2), and very low (1).

Finally, rainfall data for the last 30 years for the district of Antakya were obtained from the Hatay Meteorological Station. These data were analyzed using the 30-year monthly total rainfall amounts for Antakya. Subsequently, long-term average rainfall values were calculated for each month. These calculated data were transferred to the GIS environment as point rainfall values from meteorological stations. In the next step, the Inverse Distance Weighting (IDW) interpolation method was applied to the precipitation spatial distribution. As a result, the 30-year average annual precipitation distribution was divided into five classes and scored as very high (5), high (4), medium (3), low (2), and very low (1) (Table 4).

2.3.5 Risk evaluation

In the study, the AHP method was used to create a decision matrix for natural and anthropogenic factors, based on expert opinion. This method was used to calculate each factor’s weight. First, a 10 × 10 decision matrix was created for flood and landslide risks among natural hazards (Tables 5, 6). Then, to assess the risks of multiple hazards, a 7 × 7 pairwise comparison matrix was prepared for urban sprawl, fire, distance from modern road networks, earthquake, flood, landslide, and erosion risks, and the weights of these risks were calculated (Tables 7, 8). As a result of these analyses, earthquakes were assessed as the most significant risk, given the destructive impact of the February 6 Kahramanmaraş earthquake on cultural assets and heritage sites. After the earthquake, floods and urban sprawl came to the fore in terms of weight. In particular, the opening of the archaeological site for development and the need for housing in this area in the Kahramanmaraş earthquake show that urban sprawl is a significant threat. Finally, the consistency test showed that CR passed with a value of 0.0652.

TABLE 5

SLPASPELVPRSOLDFSLULCNDVICURTWIWeight
SLP1711/35123540.144
ASP1/711/51/51/31/51/31/311/30.025
ELV1511/27133540.154
PR35217245750.251
SOL1/531/71/711/71/51/51/31/50.026
DFS1511/27145630.166
LULC1/231/31/451/41151/20.068
NDVI1/331/31/551/51151/20.064
CUR1/511/51/731/61/51/511/40.028
TWI1/431/41/551/322410.074
CR = 0.0819

Determining the weights of flood risk factors using the AHP method.

SLP, Slope; ASP, aspect; ELV, elevation; PR, precipitation; SOL, soil; DFS, distance from stream; LULC, land use/Land cover; NDVI, normalized difference vegetation index; CUR, curvature; TWI, topographic wetness index.

TABLE 6

SLPASPCURDFRLULCGEODFFDFSPRELVWeight
SLP17783255360.281
ASP1/711/341/51/71/31/21/510.035
CUR1/73121/41/51/31/31/31/20.038
DFR1/81/41/211/51/71/41/21/510.024
LULC1/35451133140.142
GEO1/27571145340.196
DFF1/53341/31/4111/21/20.062
DFS1/52321/31/5111/410.053
PR1/353511/324130.121
ELV1/61211/41/4211/310.049
CR = 0.062

Determining the weights of landslide risk factors using the AHP method.

SLP, Slope; ASP, aspect; CUR, curvature; DFR, distance from roads; LULC, land use/Land cover; GEO, geology; DFF, distance from active faults; DFS, distance from stream; PR, precipitation; ELV, elevation.

TABLE 7

(n)12345678910
RI0.000.000.580.901.121.241.321.411.451.49

RI of the comparison matrix.

TABLE 8

Urban sprawlFireRoadEarthquakeFloodLandslideErosionWeight
Urban sprawl1331/31/3340.155
Fire1/3131/51/3220.090
Road1/31/311/71/51/310.039
Earthquake35713370.363
Floods3351/31350.226
Landslide1/31/231/31/3130.088
Erosion1/41/211/71/51/310.039
CR = 0.0652.

Determining the weights of multiple hazard risk factors using the AHP method.

3 Results

3.1 Natural and anthropogenic factors

3.1.1 Erosion

Soil erosion is adversely affected by climate change, agriculture, overgrazing, urban sprawl, and land-use changes. It is considered one of the major threats causing damage to cultural heritage sites, especially in recent times (). In light of these concerns, the RUSLE erosion risk prediction model was used in this study to determine erosion risk. This model was preferred for its compatibility with GIS-based analyses and reliability. Through these analyses, it was revealed that the high- and very-high-risk erosion area amounts to approximately 4.21 km2, corresponding to 0.72% of the study area. Furthermore, Table 9 depicts that there are 10 cultural heritage areas in regions with medium erosion risk (Figure 6). This represents 7.09% of the total heritage areas. As a result, 92.94% of heritage areas are located in low-erosion-risk zones. Additionally, 977 cultural assets (99.80%) in the study area are under very low erosion risk.

TABLE 9

Area (km2)Percentage of area (%)Number of cultural heritage sitesPercentage (%)
Very high risk0.840.1400
High risk3.370.5800
Moderate risk14.562.49107.09
Low risk57.419.813625.56
Extremely low risk508.9286.989567.38
amount to585.10100141100

Erosion risk levels in cultural heritage sites.

FIGURE 6

3.1.2 Landslide

Cultural heritage sites are particularly vulnerable to landslides due to their fragile structures. This situation poses a potential threat to the sustainable protection of cultural heritage (). Ten factors influence landslide risk in the study. When assessed for landslide risk, very high- and high-risk areas constitute 21.13% of the total area (Table 10, Figure 7). There are 19 cultural heritage sites in these areas. In contrast, there are 50 cultural heritage sites in medium-risk areas. On the other hand, 876 cultural assets are in the very low and low-risk groups, 55 are in the medium-risk group, and 48 are in the high and very high-risk landslide groups.

TABLE 10

Area (km2)Percentage of area (%)Number of cultural heritage sitesPercentage (%)
Very high risk53.079.0764.26
High risk70.5612.06139.22
Moderate risk256.3843.825035.46
Low risk122.9921.024431.21
Very low risk82.1014.032819.85
amount to585.10100141100

Landslide risk levels in cultural heritage sites.

FIGURE 7

3.1.3 Earthquake

Earthquakes, among the most destructive natural threats, have destroyed many cultural assets and disrupted urban identity from the past to the present (). The city of Antakya is located in a first-degree earthquake zone, where the Dead Sea Fault, Karasu Fault, and Cyprus-Antakya Fault intersect. Therefore, the city has experienced many severe earthquakes throughout its history (). Most recently, the 6 February 2023, Kahramanmaraş earthquake caused serious structural damage and loss of life in the city. In the city, 30,176 buildings were either destroyed or severely damaged, and 75.5% of cultural assets were damaged (; ). Accordingly, earthquake hazard maps were prepared based on tremor magnitude and distance to the fault. Earthquakes in the study area are of medium magnitude and are concentrated in the northern and eastern parts. Accordingly, areas with high and very high earthquake magnitudes cover 56.41% of the study area, and the cultural heritage sites located in these areas account for 94.33% of the total heritage sites (Table 11, Figure 8). In the area where proximity to the active fault line poses a medium risk, 5 cultural heritage sites are under threat. As a result, more than 95% of heritage sites are at risk. Of the cultural assets, 977 (99.80%) are at high or very high earthquake risk.

TABLE 11

Area (km2)Percentage of area (%)Number of cultural heritage sitesPercentage (%)
Very high risk169.6028.994531.91
High risk160.4627.428862.41
Moderate risk165.3928.2753.55
Low risk89.6515.3242.13
amount to585.10100141100

Earthquake risk levels in cultural heritage sites.

FIGURE 8

3.1.4 Floods

Floods, one of the most common natural disasters worldwide, cause devastating damage in cities, significant financial losses, and damage to cultural heritage assets (; ). This problem is also clearly evident in Antakya. In particular, the city of Antakya is located on both banks of the Asi River and has therefore been frequently affected by floods and inundations throughout history (). For this reason, a risk map was produced using 10 factors to assess flood risk in the study. According to this map, the area with very high and high flood risk is 312.74 km2, corresponding to 53.45% of the total area (Table 12, Figure 9). Furthermore, according to Table 12, 77 cultural heritage sites are located in areas with very high and high flood risk, accounting for 54.61% of all heritage sites. Finally, 787 cultural assets are under high or very high flood risk.

TABLE 12

Area (km2)Percentage of area (%)Number of cultural heritage sitesPercentage (%)
Very high risk86.5414.793826.95
High risk226.2038.663927.66
Moderate risk138.7823.723927.66
Low risk85.7714.662316.31
Extremely low risk47.818.1721.42
amount to585.10100141100

Flood risk levels in cultural heritage sites.

FIGURE 9

3.1.5 Urban sprawl

Rapid urbanization presents urgent challenges, as more than half of the world’s population now lives in cities. Uncontrolled sprawl, in particular, is irreversibly damaging heritage sites, erasing the character of landscapes, and disconnecting areas from their contexts (; ). The city of Antakya was located on the slopes of Habib-i Neccar Mountain and along the banks of the Asi River prior to the 20th century. Between 1932 and 1967, urban development progressed linearly toward the north-northwest. After 1967, with the establishment of the Antakya Small Industrial Site, urban expansion along the northern axis accelerated, leading to urban sprawl. Between 1968 and 1985, due to rapid urbanization and migration, the city expanded northwestward toward the Amik Plain. Between 1985 and 2014, especially after 2000, with the expansion of public use areas in the northwest of the city, the city experienced significant growth in the north and northwest, and limited growth in the southeast. Between 2014 and 2023, urban sprawl increased considerably in the north and west, while development continued in the north, east, and south. With Hatay becoming a metropolitan city in 2012, rural settlements were transformed into urban neighborhoods. This accelerated urban sprawl. Furthermore, the 6 February 2023, Kahramanmaraş earthquake caused significant destruction in the city. The area most severely affected by damage and destruction after the earthquake was the Antakya Historic City Center. Additionally, there was significant destruction in urban development areas established after 2000 (; ). In Antakya, 93,247 people live in ninety-two temporary shelter areas (). To address housing needs, the Ministry of Environment, Urbanization, and Climate Change has planned a residential development area in the city’s northwest. However, this area is located on a third-degree archaeological site, and the projects are in the final phase of completion. Accordingly, intensive construction activities are underway in Antakya, especially following the Kahramanmaraş earthquake. This situation poses a significant risk to cultural heritage sites. Accordingly, 50 cultural heritage sites remain in the urban sprawl areas of Antakya district, accounting for 35.46% of the total heritage sites (Table 13, Figure 10). In addition, all cultural assets (979) are located outside of urban expansion area.

TABLE 13

Area (km2)Percentage of area (%)Number of cultural heritage sitesPercentage (%)
Urban expansion area47.368.095035.46
Other areas537.7491.919164.54
amount to585.10100141100

Urban sprawl risk levels in cultural heritage sites.

FIGURE 10

3.1.6 Proximity to modern road networks

Road networks increase air and noise pollution, promote urban sprawl, and threaten cultural heritage sites during construction (; ). Buffer zone analysis examined the distance between heritage sites and roads in the research area. Thirty-nine percent of cultural heritage sites are located near the road network (Table 14, Figure 11). In total, 901 (92.03%) cultural assets are within the buffer zone.

TABLE 14

Area (km2)Percentage of area (%)Number of cultural heritage sitesPercentage (%)
Modern road29.425.035539
Other areas555.6894.978661
amount to585.10100141100

Proximity to modern road networks risk levels in cultural heritage sites.

FIGURE 11

3.1.7 Fires

Fires pose a significant threat to heritage sites. Cultural assets made of wood and natural heritage sites are particularly vulnerable to fire and at risk (; Hagage et al., 2023). The research area includes archaeological sites, natural sites, and examples of wooden civil architecture. Therefore, when assessing fire risk in the study, urban sprawl was also taken into account, Sentinel-2 imagery from 2010–2025 was used, and the area was divided into 5 risk groups. As a result, the area with very high and high fire risk covers 82.75 km2, corresponding to 14.14% of the total area (Figure 12). Additionally, according to Table 15, 42 cultural heritage sites are in areas at risk of fire, constituting 29.81% of the heritage sites. Furthermore, 977 of the cultural assets identified in the study fall into the medium- and high-fire risk categories. Burn area analyses from the study area indicate that fires are concentrated under specific environmental conditions. In particular, the distribution of burn areas derived from satellite data was evaluated in relation to environmental factors, including slope and elevation, annual average precipitation, vegetation cover density (NDVI), and land use type. Furthermore, spatial overlay analyses revealed that mid-elevation mountainous areas and forested regions with high vegetation cover density exhibit higher fire occurrence rates. In these areas, precipitation levels exceed the study area average, supporting plant growth and increasing the amount of combustible vegetation. Increased summer temperatures and drought further elevate the risk of fire. Finally, second- and third-degree natural sites are located in areas with high fire risk.

FIGURE 12

TABLE 15

Area (km2)Percentage of area (%)Number of cultural heritage sitesPercentage (%)
Very high risk21.943.7510.74
High risk60.8110.394129.07
Moderate risk377.3164.499567.37
Low risk59.6510.1921.41
Extremely low risk65.3911.1821.41
amount to585.10100141100

Fire risk levels in cultural heritage sites.

3.2 Interpretation and evaluation of multi-risk assessment results

The multi-hazard risk map for Antakya was created by combining the risk maps for erosion, landslides, earthquakes, floods, fires, urban sprawl, and modern road networks in GIS. According to the results, 36.87% of cultural heritage sites are located in high-risk areas, 39.01% in very high-risk areas, and 20.57% in medium-risk areas. Accordingly, 96.45% of cultural heritage sites are located within areas of medium or higher risk. (Table 16, Figure 13). Of the cultural assets, 875 (89.68%) are located in very high- or high-risk areas, 102 in medium-risk areas, and 2 in low-risk areas. Among the city’s cultural heritage sites at very high or high risk, 91 archaeological sites, 3 natural sites, and 18 other conserved areas are vulnerable to multiple hazards.

TABLE 16

Risk assessmentArea (km2)Percentage of area (%)Number of cultural heritage sitesPercentage (%)
Very high risk70.7012.085236.87
High risk125.2721.415539.01
Moderate risk150.3925.702920.57
Low risk166.7328.5053.55
Extremely low risk72.0112.3100
amount to585.10100141100

Assessment of multi-hazard risk levels in cultural heritage sites of Antakya.

FIGURE 13

3.3 Model validation

Since the number of disaster events occurring in the study area is limited, the risk zones and disaster-prone areas announced by AFAD and the Ministry of Environment, Urbanization, and Climate Change were used as references in validating the model. In this context, a spatial overlay analysis was conducted between the generated multi-hazard risk map and these official risk areas. Thus, the model’s spatial accuracy was evaluated. High-risk areas indicate regions that pose a risk of loss of life and property due to soil structure and/or development conditions. Disaster-prone zones, on the other hand, are areas that pose a threat to settlement safety due to disasters; consequently, settlement is prohibited or restricted in these areas (; ; ). In Antakya, areas where landslides and rockfalls have been observed, fault buffer zones, soils with stability issues, and areas sensitive to swelling and settlement have been defined as disaster-prone zones. These zones account for 27.50% of the study area, while high-risk zones account for 0.58%. Verification results indicate that these areas are predominantly classified as high- and very-high risk in the multi-hazard risk map (Figure 14).

FIGURE 14

A total of 7 floods, 203 fires, 6 rockfalls, 8 landslides, and 103 earthquakes with a magnitude of 3 or higher have been recorded in the study area (). When examining the spatial distribution of disaster events, landslides and rockfalls are concentrated in the southern part of the city. Floods, on the other hand, occurred along the tributaries of the Asi River and in low-slope areas. Fires, meanwhile, were largely clustered around the Amanos Mountains and the Amik Plain. This spatial pattern indicates that physical environmental factors are decisive in the occurrence of disaster types. Landslides and rockfalls are associated with steep-sloped, lithologically weak areas; floods with riverbeds and low-gradient basins; and fires with mountain-plain transition zones dominated by forest and maquis vegetation. Consequently, the obtained spatial distributions support the appropriateness and scientific consistency of the control parameters (such as slope, lithology, hydrological structure, and land cover) used in landslide, flood, and fire risk analyses. It is observed that the variables selected in the model are consistent with field observations.

4 Discussion

In this study, risk assessment against multiple hazards was performed using the AHP method. AHP offers a hierarchical structure for multi-criteria decision-making. The main advantages of AHP are its ability to evaluate qualitative and quantitative hazard factors together, to incorporate expert opinion into risk assessment, to measure the consistency of these opinions, and to be applied in a manner compatible with Geographic Information Systems. Thanks to these strengths, AHP is widely used, especially in complex risk analyses. However, the subjective nature of AHP and the independent evaluation of factors impose certain limitations. Future studies are recommended to develop multi-criteria decision-making methods that not only analyze the relationships and interactions among hazard factors in greater detail but also improve spatial modeling and explicitly account for exposure, vulnerability, and resilience as integral parts of risk assessment.

In addition to this research, extreme rainfall caused by climate change affects the hydrological structure. It particularly increases deterioration in archaeological sites. Environmental pollution, dust accumulation, biological pests, air pollution, and tourism pressure also pose cumulative risks over time. Therefore, future studies should consider these factors for sustainable conservation.

Research shows that earthquakes, floods, and urban sprawl pose the highest risks to cultural heritage sites. Therefore, local and central governments, along with all stakeholders, must prioritize each type of risk, develop disaster management plans, create scenarios for potential hazards, and secure the necessary budget.

5 Conclusion

This study identifies the multi-hazard risks facing cultural heritage sites in Antakya using a Geographic Information Systems (GIS)-based approach. The findings indicate that earthquake hazard is the risk factor with the highest weight, followed by flood risk and urban sprawl pressure, respectively. According to the analysis, 75.85% of cultural heritage sites and 89.38% of cultural assets fall into the high or very high risk categories. These percentages largely align with the damage assessment reports published by the Disaster and Emergency Management Presidency (AFAD) and the Ministry of Culture and Tourism following the 6 February 2023, earthquakes in Kahramanmaraş. Indeed, official records have also revealed that approximately 75% of the registered cultural assets in the Antakya Historic City Center sustained severe damage. This indicates that the developed model produces consistent results when compared with existing disaster data.

The results obtained reveal not only the quantitative distribution of risk levels but also the spatial patterns of these risks. High-risk areas are observed to cluster particularly in sections near fault lines, in the floodplains surrounding the Asi River, and in buffer zones where transportation infrastructure is concentrated. This spatial concentration indicates that the convergence of different hazard types increases the vulnerability of heritage areas. For example, the fact that urban sprawl pressure and road effects are also pronounced in regions with high earthquake hazard reveals that these areas are simultaneously exposed to both natural and human-induced risks.

The study also identified distinct differences in the vulnerability levels of various heritage types to these hazards. Monumental and religious structures are more sensitive to multiple hazards, particularly earthquakes and floods, while examples of civil architecture and the traditional street fabric are at greater risk due to pressures from urban sprawl and transportation infrastructure. This finding indicates that conservation and intervention strategies must be developed specifically tailored to the type of heritage.

In conclusion, a comprehensive risk management approach that accounts for the interactions of multiple hazards is necessary to ensure the sustainable preservation of cultural heritage sites in Antakya. In this context: (i) the widespread implementation of structural reinforcement measures against natural disasters such as earthquakes and floods, and the enhancement of post-disaster response capacity, (ii) limiting pressures from urban sprawl and transportation through planning and legal regulations, (iii) developing ecosystem-based measures against secondary hazards such as fire and erosion, and (iv) establishing protection policies tailored to different types of heritage emerge as priority steps. Furthermore, it is of great importance that large-scale decisions—such as post-disaster reconstruction processes and urban planning policies—be addressed in a manner consistent with the protection of cultural heritage.

Statements

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

Ethics statement

Ethical approval was not required for the studies involving humans because the conducted study includes a spatial risk analysis. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements because the vulnerability of cultural heritage was determined by creating risk maps. No survey or questionnaire study was conducted in this research.

Author contributions

GB: Conceptualization, Data curation, Formal Analysis, Investigation, Methodology, Resources, Software, Supervision, Visualization, Writing – original draft, Writing – review and editing.

Funding

The author(s) declared that financial support was not received for this work and/or its publication.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/feart.2026.1837277/full#supplementary-material

References

Summary

Keywords

cultural heritage sites, cultural heritage assets, multi-hazard, risk, GIS, AHP

Citation

Başdoğan Deniz G (2026) GIS-based spatial multi-hazard vulnerability assessment of cultural heritage sites: the case of Antakya. Front. Earth Sci. 14:1837277. doi: 10.3389/feart.2026.1837277

Received

23 March 2026

Revised

19 April 2026

Accepted

06 May 2026

Published

05 June 2026

Volume

14 - 2026

Edited by

Fatih Adıgüzel, Bitlis Eren University, Türkiye

Reviewed by

Qian Cheng, Shenyang Agricultural University, China

Yan Dai, Chongqing University, China

Updates

Copyright

© 2026 Başdoğan Deniz.

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Gülçinay Başdoğan Deniz,

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Outline

Figures

Cite article

Copy to clipboard


Export citation file


Share article

Share on WeChat

Scan with WeChat to share this article

Article metrics