ABSTRACT
With the intensification of urbanization and the expansion of urban heat islands in Tehran, green roofs have become an important sustainable infrastructure for improving environmental quality. The principal challenge to the long-term performance of these systems is the selection of plant species that are compatible with roof structural constraints, the semi-arid climate, and severe temperature fluctuations. This study develops a multi-criteria decision-making model for selecting appropriate green-roof plant species in Tehran. Evaluation indicators were first screened and finalized using the Fuzzy Delphi Method with a seven-member expert panel. The criteria were then weighted using the Analytic Hierarchy Process. The results showed that low water requirement and ecological resilience were the most influential considerations. Ten plant species were evaluated using the final criteria. Festuca ovina obtained the highest final score and ranked first, followed by Lavandula angustifolia and Rosmarinus officinalis. In contrast, Rhododendron simsii and Chamaecyparis lawsoniana received the lowest scores because of higher water demand and lower compatibility with the roof microclimate. The findings indicate that prioritizing ecological sustainability over purely aesthetic criteria can improve the persistence and manageability of green roofs in Tehran and provide a practical decision-support framework for designers and urban managers.

EXTENDED ABSTRACT
Introduction
Urbanization and climate change have increased environmental pressure on metropolitan ecosystems, contributing to urban heat islands, biodiversity loss and reduced ecological resilience. Green roofs provide additional vegetation on otherwise unused roof surfaces and can support thermal moderation, stormwater retention, pollution reduction and urban biodiversity. Their performance, however, depends strongly on the persistence of the selected vegetation. Roof environments in Tehran are particularly stressful because of shallow substrates, restricted water and nutrient availability, intense solar radiation, strong winds and pronounced seasonal temperature fluctuations. Field observations reported by the authors indicate that plant selection in many Tehran green-roof projects is still strongly influenced by appearance, while ecological compatibility receives less attention. This can result in plant loss, repeated replacement and increased irrigation and maintenance requirements. Accordingly, the study aimed to develop a practical, locally applicable model that integrates ecological, biological, aesthetic and economic considerations for selecting sustainable green-roof vegetation in Tehran.
Materials and Methods
The study used a combined Fuzzy Delphi Method (FDM) and Analytic Hierarchy Process (AHP). Candidate criteria and plant species were identified through review of the scientific literature and field investigation of implemented green-roof projects in Tehran. Ten plant species were selected for evaluation, including Festuca ovina, Lavandula angustifolia, Rosmarinus officinalis, Lippia nodiflora, Pennisetum sp., Pittosporum sp., Ligustrum japonicum, Sedum sp., Rhododendron simsii and Chamaecyparis lawsoniana. A panel of seven experts representing horticulture, landscape architecture, urban ecology, green-roof implementation, botany, municipal green-space management and structural engineering participated in the evaluation. Expert judgments were represented by triangular fuzzy numbers and defuzzified using a centroid approach. Indicators above the stated threshold of 0.6 were retained. AHP pairwise comparisons were then used to determine the relative importance of the final criteria. The reported consistency ratio was 0.06.
Results and Discussion
The findings showed that ecological considerations were the most influential factors in selecting plant species for green roofs in Tehran. Among the major groups of criteria, ecological characteristics, particularly low water requirement and resistance to environmental stresses, received the highest final weight of 0.45. This result reflects the importance of water limitation, drought tolerance, and climatic adaptation in the semi-arid environment of Tehran. Biological sustainability, including characteristics associated with plant longevity, growth, and persistence, ranked second with a weight of 0.25. Aesthetic and visual characteristics received a weight of 0.18, while economic criteria related to plant purchase and maintenance accounted for a weight of 0.12.
Ecological characteristics, particularly low water requirement and environmental resistance, received the highest final weight (0.45). Biological sustainability ranked second (0.25), aesthetic and visual characteristics third (0.18), and economic criteria fourth (0.12). Festuca ovina achieved the highest final score (0.942), followed by Lavandula angustifolia (0.885) and Rosmarinus officinalis (0.820). Lippia nodiflora and Pennisetum sp. occupied intermediate positions. Rhododendron simsii and Chamaecyparis lawsoniana were among the lowest-ranked alternatives because of greater maintenance requirements and lower compatibility with Tehran roof conditions. The findings show that species able to tolerate drought, temperature fluctuation, shallow substrates and low-input management provide a stronger basis for long-term green-roof performance than choices driven mainly by ornamental value.
Conclusion
The present study developed a locally oriented multi-criteria decision-support model for selecting plant species for green roofs in Tehran using a combined Fuzzy Delphi Method and Analytic Hierarchy Process. The results demonstrated that ecological considerations, particularly low water demand and resistance to environmental stresses, should have the greatest influence on plant selection under Tehran’s semi-arid climatic conditions. Biological sustainability was the second most important criterion, whereas aesthetic and economic considerations, although relevant, received lower relative weights.
Among the ten evaluated plant species, Festuca ovina, Lavandula angustifolia, and Rosmarinus officinalis were identified as the most suitable alternatives. Their strong performance was associated with relatively low water requirements, tolerance to environmental stresses, adaptability to rooftop conditions, and manageable maintenance needs. In contrast, Rhododendron simsii and Chamaecyparis lawsoniana showed lower suitability because of their greater ecological and management requirements.
The proposed model can provide a practical scientific framework for landscape designers, green-roof contractors, municipal green-space managers, architects, and urban decision-makers. Applying such a framework can help reduce the risk of inappropriate plant selection, excessive irrigation, repeated replacement, and unnecessary maintenance costs. It may also support the development of local standards and guidelines for sustainable green-roof design in Tehran.
The framework is potentially transferable to other cities, provided that the criteria and their weights are recalibrated according to local climatic conditions, structural limitations, available water resources, management practices, and regional plant characteristics. Future research should complement the expert-based model with long-term field experiments examining actual plant survival, water consumption, substrate performance, seasonal responses, and interactions among species. Evaluating user preferences and psychological responses to different plant combinations may also add a social dimension to the framework. Integrating these additional forms of evidence can further improve the accuracy of plant-selection models and contribute to the development of resilient, resource-efficient, and environmentally sustainable green infrastructure in Iranian cities.