The proposed decision framework adopts a deliberately modular structure in which criterion weighting, spatial

level using the Technique for Order Preference by Similarity to Ideal Solution, where candidate locations are

assessed as discrete, implementable planning entities. To safeguard the integrity of the results, sensitivity testing

is intentionally restricted to the final ranking stage, preventing distortion of the spatial evaluation process. The

framework is demonstrated using municipality-scale data, confirming its potential as a transparent, replicable,

and technology-oriented decision support approach for solid waste management planning. A case-based

implementation using municipal-scale spatial data is included to demonstrate the practical applicability and

robustness of the proposed framework.

Keywords: Decision Support System, Soft Computing, Fuzzy AHP, GIS-Based Analysis, TOPSIS, Weighted

Linear Combination, Solid Waste Management, Sensitivity Analysis

transparent and justifiable planning outcomes. In this regard, the coupling of Geographic Information Systems

(GIS) with algorithm-driven decision models provides a coherent computational environment for systematically

managing spatial constraints, diverse evaluation criteria, and stakeholder preferences.

Although GIS-based multi-criteria decision analysis has been widely applied to landfill site selection, many

existing studies continue to rely on crisp weighting methods and raster-level optimization, limiting their ability

to represent uncertainty and realistic decision processes [2]. Conventional implementations of the Analytic

Hierarchy Process (AHP) assume precise pairwise judgments, even though expert evaluations in environmental

planning are often vague and linguistically expressed. Moreover, spatial decision models frequently treat pixel-

level suitability values as final decisions, overlooking the distinction between spatial assessment and

implementable planning alternatives. These simplifications reduce the effectiveness of GIS-based decision

explicit ecological constraints, while final prioritization is conducted at the DSS level using the Technique for

Order Preference by Similarity to Ideal Solution (TOPSIS), where alternatives are assessed as implementable

planning units instead of raster cells. This strict separation of weighting, aggregation, and ranking improves

transparency, avoids methodological ambiguity, and conforms to established decision support system design

principles. Additionally, a case-based implementation is presented to validate the operational applicability of the

proposed framework in a real-world municipal context.

Related Work and Research Gaps

GIS-based landfill site selection has been extensively studied using multi-criteria decision analysis to integrate

spatial constraints with planning objectives. Early computational frameworks largely relied on overlay-based

suitability mapping, combining environmental and infrastructural factors through weighted aggregation within

decision-making models for spatial planning applications. In particular, fuzzy extensions of the Analytic

Hierarchy Process have been used to capture linguistic judgments and partial preferences in criteria weighting

for landfill site selection. Although these approaches improve uncertainty representation at the weighting stage,

many studies still embed FAHP within spatial aggregation or treat fuzzy raster-based suitability outputs as final

decisions. This practice obscures the separation between preference modelling and decision execution, thereby

reducing the effectiveness of such approaches as comprehensive decision support systems from a computational

perspective [5].

Beyond fuzzy weighting methods, several studies have applied ranking-based multi-criteria techniques, such as

the Technique for Order Preference by Similarity to Ideal Solution, to prioritize landfill sites and other

infrastructure alternatives. These methods are typically implemented after spatial suitability mapping to rank

The proposed decision support framework adopts a modular, multi-stage computational architecture that

integrates geospatial analysis with soft computing–based decision modeling. It follows a sequential workflow in

which data preparation, criteria weighting, spatial aggregation, alternative-level ranking, and robustness

validation are implemented as independent yet interoperable components. From a computer science perspective,

this modular design improves transparency, reproducibility, and scalability while avoiding functional overlap

between analytical stages. By clearly separating uncertainty modeling, spatial evaluation, and decision

execution, the framework ensures that each method operates within its defined role, supporting reliable and

interpretable decision outcomes in complex planning contexts [7].

In the initial stage of the framework, spatial datasets representing environmental, infrastructural, and

prevents bias propagation and ensures that subsequent stages operate on validated, analysis-ready inputs rather

than inconsistently processed data.

As illustrated in Fig. 1, decision modeling begins with a dedicated criteria weighting stage implemented using

the Fuzzy Analytic Hierarchy Process (FAHP). This stage is intentionally separated from spatial aggregation

Algorithmic Workflow of the Proposed Framework

Step 1: GIS-based preparation and standardization of spatial datasets

Step 2: Derivation of criteria weights using FAHP under uncertainty

Step 3: Spatial aggregation using Weighted Linear Combination (WLC)

Step 4: Extraction of feasible decision alternatives from suitability surface

Step 5: Ranking of alternatives using TOPSIS

Step 6: Post-decision sensitivity analysis for robustness validation

This structured workflow ensures strict separation between analytical stages and improves transparency,

reproducibility, and computational efficiency.

weights to WLC-based aggregation, the framework preserves a clear separation between preference modeling

and decision ranking, thereby avoiding methodological overlap and ensuring computational clarity.

Final decision-making in the proposed framework is handled by a TOPSIS-based decision support layer that

operates exclusively on spatially aggregated outputs rather than individual raster cells. At this stage, the

continuous suitability surface produced through WLC is translated into representative measures for a finite set

of decision alternatives, thereby shifting the analysis from spatial evaluation to implementable planning

decisions. TOPSIS ranks these alternatives based on their relative closeness to ideal and anti-ideal solutions,

providing a transparent decision logic. By restricting TOPSIS to alternative-level ranking and excluding it from

spatial aggregation or weight derivation, the framework maintains a clear separation between evaluation and

decision execution, a key requirement of robust decision support system design.

methodological integrity, reduces computational redundancy, and provides transparent evidence of decision

stability, thereby strengthening the reliability of the proposed DSS in complex planning contexts.

The criteria weighting component of the proposed framework is implemented using the Fuzzy Analytic

Hierarchy Process (FAHP) to explicitly capture uncertainty and subjectivity in expert judgment. In complex

spatial planning contexts, decision-makers often express preferences linguistically, which cannot be adequately

represented by crisp weighting methods. FAHP overcomes this limitation by extending classical AHP with fuzzy

set theory, allowing pairwise comparisons to be modeled using fuzzy numbers rather than precise ratios. Within

the framework, FAHP is applied exclusively to derive relative criterion weights and is not involved in spatial

aggregation or alternative ranking. This strict confinement localizes uncertainty handling to preference

elicitation and preserves the interpretability and modularity of subsequent computational stages [8].

approach captures relative importance as value ranges rather than precise ratios, reflecting the inherent vagueness

of human judgment in environmental and infrastructure planning. TFNs enable efficient arithmetic operations

compatibility with subsequent computational stages, the fuzzy weights are defuzzified into crisp values while

retaining the uncertainty captured during preference elicitation. The resulting normalized weights, satisfying the

unity condition (Σw = 1), are then transferred exclusively to the spatial aggregation module, completing the

FAHP-based weighting process within the proposed framework [10].

this stage, as such operations can introduce subjective bias and unnecessarily restrict decision flexibility.

Retaining the surface in continuous form preserves spatial detail and enables subsequent extraction of

based prioritization methods, TOPSIS is applied solely to a finite set of decision alternatives derived from spatial

aggregation, ensuring that ranking is performed on implementable planning units. This abstraction enables a

clear transition from continuous spatial evaluation to discrete decision analysis required for practical planning

and policy formulation. By restricting TOPSIS to post-aggregation ranking, the framework maintains

methodological clarity and avoids conflating spatial suitability modeling with decision execution [16].