Introduction

Problems arising from disasters such as earthquakes, floods and landslides severely affect daily life. In the post-disaster process, it is essential for central and local administrations to make fast and accurate decisions and to manage the disaster and emergency in a data-driven manner. Using Geographic Information Systems (GIS), effective decision-making can be provided by supporting the preparedness, mitigation, response and recovery phases of disaster management (Cova 1999; Eichelberger 2018; Gunes and Kovel 2000).

Capabilities of GIS such as data management, risk analysis, spatial accessibility and planning facilitate process management in location-based environmental applications. Since disasters are directly related to geography, GIS is used effectively in disaster management (Tomaszewski 2020). In GIS-based disaster and emergency management, there are diverse application areas such as flood risk analysis (Chen et al. 2011; Deckers et al. 2010; Tran et al. 2009; Zerger and Wealands 2004) and landslide risk analysis (Assilzadeh et al. 2010; Ayalew et al. 2004; Carrara et al. 1999).

GIS makes an essential contribution in taking fast and accurate steps to solve environmental problems that may arise after disasters (Abdalla and Tao 2005; Hashemi and Alesheikh 2011; Parizi et al. 2022; Walker et al. 2021; Wang et al. 2012). It also provides an effective management of solid wastes in the collection, storage, and disposal stages. Wastes and rubble generated after earthquakes can be handled in this context. Pre-determination of the landfill areas will shorten the response time and reduce indirect environmental damage (Grzeda et al. 2014). Thus, there is a need for a spatial model to determine suitable landfill sites systematically.

Multi-criteria decision-making (MCDM) methods can be used to transform and aggregate value judgment preferences to obtain the information necessary for decision-making (Malczewski 1999; Malczewski and Rinner 2015). Spatial MCDM applications frequently use ranking, scoring and pairwise comparison methods (Greene et al. 2011; Malczewski 2006). In GIS-based disaster management studies, criteria weighting methods such as Analytic Hierarchy Process (AHP) (Saaty 1990), ELimination Et Choix Traduisant la REalité (ELECTRE) (Roy 1968), Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) (Hwang and Yoon 1981), VIseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR) (Opricovic 1998), Complex Proportional Assessment (COPRAS) (Zavadskas and Kaklauskas 1996), Preference Ranking Organization Method for Enrichment of Evaluations (PROMETHEE) (Brans and Vincke 1985), COmbinative Distance-based ASsessment (CODAS) (Keshavarz Ghorabaee et al. 2016), and Best-Worst Method (BWM) (Rezaei 2016) can be used to determine the criteria weights (Anton et al. 2006; Cegan et al. 2017; Gilliams et al. 2005; Nyimbili and Erden 2020; Malczewski 2006; Rezaei 2016). Due to the number of criteria used in landfill site selection studies, weight calculation becomes quite complex with pairwise comparison-based methods. Since more comparisons are made in BWM compared to other MCDM methods (such as AHP), the calculation step is reduced, and weights can be determined with a high consistency rate (Mete and Yomralioglu 2019; Rezaei 2016).

Determination of the earthquake demolition waste landfill areas is considered similar to solid waste management (Rahimi et al. 2020) and land suitability assessment studies (Benaissa and Khalfallah 2021) in terms of methodological approach. GIS-based MCDM methods are used for site selection of landfill areas in existing studies (Tercan et al. 2020; Kang et al. 2024; Ali et al. 2023; Bhowmick et al. 2024; Armanuos et al. 2023; Liu 2022). In GIS-supported disaster management studies, the AHP method is generally used with limited criteria (Greene et al. 2011; Aksoy and San 2019; Roy et al. 2022; Kazuva et al. 2021; Zewdie and Yeshanew 2023). In this study, openly licensed spatial data are obtained from different data sources such as OSM and EEA Copernicus and weights are determined for 14 criteria using the BWM method with high consistency.

Research questions of the study can be named as: How to create a scalable, highly accessible, secure, functional, and reliable disaster management system that can be used immediately after earthquakes? Can we utilize the cloud-GIS approach in earthquake demolition waste management for prompt action? Storing, processing, analyzing and sharing spatial data within an information system facilitates disaster and emergency management. On the other hand, due to the damaged technological infrastructure, decision support systems may be disabled after the disaster and spatial and non-spatial data of the region may be lost. To prevent this issue, a decentralized cloud computing infrastructure is needed. In case of a problem in a data center, services can be continued by providing access from other regions.

The primary purpose of this study is to determine the most suitable areas where earthquake debris wastes can be transported by analyzing openly licensed spatial data on a cloud-GIS framework. This is the first study on landfill site selection for the Kahramanmaraş, Türkiye Earthquake Sequence. The study brings a new perspective to the current disaster management approaches by developing a cloud GIS-based disaster management portal and creating a high-performance and low-cost infrastructure with serverless cloud architecture. Using open-source data and software, a methodology that enables rapid action immediate aftermath of the disaster is followed. The study also covers the determination of weights by using BWM to produce suitability maps based on these weights and identify potential sites. Unlike existing studies, this study provides significant improvements in disaster management processes in terms of speed, cost, and security with a cloud-GIS-supported data management approach.

Material and methods

On February 6, 2023, two massive earthquakes centered in the districts of Pazarcık (Mw 7.8) and Elbistan (Mw 7.7) in Kahramanmaraş province of Türkiye struck a large region. This disaster caused the collapse of many buildings in 11 cities, resulting in millions of cubic meters of rubble (Fig. 1). After this devastating disaster, there was a need to quickly identify the areas where the earthquake demolition waste would be transported so that recovery work could begin. Although there are landfill sites previously determined by the municipalities, it is observed that new areas are needed because the specified sites are inadequate since they do not meet the necessary conditions or have lost their suitability.

Fig. 1
figure 1

Study area: provinces that have been affected by the earthquake

Full size image

Within the scope of the study, criteria are determined for the location selection of temporary and permanent waste disposal areas with cloud GIS-supported disaster management approach, necessary data are obtained from open data sources, spatial analyses are performed with QGIS open source GIS software, suitability maps are produced by determining criteria weights with the BWM, and a Cloud GIS-based disaster management portal is developed using Amazon Web Services (AWS) (Fig. 2).

Fig. 2
figure 2

Workflow diagram of the Cloud-GIS based disaster management approach

Full size image

Determination of the decision criteria

Solid waste management includes collecting, treating, and disposing of waste from human activities or disasters to avoid environmental damage. Selecting a suitable site for waste management is done by considering various factors. These factors include geographical features, soil characteristics, population density, transportation, environmental sensitivity, and economic factors (Asefa et al. 2022; Zewdie and Yeshanew 2023).

Geographical features include proximity to underground and surface water resources, the direction of surface water valleys and rivers, and distance from fault lines (Sener et al. 2010; Wang et al. 2009). Soil properties include permeability and mineral and organic matter content, which determine the risk of waste penetrating the soil and polluting the environment. The selected landfill areas should be distant from densely populated regions within a specified distance from settlements (Fard et al. 2022). Within the scope of the transportation criterion, the waste disposal location should be easily accessible by vehicles. In addition, it should not be close to protected areas, forests, agricultural lands, or settlements. Thus, public health and the environment can be prevented from harmful effects of the waste (Nickdoost and Choi 2023). Finally, economic factors require consideration of the necessary costs to carry out waste disposal operations and manage the site. Therefore, the existing road network and the distribution of building demolition waste should be considered.

While determining decision criteria and their weights, the Delphi method is used with the BWM. The Delphi approach aims to collect expert opinions on the relative importance of criteria within a panel. The expert panel consists of multiple rounds of questionnaires to gain consensus opinion (Grisham 2009; Han et al. 2023). In this sense, an expert group is created after the earthquake with the participation of five academics from different departments (Table 1). In the expert panel, a structured questionnaire is provided to rate the importance of each criterion.

Table 1 Delphi analysis expert group information
Full size table

According to the three-stage Delphi method, the criteria set is determined in the first step (Table 2). In the second step, pairwise comparisons of the criteria are made. In the third step, the relative importance (weights) of the criteria are determined using the BWM.

Table 2 Classification of landfill site selection criteria
Full size table

At the end of the first round, fourteen criteria are determined for the site selection of landfill areas: C1: Proximity to the Settlement Areas, C2: Bareland, C3: Proximity to the Water Bodies, C4: Proximity to the Demolished Buildings, C5: Slope, C6: Proximity to the Roads, C7: Proximity to the Railway, C8: Geology, C9: Proximity to the Airports, C10: Forest, C11: Agricultural Land, C12: Shrub, C13: Pasture, C14: Population Density.

In the second round, a pairwise comparison matrix is created by determination of the best criterion over all the other criteria (BO) and other criteria over the worst criterion (OW) (Table 3). In order to calculate the consistency of the scores given after the pairwise comparisons, the internal consistency of the experts’ scoring is calculated with Cronbach’s \(\alpha\) method. Cronbach’s \(\alpha\) is a reliability coefficient and a measure of the internal consistency of tests and measures (Cronbach 1951). It is specified by calculating individual scores assigned to each item within a scale, followed by correlating these scores with the overall score for each instance. The resulting correlations are subsequently compared with the variance observed among all individual item scores (1). Cronbach’s alpha is intricately tied to factors, including the number of questions or items within a measurement, the mean covariance among pairs of items, and the comprehensive variance of the overall score being measured. The alpha value of the pairwise comparison is calculated as 0.841 (exceeded the threshold of 0.7), thus validating the consistency of the judgment of the experts in the Delphi panel.

Table 3 Pairwise comparison matrix created by experts’ opinion
Full size table
$$\begin{aligned} \alpha = \frac{k}{k-1}\left[ \frac{\sigma _{t}^{2}-\sum \sigma _{i}^{2}}{\sigma _{t}^{2}} \right] \end{aligned}$$
(1)

where k is the number of criterion, \(\sigma _{t}^{2}\) is the variance of the sum column, \(\sigma _{i}^{2}\) is the variance of each variable.

Calculation of the criterion weights

A decision is the act of choosing among alternatives. Criteria are the rules for a decision that can be measured or evaluated. MCDM methods are all tools and approaches used to transform and combine value judgment preferences in obtaining the information necessary for decision-making. There are three basic concepts in solving MCDM problems. These concepts are value measurement (or standardization), criteria weighting and decision rule. There are a variety of MCDM techniques, such as AHP, COPRAS, TOPSIS, CODAS, BWM that can be preferred for landfill site selection study. In this study, the BWM is adopted as a decision-making method.

BWM has a new approach compared to other methods. With BWM, pairwise comparisons are made to find the weights to see how much decision-makers prefer one criterion over another. BWM is an easy-to-understand and easy-to-use MCDM method. In addition, it facilitates decision-making by scoring pairwise comparisons in a particular order according to the Likert scale, making it possible to make consistent comparisons. The resulting weights are consistent and reliable. The calculation step is reduced since fewer comparisons are made in BWM compared to many other MCDM methods (such as AHP).

The decision-making process according to BWM can be briefly summarized as follows: First, a set of decision-making criteria is determined. After determining the criteria, in the second stage, the best and the worst among these criteria are selected. In the third stage, the degree of preference for the best criterion over the other criteria should be determined. If a range between 1 and 9 is used as a comparison scale, a score of 1 is given for the criterion equally important to the best criterion. Conversely, a score of 9 is given when the best criterion significantly outweighs the other criteria. In the fourth stage, the process of pairwise comparisons is finalized by ascertaining the extent of preference for the worst criterion in relation to the others.

The weights and consistency ratio (\(\xi\)) are determined by solving the following inequalities (2):

$$\begin{aligned} \begin{aligned} \text {minimum } \xi ^{L} \text { such that;}\\ \left| w_{B} - a_{Bj}w_{j} \right| \le \xi ^{L} \text {, for all j,}\\ \left| w_{j} - a_{jw}w_{w} \right| \le \xi ^{L} \text {, for all j,}\\ \sum \limits _{j}^{}w_{j} = 1,\\ w_{j} \ge 0 \text {, for all j} \end{aligned} \end{aligned}$$
(2)

where \(a_{Bj}\) is preference of the best criterion over criterion j, \(a_{jW}\) is preference of criterion j over the worst criterion.

By solving this linear model, unique criteria weights are obtained (Fig. 3). \(\xi\) is considered as a good indicator of the consistency of the model. This ratio can vary between 0 and 1, but for a high consistency, \(\xi \le 0.25\) is expected. The consistency ratio of the model is calculated as 0.072, which validates the consistency of the judgment of the experts in the pairwise comparison.

Fig. 3
figure 3

Criteria weights determined by using the BWM

Full size image

Spatial analysis

After determining the criteria for the analysis of the most appropriate landfill site selection areas, data for the disaster area are obtained from the relevant sources, and spatial analyses such as proximity and slope are performed using QGIS software.

One essential criterion for the site selection of landfills is that the land should have a solid geological formation with low liquid permeability. The General Directorate of Mineral Research and Exploration (GDMRE) data are used to analyze the geological status of the cities in the disaster area, and soil structures such as sedimentary and metamorphic for landfills are classified as highly suitable in the site selection analysis.

In site selection, the topographic characteristics of the land are significant for both the establishment of the landfill and the ease of transportation of waste to the landfill areas. In this context, slope analysis is performed using EUDEM v1.1 Digital Elevation Model data shared by the Copernicus Land Monitoring Service (CLMS) program of the European Environment Agency with a spatial resolution of 25 ms. As a result of the analysis, the suitability intervals for site selection are reclassified between 0 and 5, and regions with low slopes are given high scores.

Another characteristic of the areas to be identified for waste management is proximity to the road network. A location close to main roads and intersections should be preferred for easy transportation to the relevant locations. In order to examine this criterion, proximity analysis is carried out using OpenStreetMap (OSM) road network data, which is significantly updated by volunteers after the disaster.

The regulations in the waste management legislation in Türkiye state that landfills should be located at least 1 km away from the settlements (Official Gazette 1991, 2019). CLMS European Settlement Map open data is used for the analysis of proximity to settlements and classified in such a way that places far from settlements receive higher scores. On the other hand, in order to select landfills away from densely populated centers, raster data from the Turkish Statistical Institute (TurkStat) for 2021 indicating the population density per 1 \(km^{2}\) area is used.

Land use and land cover are of great importance in GIS-based site selection studies. Land use types are classified into five basic classes: artificial surfaces, agricultural areas, forests, wetlands and water bodies. In the study of site selection of landfill areas where construction and demolition waste will be temporarily or permanently stored and transformed, it should be located at a certain distance from land use classes, such as urban areas, industrial and commercial zones, and transportation networks, among artificial surfaces. In the study, agricultural areas, forests, wetlands and water bodies are also considered unsuitable locations.

The proximity criterion to the airport is vital in terms of not jeopardizing flight safety and not disrupting transportation in disaster areas. In the study, buffer analysis is performed for airports, and the surrounding region is included in the unsuitable areas.

After calculating criteria weights and analyzing spatial criteria, a suitability map is produced in QGIS using the Weighted Linear Combination (WLC) approach to determine the areas where earthquake debris waste can be transported (3). In the WLC approach, the normalized analysis results of the criteria are multiplied by the weight of the relevant criterion to obtain a weighted criterion score, and a suitability map can be produced with the weighted sum of all criteria in GIS software. An algorithm pseudocode is also given for describing coding operations and their correct sequence within the algorithm.

$$\begin{aligned} V(x_{i}) = \sum \limits _{j} w_{j}v_{j}(x_{i}) = \sum \limits _{j} w_{j}r_{ij} \end{aligned}$$
(3)

where \(V(x_{i})\) is suitability value, \(w_{j}\) is criterion weight, \(v_{j}(x_{i})\) is the value function for the j-th attribute, \(r_{ij}\) is the criterion score.

Algorithm 1
figure a

Calculate suitability index

Full size image

By integrating the data for site selection of waste facilities, suitable and unsuitable areas are identified. Figure 4 shows the site selection map of the study area as a result of the suitability analysis for waste storage purposes. Potential landfill sites are identified on the suitability map for Kahramanmaraş and Hatay provinces.

Fig. 4
figure 4

Landfill suitability map of the study area

Full size image

Sensitivity analysis

Sensitivity analysis can be defined as an analytical process that identifies the change in the calculated weights under different scenarios and priorities. In order to obtain objective criteria weights, sensitivity analysis ensures reliability in a decision-making process. In this study, three sustainability criteria are determined as decision alternatives: environmental, social, and economic. In the Delphi expert panel, respondents are also asked to give different scores for each scenario while comparing criteria. Then, criteria weights for each scenario are calculated by using the BWM, and final criteria weights are obtained by averaging five expert opinions (Table 4). After determining the weights for each of the three scenarios, suitability maps of the relevant priorities are produced (Fig. 5).

Table 4 Criteria weights obtained from sensitivity analysis
Full size table
Fig. 5
figure 5

Suitability map of sustainability alternatives

Full size image

Development of disaster management portal with serverless cloud architecture

In disaster management, map-based decision support systems provide critical infrastructure for various purposes such as planning, coordination and risk analysis. With GIS portals that can be accessed over the web, spatial data management can be realized, and functional applications can be developed. In this study, a serverless cloud architecture framework is created on AWS to develop a disaster management portal (Fig. 6). In serverless cloud architecture, creating a virtual machine for running a GIS application is unnecessary. The cloud provider performs all operations, including software installations, updates, system management, and maintenance. Amazon Aurora DB - Serverless PostgreSQL and PostGIS extension are used for spatial data storage. Aurora Serverless PostgreSQL database can be accessed through a VPN connection, SSH or RDS Query Editor.

Fig. 6
figure 6

AWS serverless cloud GIS architecture

Full size image

After completing the database setup, vector data tiles are published in MVT format to optimize the visualization of big vector data on web platforms. The data tiles are served directly from the database using PostGIS ST_AsMVTGeom() and ST_AsMVT() functions dynamically. This method removes the necessity for an intermediary server installation. In this approach, incoming map tile requests are forwarded to AWS Lambda by AWS API Gateway in order to check the location information of the API call. The relevant vector data tile is retrieved from the database through the Data API for Aurora Serverless v1 and sent to the client. SQL statements can be executed in a secure HTTP endpoint using Data API without connecting to the database cluster persistently (Mete and Yomralioglu 2021).

On the other hand, in order to share raster-based suitability maps as a web service, raster tiles are generated in MBTiles format using the “Generate XYZ Tiles (MBTiles)” tool in QGIS and uploaded to the Amazon S3 object storage bucket. Object sharing settings and policies such as public access and cross-origin resource sharing (CORS) are configured, and raster tiles are shared on the web by using S3 and Lambda serverless services. Edge caching is also enabled using AWS CloudFront Content Delivery Network (CDN) to increase the speed at which the data is displayed on the user side. The web service of the pre-rendered raster image tiles with zoom level and x/y coordinate information can be accessed using the XYZ data tile protocol.

Raster and vector data of the earthquake-affected areas can be accessed through both desktop GIS software and browsers using web mapping libraries. There are several JavaScript web mapping libraries, such as ArcGIS Maps SDK for JavaScript, Mapbox GL JS, MapLibre GL JS, Leaflet JS, and OpenLayers for interacting with maps in a web browser. In this study, the Leaflet open source web mapping library is used, considering features such as design, readability of the source code, availability of extensions, and detailed documentation. This JavaScript-based mapping library uses the Web Graphics Library (WebGL) to create web maps for desktop and mobile platforms. Web mapping libraries are compatible with desktop and mobile platforms and can be customized with many plugins to increase functionality.

The landfill suitability map is presented on the web through the serverless cloud-GIS-based disaster management portal. Users can view the most suitable areas on the raster map by using a web browser, query the information of potential sites and get information about many attributes such as surface area, ownership status, land use, soil condition, proximity to the city centre and average slope. On the other hand, satellite images, orthophotos, topographic maps and various vector maps can be added to the platform as base maps via REST API and OGC web services. It is also possible to develop navigation tools, distance measurement, and address search functions on the portal.

Results and discussion

Open data are valuable resources for natural disaster studies. Moreover, the data should be valid, accurate and up-to-date. Hence, developing open data policies by central and local governments will significantly contribute to GIS-based disaster management applications. After Kahramanmaraş, Türkiye earthquakes, several non-governmental organizations and individuals have started disaster relief projects such as Humanitarian OpenStreetMap (HOT), Yer Çizenler, Açık Yazılım Aǧı (AYA). Volunteered Geographic Information (VGI) projects have led to saving lives and facilitated humanitarian aid in the disaster region with data and maps.

In this study, several open data sets are used to create a disaster management database. Criteria weights for landfill site selection were determined using BWM with a high consistency rate. It is seen that the use of the BWM method in spatial decision-making problems is appropriate due to its advantages, such as the low number of pairwise comparisons and high consistency ratio. Using the suitability map produced after GIS analysis, unsuitable areas such as underground and surface water bodies, forests, agricultural lands, and natural protection zones were masked, and the negative impact of landfills on the environment and people was prevented. A certain threshold value was set for suitable areas, and detailed controls such as road access and ownership status were carried out for areas above this threshold value. In addition, potentially suitable areas were also examined through the 3D topographic base and satellite images to verify the slope of the land and the current status of the sites. Thus, the most suitable landfills for the transportation of earthquake debris wastes were determined in a data-driven manner.

Immediately after the earthquake, the Ministry of Environment, Urbanization and Climate Change (MEUCC) of Türkiye conducted damage assessments and shared statistics on building damage by province. According to these data, 71,735 independent sections in Hatay, 60,051 in Kahramanmaraş and 29,703 in Adıyaman were found to be in urgent need of demolition, heavily damaged or collapsed (MEUCC 2023). Within the scope of the site selection study, landfills with an area of 34.7 million \(\text{m}^{2}\) in Hatay, 29.7 million \(\text{m}^{2}\) in Kahramanmaraş and 7.9 million \(\text{m}^{2}\) in Adıyaman were identified. On the other hand, in the earthquake report published by Istanbul Technical University (ITU), approximate amounts of earthquake demolition waste were calculated on a provincial basis according to the statistics of damaged buildings shared by the ministry (ITU 2023). When the capacities of the most suitable landfills are analyzed on a provincial basis, it is seen that they are sufficient to store large amounts of waste resulting from earthquakes.

The sensitivity analysis indicates how responsive is the suitability index under different priorities of decision-makers. The area of suitability classes changes in the study region according to different scenarios. In the environmental priority, 32.92 % (29,633 \(\text{km}^{2}\)) of the study area was calculated as suitable, and 14.46 % (11,932 \(\text{km}^{2}\)) was calculated as very suitable for landfill sites. In the scenario where economic impacts are predominant, 28.90 % (23,837 \(\text{km}^{2}\)) of the region was calculated as suitable, and 9.15 % (7549 \(\text{km}^{2}\)) was calculated as very suitable. In the scenario in which social criteria were at the forefront, a suitable class covered 27.43 % (22,627 \(\text{km}^{2}\)) of the study area, while the very suitable class was calculated as 6.79 % (5598 \(\text{km}^{2}\)) (Fig. 7). When these changes are analyzed, it is seen that there are limited suitable areas for landfill of the earthquake demolition waste in the study region. For instance, the city center of the Hatay province was established around the Asi River. There are also a great amount of olive groves in the city. Therefore, it is hard to find suitable sites for landfill areas near the demolished buildings. Integrating MCDM and cloud-based GIS facilitates the selection of suitable sites consistently. Depending on many factors, manually determining adequate landfill sites is a complicated and time-consuming process. Thanks to the holistic disaster management approach, the site selection study was carried out quickly and consistently. Figure 8 shows determined potential landfill sites in Hatay and Kahramanmaraş provinces.

Fig. 7
figure 7

Suitable areas in different sustainability alternatives

Full size image
Fig. 8
figure 8

Determined potential landfill sites in Hatay and Kahramanmaraş provinces

Full size image

After identifying suitable sites, the results were shared through the Cloud GIS-based disaster management portal (Fig. 9). A navigation application was developed to enable teams to access the sites identified for each province via the appropriate route on the map, and demolition waste was transported to the nearest available landfill site. It was seen that a significant advantage was obtained in planning and coordination by using the data of highway network, most suitable waste storage areas and demolished structures in GIS-supported proximity analysis.

Fig. 9
figure 9

Cloud-GIS based disaster management portal

Full size image

Access to the disaster management portal with a serverless cloud computing infrastructure was provided seamlessly from all platforms, such as phones, tablets and computers. Thanks to the scalable nature of this approach, which offers high performance, services are suspended when not in use, providing a significant advantage in terms of cost.

In GIS-supported disaster management studies, the AHP method is generally used with limited criteria (Greene et al. 2011; Aksoy and San 2019; Roy et al. 2022; Kazuva et al. 2021; Zewdie and Yeshanew 2023). In this study, openly licensed spatial data were obtained from different data sources such as OSM and EEA Copernicus and weights were determined for 14 criteria using the BWM method with high consistency. The originality of this study can be named as determining suitable landfill areas with open-source data and developing a scalable GIS environment on the cloud for effective disaster management. In order to provide a systematic and expert-driven approach in justifying weight calculation, modified Delphi-BWM was used. Cronbach’s \(\alpha\) reliability coefficient and consistency ratio of the BWM showed that the decision-making process was completed with several successive rounds, and expert consensus was achieved. In addition, the suitability maps were shared on the disaster management portal using serverless cloud architecture with high performance and accessibility rates. In this respect, it was observed that the study brought a new dimension to the holistic disaster management approach with its elements such as data acquisition, visualization, analysis and sharing.

Conclusion

Earthquakes have multidimensional consequences for both cities and citizens. Therefore, ensuring the planned collection, storage and recycling of waste is crucial to effectively carrying out disaster management processes. There is a need for a systematic approach to determine realistic and feasible planning in disaster relief processes.

In this study, a site selection model was developed to determine the most suitable landfill areas after the Kahramanmaraş, Türkiye earthquakes. In this context, the Delphi Method was utilized to determine the decision criteria set in an expert panel. In the second round, pairwise comparisons were carried out. To determine the criteria weights with an objective approach, the BWM method was applied in the final stage. To produce a suitability map of landfill areas, site selection criteria were analyzed in the QGIS open source software. As a result, the potential landfill sites were determined by identifying high suitability scores, and the spatial data were shared on the web utilizing cloud GIS services. After setting up solid waste recycling facilities in the selected landfill areas, transported earthquake debris can be sorted and recycled. This will contribute to protecting the environment by ensuring that harmful wastes are disposed of and useful wastes are recycled and reused.

Using open-source software and openly licensed data for developing a GIS-based disaster management portal with serverless cloud architecture stands out as the originality of this study. Pre-development of disaster-specific conceptual and physical models will help quick response after disasters. Within the scope of provincial disaster risk mitigation plans, developing the model for landfill site selection model for earthquake demolition wastes will contribute to the preparedness phase of disaster management by defining the necessary data sets, data types, attributes and analyses. While preparing disaster-specific data models, The Federal Emergency Management Agency (FEMA) resources and standards can be utilized (FEMA 2023).