Introduction

Delta megacities, hubs of burgeoning urban-industrial development, offer vital socioeconomic opportunities in developing countries worldwide (Youssef et al., 2024). Unfortunately, these countries, with only 16% of the global population, generate 34% of the world's waste, posing significant environmental challenges in highly populated riverine areas (Van den Brandeler et al., 2019; Xu et al., 2019). Decades of rapid urbanization have worsened sanitation and solid waste (SW) management issues (Yin et al., 2021). Unmanaged SW disposal near agricultural lands and water resources cause diverse environmental and health risks due to hazardous pollutants, such as heavy metals, industrial waste, pesticides, and volatile organic compounds ( Blight & Mbande, 1996; Lim et al., 2016; Santhosh & Babu, 2018; Akhtar et al., 2021). This disrupts surrounding aquatic and terrestrial ecosystems in riverine megacities (Tortajada, 2008; Youssef et al., 2024).

Globally, developing countries allocate 20–50% of municipal budgets to waste management (Guerrero et al., 2013), yet only collect 50–80% of generated waste due to high costs (80–95%) associated with collection and transportation (Abdel-Shafy & Mansour, 2018; Simion et al., 2013). Inadequate waste management, such as poor collection and disposal, poses significant environmental threats in delta megacities like Cairo in the Nile Delta (Abdallah et al., 2020; Fahmi & Sutton, 2010), Dhaka in the Ganges–Brahmaputra delta (Ahmed & Rahman, 2000; Chowdhury & Afza, 2006; Yin et al., 2021), Bangkok in Chao Phraya Delta, Karachi in Indus delta (Soudani et al., 2024), and Mumbai in western Indian deltas (Iyer, 2016; Shah, 2021). South Asia, the largest waste-generating region globally (Kaza et al., 2018), produces around 70 million tonnes annually, with 426 million residents (Hoornweg & Bhada-Tata, 2012). In 2014, Bangladesh generated 23,688 tons of urban solid waste per day, with Dhaka accounting for 6,026 tons daily (Waste Concern, 2014). Within these challenges, delineating Megacity Solid Waste Disposal (MSWD) suitability sites is essential for achieving Sustainable Development Goals (SDGs), including 6 (clean water and sanitation), 11 (cultivating sustainable cities and inclusive communities), 13 (climate action), and 15 (conserving terrestrial ecosystems) in these metropolises (United Nations, 2018; Xu et al., 2019; Bilgilioglu et al., 2022).

Analyzing MSWD suitability requires considering various factors, including geomorphological, geological, hydrogeological, socioeconomic, and climatological conditions (Beskese et al., 2015; Santhosh & Babu, 2018). Traditional methods are limited, and integrating Remote Sensing (RS) and Geographic Information Systems (GIS) integration is imperative is essential for identifying potential MSWD sites (Halvadakis, 1993; Bilgilioglu et al., 2022). The multispectral Landsat series serves as a pivotal resource for mapping Land Use and Land Cover (LULC), enabling the delineation of human activities, socioeconomic aspects, and environmental conditions (Elmahdy & Mohamed, 2016; Youssef et al., 2021). Satellite-derived Digital Elevation Models (DEMs) have undoubtedly facilitated the acquisition of precise geomorphological information, encompassing topography, stream networks, and flow patterns (El-Aal et al., 2024; Embaby et al., 2024; Abu El-Magd et al., 2024). Noteworthy, the integration of Landsat and DEM imagery with subsurface (lithological and hydrogeological) datasets in GIS has recently improved our understanding of the near-surface environment, including pollutant accumulation and contaminant infiltration in the topsoil (Khalil et al., 2022; Youssef et al., 2021).

Diverse GIS-based Multi-Criteria Decision Analysis (MCDA) techniques, such as Weighted Linear Combination (WLC), Analytical Hierarchy Process (AHP), and fuzzy logic overlay, have been widely employed to determine optimal SW disposal locations (Beskese et al., 2015; Lim et al., 2016; Paul & Ghosh, 2022). Among these, the AHP excels in structuring expert judgment through pairwise comparisons, producing a list of factors weighing (Saaty, 1980). Yesilnacar et al. (2012) found that WLC may not always be suitable due to rapid parameter changes. Bilgilioglu et al. (2022) utilized a GIS-integrated AHP approach for landfill site selection, demonstrating its effectiveness in the densely populated regions. As a result, several studies have successfully employed AHP to select sustainable SW disposal sites across various regions of the world, i.e., Cao et al. (2006), Paul et al. (2014); Khan & Samadder (2015); Sam and Steven (2017); Ali et al. (2017); Tercan et al. (2020); Bilgilioglu et al. (2022); Paul and Ghosh (2022).

In the context of South Asia, the Ganges–Brahmaputra area is the most populous delta, hosting over 150 million people across Bangladesh, Bengal, and India (Chakraborty et al., 2021; UNDESA/PD, 2019; Edmonds et al., 2020). Dhaka City, the capital of Bangladesh (Fig. 1a), houses over 19% of this population and experiences a rapid growth rate of 37% (Rahman et al., 2020; Razia et al., 2023), making it on of the most world's densest megacities (Demographia, 2019). Dhaka's metropolitan area experienced significant growth in 2020, with a population of 21,006,000, reflecting a 3.56% increase from 2019 (Yin et al., 2021). This population rumble places significant strain on waste management infrastructure and resources, driven by escalating food demands and consumption patterns, leading to extraordinary challenges for sustainable SW disposal (APO, 2007; Streatfield & Karar, 2008; Yin et al., 2021). Typically, the waste management system is community-based in the city, lacking involvement from the scientific community in the planning and mitigation efforts (Ahmed et al., 2018; Salam, 2000; Yousuf, 1996). For example, Dhaka residents generate approximately 4632 tons of waste daily (Rahman & Kashif, 2009), with an average of 40% of the generated waste remaining uncollected (Islam, 2016). Ahmed and Rahman (2000) and APO (2007) indicate that annually, around 1.65 million tons of solid waste is produced from households, industries, and growth centers in Dhaka, translating to approximately 0.29 to 0.60 kg/person/day. Given Dhaka's population exceeding 10.2 million, a comprehensive study of this area is imperative for informed decision-making (Fig. 2).

Fig. 1
figure 1

a Location of the study area; b Elevation map displays Bangladesh country in the southern part of Asia, c Administrative map of Dhaka metropolis in the middle part of Bangladesh showing the capital city (light green polygon) boundary, and (d) the location of the study area, as well the current locations of Solid Waste (SW) disposal site by black points

Full size image
Fig. 2
figure 2

Some collected photos during field observations in Dhaka city describing a) Solid waste was traditionally collected from households in the Kallyanpur region (Date: 22.10.2022), b) Unmanaged solid waste disposal practices near the Roadside In-Mirpur-10 (Date: 18.11.2022), c) Amin Bazar waste disposal site was designed as a landfill in the western part of Dhaka city (Date: 27.10.2022), d) Initiative of power generation 42.5MW at Amin Bazar waste disposal site was designed as a landfill in the western part of Dhaka city (Date: 25.05.2024), e) the Matuaill landfill site beside the building (Date: 20.12.2022), f) Kallayanpur wastewater canal was established to transport both domestic and industrial water (Date: 22.10.2022)

Full size image

Several researchers therefore studied waste disposal management strategies and rehabilitation approaches in the Dhaka megacity, i.e., Yousuf (1996), Salam (2000), Chowdhury and Afza (2006), Zahur (2007), Tania (2014), Islam (2016), Ahmed et al. (2018), and Yin et al., 2021. On a local scale, Hasan et al. (2009) developed an integrated fuzzy set theory and WLC techniques in a GIS to improve landfill mapping in three districts, namely Dhaka, Narayanganj, and Gazipur. Islam et al., (2016) utilized a GIS-based optimization scenario to identify optimal relocations for the insufficient waste collection containers and design more efficient collection routes in Mirpur-Pallabi in northern Dhaka City. Recently, Akther et al. (2019) integrated MCDA and GIS techniques to find suitable SW sites considering environmental, social, and economic determinants in a zone of Dhaka. Despite these efforts, AHP-based GIS modelling has not been entirely successful in pinpointing suitable MSWD sites for the growing population in Dhaka. This research gap highlights the urgent need for holistic and data-driven techniques to effectively address solid waste disposal in this megacity. To this end, this work typically considers the impact of both surface (socioeconomic, geomorphological, and climatological) and subsurface (lithological and hydrogeological) factors by integrating site suitability with a land-use and hydro-environmental evaluation analysis. The current research, therefore, aims to: (1) find out suitable sites for the MSWD using WLC; and (2) defining the optimal future waste disposal locations considering the hydro-environmental aspects and related problems in Dhaka city.

Study area characteristics and environmental challenges

The study area, situated in the central region of Bangladesh, encompasses the Turag, Balu, Lakkya, Shitalakshya, and Burigonga rivers; covering a total area of 306 km2 (Fig. 1b). It falls within the North and South City Corporations of Dhaka City (Abir et al., 2021), which is situated between latitudes 90°16′13" -90°33′86" E and longitudes 23°56′44" -23°37′27" N (Fig. 1c). The proposed study area within the Madhupur tract is geologically significant due to its uplifted position relative to surroundings, the Enachulen fault structure, and lineaments influencing river flow (Alam et al., 1990). These features indicate ongoing tectonic activity (Maitra & Akhter, 2011). Additionally, the area's aquifer is recharged by nearby rivers during monsoons (Hasan et al., 1999). Kamal and Midorikawa (2006) produced a geomorphological map of Dhaka city using aerial photographs from 1954 (scale: 1:40000), incorporating information from fill-thickness, boreholes, and an old topographic map (see Fig. 3). The delineation revealed eighteen geomorphic units, with thirteen of them covering 65% of the area, necessitating landfill practices for urban development. These units represent the natural soil conditions or surface geology of Dhaka, with minimal anthropogenic intervention. The study area experiences a tropical climate characterized by hot, wet, and humid conditions, with distinct monsoon seasons being a prominent feature. The rainfall is so high that it ranges from 1942 to 2059 mm (Arefin, 2020). In the study area, the minimum temperature typically reaches 18 °C in January, while the maximum temperature peaks at 29 °C in August.

Fig. 3
figure 3

A geological map showing the surficial lithological and geomorphological units in the Dhaka region (Kamal & Midorikawa, 2006)

Full size image

Notably, the city has exhibited rapid expansion, extending even into low-lying geomorphic units through fill practices since 1960. Dhaka, for instance, has a significant informal waste sector, with an estimated 1.7% of its population working in this sector (Wilson, 2010). The researchers also categorized fill sites into four classes based on the thickness of the fills. The Amin Bazar and Matuail landfill areas, located in the northwest and southeast, respectively, serve as the current dumping sites. The Amin Bazar site is located 7 km away from the city boundary area, and therefore, it is not included in this study. Based on our field observations, SW has been collected from the Kalabagan, Dhanmondi, Banani, Gulshan, Baridhara, and Uttara residential areas by poor private management (Fig. 2a). Local transports, such as rickshaws and vans, collaborating to collect waste from houses to municipal containers is a promising approach to SW management (Fig. 2b-c). A 42.5-megawatt (MW) waste-to-energy (WtE) project was developed at the Amin Bazar under Savar Upazila landfill site in western Dhaka City, adjacent to agricultural land (Fig. 2d), generating large amounts of gas for power production (Toha & Rahman, 2023). Greenhouse gas emissions from the plant are estimated at 205.18–293.87 kilotonnes per year, totaling 5.13–7.35 million tonnes over its operation period. GHGSat (satellite-based and aerial remote sensing technology and emission) data indicates the site emits 35.04 kilotonnes of methane annually (Amin, 2021). However, direct measurements show 19.02 Gigagrams (Gg)/year, while the Land-GEM-V-3.02 (SP1) model estimates 25.95 Gg/year of methane emissions (Toha & Rahman, 2023). The current Matuail landfill areas (south east side of Dhaka city)—neglecting the buffer zones from urban settlements and infrastructures—were designed to manage the dumping process (Fig. 2e). The wastewater canal in the Kallayanpur area, west side of the Dhaka city, is used as a sewerage line, as shown in Fig. 2f.

Datasets and methodology

This study employed diverse data sources to create a comprehensive framework for evaluating potential MSWD sites in Dhaka City. Table 1 outlines the datasets utilized in this study, along with their respective sources. The Dhaka GIS database was built using accessible satellite (Landsat-8 and SRTM) imagery, groundwater well data, and supplementary datasets such as topographic sheets and local administrative maps. All thematic images were co-registered to a common projection (WGS1984, zone 46N) using the ArcMap software. An integrated five-step approach within the GIS environment was implemented to achieve the research aims (Fig. 4), which included the following (Table 2):

Table 1 Data and relevant sources for the study
Full size table
Fig. 4
figure 4

Flow chart for the solid waste landfill site selection

Full size image
Table 2 Relative class rate scale according to Saaty (1980)
Full size table

Thematic mapping and classification of controlling factors

Following the comprehensive literature review, eleven controlling factors were selected for this work: slope, flow accumulation, soil type, lineaments, LULC, distance from roads, distance from settlements, distance from airports, distance from water bodies, depth to groundwater, and wind flow. The score of the thematic layers classes was chosen from 1 to 5 according to their importance (Khan & Samadder, 2015; Mussa & Suryabhagavan, 2019; Santhosh & Babu, 2018). A score of 1 is designated for the category deemed suitable, while a score of 5 is allocated for the class considered comparatively less suitable, as suggested in the literature review (see Table 3).

Table 3 Thematic layers class range and their score and final rate
Full size table

Topographic factors: The shuttle radar topography mission (SRTM)-DEM data were acquired with a spatial resolution of 30 m. The spatial distribution of the surface elevation (Fig. 1c), slope, aspect, and flow accumulation maps were determined using the following steps: (1) a mosaic DEM image was created using ArcMap; (2) sinkholes were corrected by filling their elevation levels with those of their neighbors; (3) a flow direction map was generated from the resulting filled image using the eight-direction algorithm (D8). The flow accumulation raster imputes the cell number of the stream flow path. It also determines the size of the flow path after precipitation. The relationship between runoff and peak discharge can be determined from this raster. In this study, the slope and flow accumulation maps were classified into five classes following the classification scheme proposed by Khan & Samadder (2015), (see Fig. 5).

Fig. 5
figure 5

Topographical and climatological factors of the study area showing (a) Slope, (b) Flow accumulation, and (c) Wind flow direction

Full size image

Climatological factors: Wind direction was obtained from the Bangladesh Meteorological Department (BMD), to assess its impact on odor dispersion from the landfill site. Since prevailing winds can carry odors, minimizing the impact on settlements is crucial. Aspect and wind directions are interconnected with the flow path, as indicated by Şener et al. (2010). Therefore, the resulting aspect map from SRTM-DEM was utilized to enhance the classification of the wind direction map, categorizing it into north (N) to north-east (NE), south-west (SW) to west (W) to north-west (NW), east (E), south-east (SE), and flat to south (S) for the area (Fig. 5c).

Geological factors: The type of soil underlying SW dumpsites can significantly influence how contaminants leach into the underlying aquifer, potentially leading to water quality degradation and associated health risks (Gemail et al., 2017). The official soil map was obtained from the Geological Survey of Bangladesh to identify the soil types in the study area (Fig. 6a). Lineament is a geological structure that lies just below the weathered soil surface and acts as a conduit between surface water bodies and the aquifer (Gemitzi et al., 2007; Mussa & Suryabhagavan, 2019). In this work, the lineament raster map was obtained from Khan & Samadder (2015), which was later classified using multi-buffering tools in ArcMap (Fig. 6b).

Fig. 6
figure 6

Geological factors of the study area showing (a) Lithological maps, and (b) Lineament map

Full size image

Socioeconomic factors: The LULC map was produced from Landsat 8 (OLI) (https://earthexplorer.usgs.gov/, acquired in February 2014) using the Maximum likelihood method. Preprocessing of the Landsat OLI, including radiometric and atmospheric corrections, was performed using the calibration and dark object subtraction tools in ENVI 5.1 software (Chavez, 1988). After that, the LULC map was produced by using the Maximum likelihood method. Six LULC classes were determined, namely vegetation, sand bar, water, open land, buildings, and roads (Fig. 7a). The accuracy of the LULC classification was verified using Google Earth imagery, which stood at 87%. To safeguard urban and economic projects from environmental hazards, landfill sites should be situated at a considerable distance from settlement areas, roads, and airports. All these factors were extracted from the LULC map using the clip option toolbox in ArcMap. Dhaka city boasts two airports: Hazrat Shahjalal International Airport, serving as the main gateway for international passenger flights, and Tejgaon Airport, primarily used by the Bangladesh Air Force for military operations. Consequently, the multi-buffer zones were created for landfill purposes to restrict human movement and development close to the landfill, building upon the approach proposed by Khan & Samadder (2015) for similar contexts. (Fig. 7 b-d).

Fig. 7
figure 7

Socioeconomic factors of the area with ranking: a) LULC map, b) Distance to settlement, c) Distance to road, and (d) Distance to airport

Full size image

Hydrogeological factors: SW releases harmful gases and leaches toxic solutes, posing a double threat to the environment through climate and water degradation (Lim et al., 2016). Groundwater Table (GwT) depth plays a pivotal role in evaluating environmental hazards associated with groundwater contamination (Khalil et al., 2022). The database includes GwT data and detailed lithological information up to 250 m, collected from over 1000 tubewells by the Bangladesh Water Development Board (BWDB, 1991). Five regular interval classes, i.e., 13.6 m to 20.3 m, 20.3 m to 27 m, 27 m to 33.7 m, 33.7 m to 40.4 m, and 40.4 m to 47.1 m, were created using the inverse distance weighting (IDW) interpolation method in ArcMap (Fig. 8b). Therefore, the landfill site must be away from the proximity of the groundwater well and surface water bodies. The water bodies raster map was obtained from the LULC map. Additionally, the distance from water bodies was produced using multi-buffering tools in ArcMap (Khan & Samadder, 2015).

Fig. 8
figure 8

Hydrogeological factors of the area with ranking: a) Surface waterbody and b) Depth of ground water table (GwT)

Full size image

GIS-based multicriteria evaluation techniques

Homogenization and rating of input factors

In GIS modelling, multicriteria evaluation is a systematic process wherein a new criterion is derived by combining multiple criteria through informed judgments (Şener et al., 2010). The input data (criteria) of the same projection governed the MCDM. The homogenization step has become a standard process to build both suitability and susceptibility models (Aly et al., 2005). So, all the input factors were converted to raster layers of 30-m spatial resolution.

The decision is taken according to the input information of the eleven assigned rasters, then rearranged, and manipulation of the raster information cell by cell takes place according to the decision-making rules. In this work, eleven criteria have been evaluated, also known as Multi-Criteria Evaluation. According to Moeinaddini et al. (2010), well-defined criteria are crucial for transforming thematic maps into meaningful units within the MCDM framework. In this study, these criteria will guide the weighting of various topographical, climatological, socioeconomic, and hydrogeological factors in the MSWD site selection process. The related thematic layers were arranged with numeric values between five (as the lowest suitable rate) and one (taken as the highest suitability rate). Later, these criteria were applied to the calculation of the MSWD sites on a GIS platform (Table 3).

Assessing criteria weight using AHP

The analytical hierarchy process is one of the functions of MCDA, where pairwise comparison matrices are used to understand the layer's impotence (Saaty, 1977). The AHP technique helps to evaluate each criterion since multiple important objectives are challenging to prioritize; therefore, it is helpful for the choice of importance. The weight of each layer is fixed using Saaty’s assumption (Saaty, 1977), where a normalized matrix is generated in comparison with the other criteria associated with the relative importance of the layers. Table 2 shows Satty’s assumption that has been used for the layers criterion selection to calculate the WLC weightage.

The imputation of the MCDM procedure using AHP (Saaty, 1980) to detect thematic layers' weight is described below (Table 4):

  1. (a)

    Impute the sum of each column of the pairwise comparison matrix (PWCM) as Eq. 1.

    $$\text{Judgment Matrix}\left(\text{P}\right)=\left[\begin{array}{cccc}P11& P12& ....& P1n\\ P21& P22& ....& P2n\\ ....& ....& ....& ...\\ Pn1& Pn2& ....& Pnn\end{array}\right]$$
    (1)

    Where, Pnn = Indicator of the judgment matrix element.

  2. (b)

    Normalization by reciprocal of each element by dividing the column total as Eq. 2.

    $$\text{Cij}=\frac{{P}_{ij}}{{\sum }_{n=1}^{{n}_{f}}{P}_{n}}$$
    (2)

    Where, Cij = normalized value of ith element, Pij = element of the ith row and jth column of the PWCM.

  3. (c)

    Impute the mean of each row of the normalized matrix to get the expected weight for the thematic layers as Eq. 3 (Table 5).

    $$\text{W}=\frac{\sum_{i=1}^{n}{C}_{i}}{n}$$
    (3)

    where, W = Weight of the row (thematic layers), Ci = normalized value of ith element, n = number of rows.

  4. (d)

    Consistency ratio (CR) was imputed (Saaty, 1980) to justify the consistency using Eq. (4) and Eq. (5).

    $$CI=\frac{{\lambda }_{max}-n}{n-1}$$
    (4)
    $$\text{CR}=\frac{CI}{CR}$$
    (5)
Table 4 Pairwise comparison matrix of the eleven thematic layers
Full size table
Table 5 Normalized matrix and weight of the eleven thematic layers
Full size table

Where, \({\lambda }_{max}\) = Prime eigen value = sum of the products of each vector and column summation, n = criteria, CI = index of consistency and RI = index of random inconsistency.

If the CR value becomes > 0.10, then some comparison value needs to be readjusted so that it becomes < 0.10 (Table 5).

Utilizing WLC for mapping the suitability of MSWD

The GIS mapping needs two constraints, such as rating and weightage for each involved controlling factor based on judgment standards (Pekkan et al., 2015). After assigning weights reflecting their relative importance to establish the suitable landfill area, the scored layers were combined using a WLC approach. WLC is a widely recognized and popular method that has been applied in many urban suitability/susceptibility assessments through a linear function (i.e., Malczewski, 1997; Youssef et al., 2021). The preliminary MSWD map through the WLC was produced by assigning designed weights to all scored factors and evaluating the combined intersection for each class by the arithmetic overlay using Eq. 6 (Fig. 9a).

Fig. 9
figure 9

The map displays (a) the suitability for MSWD sites using the WLC method, classified into five categories and (b) the preliminary Megacity Solid Waste Disposal (MSWDP) sites obtained after removing areas less than 10 hectares

Full size image
$$\text{MSWD}={\sum }_{i=1}^{n}{X}_{i}\times {W}_{i}$$
(6)

Where, MSWD = The Megacity Solid Waste Disposal (MSWD) sites map, Xi = Criterion score, and Wi = weight assigned to each criterion.

Balancing MSWD management: hydro-environmental and health perspectives

Toward a sustainable MSWD suitability map, it is crucial to emphasize effective waste management systems regarding environmental and public health preservations by considering hydro-environmental factors. To this end, the subsequent procedures were utilized:

  1. 1.

    Due to their limited capacity and potential logistical challenges, the preliminary Megacity Solid Waste Disposal (MSWDP) sites smaller than 10 hectares (ha) were excluded from further analysis within the GIS environment (Table 6).

    Table 6 Area of landfill site obtained from the WLC
    Full size table
  2. 2.

    The MSWDP was statistically compared with the constraints (settlement, surface water bodies, and land use patterns) factors (Fig. 10). This analysis aimed to validate the effectiveness of the proposed waste management system in considering these critical physical and environmental factors (Table 7).

  3. 3.

    The AHP calculations were conducted for the land use, water bodies, and settlement area to determine the influence of the selected landfill site on the final selection and ranking (Tables 7, 8, and 9).

  4. 4.

    The optimal Megacity Solid Waste Disposal (MSWDO) site map of Dhaka city was obtained by multiple ranks of the MSWDP sites with the AHP calculation weight of the land use, settlement, and water (Table 9). The resulting MSWDO is classified based on the degree of suitability (Fig. 11), with the lowest value, equal to 1, assigned as the suitable site for landfill.

  5. 5.

    Afterwards, a more in-depth qualitative analysis was performed between the promising MSWDO sites and three-dimensional (3D) hydro-stratigraphic models. The 3D hydro-stratigraphic units were evaluated — neglecting field observations and quantitative analysis — to investigate the contribution of clay aquitard thickness and depth to the MSWDO sites. The 3D model was generated from borehole lithological data using 3D gridding Voxel tools, applying the lithoblending algorithm theory in the Rockworks package (Fig. 12a-c). The thickness of the clay layer was also created using the IDW interpolation technique (Fig. 12d-e)

Fig. 10
figure 10

The percentage area of Land Use and Land Cover (LULC) classes over the fifteen preliminary landfill sites

Full size image
Table 7 Pair-wise comparison for land use, settlement and water
Full size table
Table 8 Normalized matrix from the pairwise comparison matrix
Full size table
Table 9 Shows the comparison of the fifteen sites concerning land use, settlement and water and the rank
Full size table
Fig. 11
figure 11

The optimal future Megacity Solid Waste Disposal (MSWDOF) sites map of Dhaka city

Full size image
Fig. 12
figure 12

a 3D view of Hydro-stratigraphic model of the study area, b 3D view of weathered and aquitard (Clay) zone thickness, c 3D view of weathered and Clay zone with thickness less than 5 m, d 2D view of Inverse Distance Weighting (IDW) interpolated clay thickness map, and (e) With the existing location of tubewell as shown in black color symbol

Full size image

Results and discussion

The current study has addressed a critical gap in the selection of effective solid waste disposal sites using GIS and MCDM for efficient management purposes in the Dhaka megacity. The following sections demonstrate the successful application of the GIS-based AHP and statistical approach to determine appropriate locations for MSWD from various and large spatial information in the delta megacities.

Characterization of controlling factors for MSWD

Eleven factors were used to select optimal urban landfill sites in the current study. These layers are associated with geomorphological, geological, hydrogeological, climatological, and socio-economic aspects. Table 3 presents thematic layers with their respective classes, criteria, and standardized scores. Table 4 displays the PWCM for the eleven thematic layers, with criteria selection based on Saaty's assumption (refer to Table 2). The results of thematic controlling factors were described in the following section:

Topographical factors

Topography defines the local and regional relief that provides an idea of groundwater flow (Gintamo, 2015). The sloping area of more than 20° was excluded (Şener et al., 2010), while the slope surface (8° − 12°) was found to be suitable for dump sites (Khan & Samadder, 2015; Lin & Kao, 1998). In this study, the slope area between 8° and 12° was defined as more suitable for the landfill site than the steeper (> 12°) slope that permits high runoff and impurities that are mobile for longer distances from the source area (Mussa & Suryabhagavan, 2019). Thus, the score and rating for the slope map classes are as follows: 4 with a rate of 0.792 for 0°- 4°, 3 with a rate of 0.594 for > 4°- 8°, 1 with a rate of 0.198 for > 8°- 12°, 2 with a rate of 0.396 for > 12°—16°, and 5 with a rate of 0.99 for > 16°. Table 5 shows the normalized matrix and weight (0.198) of the slope map (Fig. 5a).

Concerning water flow, zones with high flow accumulation pose a risk of water pollution by solids and waste (Khan & Samadder, 2015). So, the large cell number zone is not suitable for MSWD, whereas the low cell number zone is highly suitable. In this study, the flow accumulation scores and rates were delineated across different ranges: 0–600 cells (rate 0.034), 600–2500 cells (rate 0.068), 2500–6000 cells (rate 0.102), 6000–15000 cells (rate 0.136), and 15,000–318867 cells (rate 0.17) (Fig. 5b and Table 3). Table 5 shows the weight (0.034) of this raster. Figure 4h shows the flow accumulation map for the proposed area.

Climatic factors

Solid waste creates an odour that spreads with the wind direction. In the study area, the dominant wind direction is from the South (Fig. 5c). The hillsides facing N and NE directions experience limited wind flow throughout the year and are assigned a score of 1 (refer to Table 3). Aspects aligned with the prevailing wind direction throughout the year receive a higher score of 5. Therefore, city dwellers will be affected by the wind flow over the landfill site. Table 5 shows the weight (0.047) of the aspect map.

Geological factors

Lithology plays a crucial role in selecting suitable waste dumping sites, as highlighted by Rahmat et al. (2016). Surface lithology includes Madhapur clay residuum, marsh clay and peat, alluvial silt, and clay (Fig. 6a). Hazardous waste disposal in porous and non-cohesive soils, i.e., clay residuum and peat, is particularly dangerous due to the elevated risk of contaminant leaching into aquifer, potentially impacting groundwater quality (Abu Salem et al., 2021; Alfy et al., 2010). In contrast, clay soil offers a natural barrier against contaminant leaching, significantly minimizing the risk of groundwater contamination (Abu Salem et al., 2021; Gemail et al., 2017). Consequently, Madhupur clay is considered suitable for waste disposal and assigned a score of 1 (Fig. 6a), whereas alluvial silt is considered less suitable and assigned a score of 4. Notably, Madhapur clay residue covers 38% (116 km2, rate 0.014), marsh clay and peat occupy 24% (73 km2, rate 0.028), alluvial silt and clay span 29% (89 km2, rate 0.042), and alluvial silt encompasses 9% (28 km2, rate 0.056). The weight of the soil type was 0.014 obtained from PWCM, as shown in Table 5.

Concerning geological structures, it is advisable to construct landfill sites away from areas susceptible to endogenic disturbances in the Earth, including faults, fissures, and joints (Mallick, 2021). According to Demesouka et al., (2014), a SW site needs to be selected at least 100–200 m apart from any weak zones (i.e., joints and faults). For the current study, the lineament raster was classified using 100 m multi-buffering that continued up to 500 m (Khan & Samadder, 2015). For MSWD, a distance of 500 m is considered suitable and assigned the value 1, whereas areas closer to 100 m are deemed less appropriate and assigned the value 5. The lineament map is classified into five regular interval classes (Fig. 6b), with scores and rates as follows: 5 with a rate of 0.31 for 0–100 m, 4 with a rate of 0.248 for 100–200 m, 3 with a rate of 0.186 for 200–300 m, 2 with a rate of 0.124 for 300–400 m, and 1 with a rate of 0.062 for distances exceeding 400 m (Table 3). The weight is obtained at 0.062 for the lineament map, as shown in Table 5.

Socioeconomic factors

Understanding the distribution of surface water bodies (lakes and rivers), vegetation, and other covers through LULC analysis is crucial for assessing the potential impact of SW disposal on water quality and aquatic ecosystems (Ibrahim-Bathis & Ahmed, 2016). Restricted areas, such as well points, rivers/streams, roads, built-up areas, and forests, should be avoided for SW dumping. On the other hand, bare lands are considered the most suitable option for SW disposal (Mussa & Suryabhagavan, 2019). The LULC map comprises vegetation covering 6% (20 Km2) area, sand bar covering 2% (6 Km2) area, water covering 6% (17 Km2) area, open land covering (barren land) 47% (145 Km2) area, buildings covering 33% (102 Km2) area, and roads covering 5% (16 Km2) area (Fig. 7a). Open land with scrub vegetation received a high suitability score (1) and a weight of 0.265 due to its availability for waste disposal. Conversely, restricted areas (settlements, roads, water bodies, and sand bars) were excluded (score 3) to prevent water pollution and aquifer infiltration. The combined rate of these classes is 0.795. The weight of the LULC was 0.265 obtained from PWCM, as shown in Table 5.

Analysis of the LULC data revealed a remarkably high population density of over 41,000 people per Km2 within the built-up (33% of the region) areas. This density is comparable to, if not exceeding, major metropolises globally, potentially placing it among the top contenders for the highest density based on readily available data (Demographia, 2019; Razia et al., 2023). Previous research shows that the landfill area should not be established within 1 km of the settlement area (Şener et al., 2010). Demesouka et al. (2014) also argued that the 500 m distance is suitable for landfill purposes. In the context of the settlement map, a 200 m multi-buffering process was applied, extending up to 800 m (Khan & Samadder, 2015). Figure 7b shows the distance from the settlement map and their classes. The distance from the settlement map was categorized into five classes with associated scores and rates as follows: 0–200 m (class 5, rate 0.545), 200–400 m (class 4, rate 0.436), 400–600 m (class 3, rate 0.327), 600–800 m (class 2, rate 0.218), and > 800 m (class 1, rate 0.109). PWCM provides a weight of 0.109 for this layer (Table 5).

To prevent adverse effects on public movement, Mussa and Suryabhagavan (2019) recommended a distance of over 4900 m from roads as suitable for landfill dumping. However, it is important to consider placing the landfill site closer to the road network, ideally within 200 m, to optimize cost-effectiveness (Şener et al., 2010). Herein, 100 m multi-buffer zones were created up to 500 m distance for SW purposes to avoid adverse effects on human movement. The road network map is classified, each assigned a score, rate, and distance as follows: 5 with a rate of 0.125 for 100 m, 4 with a rate of 0.1 for 200 m, 3 with a rate of 0.075 for 300 m, 2 with a rate of 0.05 for 400 m, and 1 with a rate of 0.025 for 500 m, as outlined in Fig. 7c and Table 3. The obtained weight for the road network map was 0.025 using PWCM (Table 5).

A constraint landfill buffer, ranging from 200 to 1000 m in five classes, was considered in this study (Khan & Samadder, 2015). This measure was taken to prevent the presence of birds in the vicinity, ensuring that it does not pose any obstacle to airborne vehicles (Moeinaddini et al., 2010). The distance up to 1 km from the airport is deemed more suitable and is assigned a score of 1. The distance to the airport map was classified into five distance classes with associated rates, as shown in Table 3: 0–200 m (rate 0.09), 200–400 m (rate 0.072), 400–600 m (rate 0.054), 600–800 m (rate 0.036), and 800–1000 m (rate 0.018). This layer was assigned a weight of 0.018 (Table 5).

Hydrogeological factors

Understanding the hydrogeological setting, including the depth of the aquifer, soil permeability, and distribution of water bodies, is crucial for predicting the movement of pollutants from landfills (Mekuria, 2006). The prominent water features for Dhaka city and its adjacent areas include rivers (Turag, Balu, Lakkya, and Burigonga), lakes (Gulshan, HatirJheel, and Dhanmondi), and ponds. The buffer zone of unsuitable sites from water bodies is classified into different scores and distance intervals: score 1 for 0–100 m (rate of 0.735), score 2 for 100–200 m (rate of 0.588), score 3 for 200–300 m (rate of 0.441), score 4 for 300–400 m (rate of 0.294), and > 400 m (rate of 0.147). A distance of 500 m from water bodies is considered suitable for SW site selection and is assigned a score of 1. The obtained weight was 0.147 for this layer, as shown in Table 5.

Figure 8b shows the GwT depth and groundwater pumping well map for the study area. The GwT depth ranges from 13.6 m in the northwestern part to 47.1 m in the middle parts. Substantial groundwater extraction of about 2.0 Mm3/day groundwater was observed from the Dupi-Tila formation of the upper Plio-Pleistocene age, where 563 deep tube wells belonged to BWDB and 2000 private tube wells were assigned for this purpose (Arefin & Alam, 2020; Arefin, 2020). Hence, MSWD sites should be away from the existing groundwater wells in the western and middle parts (Khan & Samadder, 2015; Mussa & Suryabhagavan, 2019). The scores, rates, and depth intervals are as follows: 5 with a rate of 0.41 for 13.6–20.3 m, 5 with a rate of 0.328 for > 20.3–27 m, 5 with a rate of 0.246 for > 27–33.7 m, 5 with a rate of 0.164 for > 33.7–40.4 m, and 5 with a rate of 0.082 for > 40.4–47.1 m (Table 3). Table 5 shows the weight (0.082) of the GwT depth map.

Megacity solid waste disposal (MSWD) site suitability mapping

To identify the MSWD site, the AHP was applied to the selected thematic layers to determine their importance. In the AHP calculation, weights assigned to each thematic layer were carefully determined considering their contribution to both environmental suitability and potential risks to human health. Experts in waste management and public health informed the weighting process, ensuring a balanced approach that prioritizes both sustainable practices and community well-being. PWCM was conducted to produce the weight after assigning the rational environmental criteria, as shown in Table 4. Table 5 shows the normalized matrix where the value of each column drives the ratio of the column value divided by the corresponding column total. Finally, the weights are obtained by taking the average value of the row for each layer. The whole procedure has been narrated in the methodology section. The CR value for each AHP was kept below 0.1 for suitable weight assignment. According to the weightage of AHP model, the LULC has the highest value (0.265), followed in order of importance by SL (0.198), DW (0.147), DS (0.109), GwT (0.082), DL (0.062), WF (0.047), FA (0.034), DR (0.025), and DA (0.018), whereas the ST has the lowest weight value (0.014). Finally, the WLC method was applied to aggregate the selected thematic layer and generate potential MSWD sites using ArcGIS.

The area identified by WLC was classified into five suitability classes from very highly suitable (1) to unsuitable (5), as shown in Fig. 5a. The highly suitable area of class 2 is negligible; therefore, class 1 areas were taken into consideration for MSWDP purposes. Additionally, the SW disposal sites with an area exceeding 10 hectares (ha) were deemed suitable for further analysis (Fig. 9b), considering the projected population growth for the next 20 years as suggested by Khan & Samadder (2015). As a result, areas less than 10 hectares were excluded, as indicated in Table 6. Figure 9b shows fifteen sites obtained from the WLC. Areas of the generated fifteen sites are given in Table 6. The results show that the 1st, 4th, 6th, 12th, and 13th sites encompass considerable areas of 2175, 157, 195, 2255 and 106 ha, respectively.

Potential landfills considering hydro-environmental conditions

Developing a sustainable MSWD suitability map requires prioritizing hydro-environmental factors to ensure effective waste management and protect public health and the environment. Understanding the hydrogeological setting, including aquifer depth and water body distribution, is crucial to avoid dumping SW into main water resources (Datta et al., 2024; Eleraki et al., 2010). Initially, fifteen SW disposal sites were statistically compared with the LULC pattern using a bar diagram (Fig. 10). The AHP of the three layers —land use, settlement, and surface water bodies—was calculated (Fig. 3). Table 7 presents the PWCM. Table 8 explains the normalized matrix and rate of these three layers. Table 9 outlines the ranking of the fifteen landfill sites concerning their suitability for land use, distance from settlements, and water bodies. Final rankings for the fifteen identified sites were determined based on their suitability concerning land use, distance from settlements, and proximity to water bodies, using the AHP-produced rates. The analysis indicates that sites 4, 9, and 15 are distant from land use classes, whereas sites 10, 13, and 14 fall within these critical classes. Therefore, site 15 is ranked 1, site 9 is ranked 2, and site 4 is ranked 3. In contrast, sites 10, 13, and 14 are ranked 13, 14, and 15, respectively. Despite its proximity to site 6, the Amin Bazar area is 7 km from the city boundary. Sites 1, 4, and 12 are more securely distanced from settlement areas and are ranked 1, 3, and 2, respectively, while sites 2, 3, and 9 are closer to settlements and ranked 13, 14, and 15, respectively. Sites 1, 2, and 12 are suitably distanced from water bodies, earning rankings of 1, 3, and 2, respectively, whereas sites 4, 10, and 14 are less suitably located and ranked 13, 15, and 14, respectively.

Secondly, a 3D hydro-stratigraphic model was employed to examine soil discontinuities and groundwater conditions (Royse et al., 2009). The model shows a heterogeneous hydrological condition of different aquifers (Fig. 12a), and the soil units are interdigitated rather than interlayered. Moreover, the depth of GwT ranges from 13.6 to 20.3 m in the northern areas (Fig. 8a), as the clay thickness decreases between 1 and 5.43 m (Fig. 12d-e). These conditions can facilitate contaminants transportation, increasing the risk of groundwater pollution (Gemail et al., 2017; Khalil et al., 2022). This observation also aligns perfectly with the statistical findings, which revealed that northern sites, including 1st to 11th, have a significantly higher vulnerability to environmental risks. Moreover, the resulting MSWDOF map is deemed to be validated against the existing landfill areas in the southeastern and southeastern parts (Figs. 1c and 11). Consequently, the 12th site (Zone A) and the 15th site (Zone B) were identified as relatively optimal future selections for waste disposal sites (Fig. 11). Zone B is located close to the existing Matuail landfill area. The suitable Zones A and B sites cover 2255 and 30 ha, respectively.

In line with De Feo and De Gisi (2014), selecting suitable landfill sites becomes especially critical in developing countries due to weak environmental regulations, rapid urbanization, and scarcity of appropriate land. This study demonstrates the superior efficiency of integrating a 3D hydro-stratigraphic model with MSWD mapping compared to solely relying on information from a GIS-based AHP model for comprehensive site suitability assessment in complex megacity environments. The integration of land use analysis, statistical methods, and subsurface assessments has led to the identification of more suitable landfill sites compared to traditional models. This comprehensive approach significantly enhances decision-making for waste disposal, facilitating access to clean water (SDG 6). This study provides urban planners with existing and potential future waste disposal locations in Dhaka through the utilization of a remote sensing and GIS-AHP framework, coupled by land-use and hydro-environmental analyses. This structured approach bridges the gap between researchers and policymakers, facilitating the implementation of effective waste management practices that align with climate change mitigation policies (SDGs 11 and 13) in Bangladesh (Martin et al., 2013). By following this framework, Dhaka North and South city corporations and other stakeholders can make more informed decisions regarding future landfill site selection, thereby promoting sustainable waste management and environmental protection. Waste-to-energy projects can generate employment opportunities and minimize landfilled waste, promoting sustainable land management for agricultural activities, aligning with SDG 15 (Mohod & Bagal, 2023; Nasr, 2024). This highlights the importance of moving beyond single-method approaches and embracing comprehensive strategies for sustainable landfill development in challenging delta megacities.

Conclusions

Selecting suitable solid waste disposal sites in densely populated deltas of developing countries is challenging due to limited public awareness about waste management. To address this, a structured approach was used to select MSWD sites in Dhaka, minimizing environmental risks. This study employs an AHP-integrated GIS approach to synthesize geomorphological, geological, socioeconomic, hydrogeological, and climatological factors, combined with land-use and hydro-environmental analyses, for mapping MSWDOF sites. Eleven thematic layers, rated from 1 to 5 based on suitability, were weighted using PWCM to generate the MSWD map via the WLC method. The analysis revealed LULC, SL, DW, and DS significantly influenced SW disposal site suitability, while slope ST had the least impact. The MSWD map showed five suitability classes from very highly suitable (1) to unsuitable (5). Due to limited highly suitable areas (class 2), focus shifted to class 1 areas, identifying 15 potential MSWDP sites, each over 10 ha. Integrating statistical land-use suitability analysis with AHP-based ratings for land use, distance to settlements, and water bodies produced an MSWDOF map identifying two suitable waste disposal sites. The 3D hydro-stratigraphic model confirmed northern sites' unsuitability due to shallow groundwater (< 5.43 m) and thin clay, excluding 11 options. Thus, sites 12 (Zone A, 2255 ha) and 15 (Zone B, 30 ha), with deeper groundwater tables and thicker clay layers, were identified as optimal for future waste disposal. To this end, this study significantly impacts sustainable waste disposal planning in Southeast Asian delta megacities, especially those experiencing rapid urbanization and rising sea levels. By offering crucial insights into the interplay between geospatial factors and site suitability, our comprehensive approach enables planners to make informed decisions that minimize environmental impact and transportation costs. The valuable application of GIS and remote sensing in analyzing land use patterns and subsurface conditions demonstrates the potential of these tools for large-scale waste management. Prioritizing the investigation of population growth patterns, extreme weather events, and economic considerations will enhance the accuracy of waste disposal modeling and ensure long-term sustainability for delta megacities.