Metropolitan nature becomes analytically useful when it stops functioning as a generic synonym for green space and begins to reveal the unequal spatial systems through which climate, access, governance and ecological form are produced. The four studies assembled here—Lagos, Cape Town, Addis Ababa and Tunis—describe very different metropolitan conditions, yet together they demonstrate why urban nature cannot be understood through quantity alone.
In Lagos, Obe, Morakinyo and Mills construct heat risk as the interaction of hazard, exposure and vulnerability rather than as temperature by itself. Their method combines the urbanised Weather Research and Forecasting model, Humidex-based heat stress, open-source demographic and geospatial datasets, Local Climate Zones, Moran’s I and Optimized Hot Spot Analysis. The resulting map identifies approximately 423 km² of Critical Heat Risk Zones at the highest confidence level, with much of that risk concentrated in densely urbanised LCZ3, LCZ6 and LCZ7 and overlapping Makoko, Ajegunle and parts of Alimosho. The ecological meaning of the city is therefore inseparable from density, vegetation, wealth, age, living conditions, water, electricity, drainage and infrastructural capacity. Heat is not a meteorological layer placed on top of society; it is a metropolitan relation produced by urban form and differentiated adaptive capacity. The study is particularly valuable because it turns limited data availability—a recurrent problem in rapidly urbanising African cities—into a methodological question rather than a reason to exclude the city from analysis. Numerical modelling and open-source evidence do not erase uncertainty, and the authors explicitly recognise biases in model performance and socioeconomic data resolution, but their framework shows how urban environmental knowledge can still become spatially actionable. Cape Town makes an apparently different problem structurally comparable. Milliken’s study of eight nature reserves and conservation areas begins with a city famous for exceptional biodiversity but asks a more difficult question: who can actually reach and use that biodiversity? Across 672 park users and 317 members of the surrounding public, access is shaped by proximity, private-car dependence, public-transport scarcity, safety, information and Cape Town’s inherited spatial structure. Only three surveyed park users arrived by public transport, most visitors lived relatively close to the spaces they used, and more than a quarter of non-users identified lack of information or lack of awareness of what a reserve offered as a reason for not visiting. Biodiversity can therefore be geographically present and socially remote. Environmental value is mediated by transport networks, neighbourhood geography, fear, knowledge and the historical distribution of populations across the metropolis. This is crucial for a comparative urban-nature project because it displaces the assumption that ecological provision automatically produces ecological access. A protected landscape may succeed as habitat while remaining weak as metropolitan public infrastructure. Addis Ababa shifts the problem again, from access to the institutional production of green-space durability. Ayele, Megento and Yeshitela examine urban parks and community green spaces through management models and governance principles, identifying generally poor park governance alongside an emerging tendency toward user-centred community management. Their evidence situates depletion within rapid urbanisation, transfer of land allocated as green space to private uses, weak sectoral cooperation and institutional fragmentation. Yet the paper avoids reducing this condition to institutional failure. Community-managed spaces indicate that alternative governance capacities already exist within the city and can become part of long-term design and management. The distinction is important: urban nature is not only planned or destroyed; it is continuously reproduced through strategic, tactical and operational levels of authority. Governance determines whether a green space is maintained, appropriated, reclassified, protected or gradually lost. Tunis then introduces a fourth register: hydrological morphology and design knowledge. Mili’s research on “instant cities” proposed for Tunisian wet coastal zones begins from the observation that large real-estate developments were repeatedly targeted at sebkhas—saline, periodically flooded depressions that store water, filter runoff, buffer coastal erosion, sustain biological production, host birds and support halophytic vegetation. The design problem is therefore not how to insert greenery into a new city but whether the existing wet landscape is recognised as infrastructure before development begins. The paper’s pedagogical and architectural method is striking: students were asked to treat geology, geomorphology and ecological processes as primary project information, and to accept that the correct design conclusion might be to build differently, build less or not build at all. In this formulation, the ecological site is not an empty substrate awaiting urban form. It already performs metropolitan work. Across Lagos, Cape Town, Addis Ababa and Tunis, metropolitan nature emerges not as a bounded object but as a relation among morphology, climate, access, governance, infrastructure and knowledge. Each city changes the meaning of the shared category: Lagos turns nature into thermal and socioeconomic risk modulation; Cape Town turns it into a problem of unequal access to ecologically valuable territory; Addis Ababa makes it dependent on institutional and community governance; Tunis reveals water-bearing landscapes whose ecological operation precedes and constrains architecture. The common analytical object is therefore not “green space” but the metropolitan configuration through which ecological capacities become lived, protected, measurable and actionable.
This four-city constellation also clarifies what comparison should and should not do. A conventional comparative framework could easily convert these studies into a table of indicators—heat risk, hectares of protected land, governance scores, wetland area—and lose the strongest knowledge each city produces. The more productive operation is to compare relations while preserving the epistemic conditions of each source. Lagos demonstrates that environmental risk becomes legible only when climatic hazard is multiplied by exposure and vulnerability; Cape Town demonstrates that ecological abundance has to be crossed with mobility, safety, information and post-apartheid spatial form; Addis Ababa demonstrates that the survival of green infrastructure depends on formal and informal actor networks rather than on master-plan designation alone; Tunis demonstrates that hydrological and geomorphological systems can be erased conceptually before they are erased physically, simply by being classified as development land rather than ecological infrastructure. These are not four instances of one universal law. They are four different tests of what must be added to a minimal comparative grammar before a claim about metropolitan nature can become credible. For METROPOLITAN NATURES, this suggests a precise operational architecture. FORM must include density, morphology, edges, accessibility and the infrastructural networks that distribute exposure; NATURE must include vegetation but also shade, water, wetland storage, ecological condition and maintenance; LIFE must record actual use, barriers, safety, thermal experience, belonging and the practical ability to occupy environmental space; PROJECT must document the institutional and design decisions through which ecological capacities are maintained, transformed or removed. The temporal sequence MATURE–TRANSFORMED–FUTURE also gains substance from these sources. Cape Town’s reserves are mature landscapes whose ecological value has accumulated while barriers of access remain embedded in metropolitan structure. Addis Ababa’s community green spaces expose transformation as a governance process in which local management can either stabilise or fail to stabilise environmental provision. Tunisian sebkhas are future-project territories in the most literal sense: development decisions can preserve, reinterpret or suppress their hydrological function before construction fixes the outcome. Lagos demonstrates why a FUTURE decision cannot be evaluated through project imagery alone, because a design intervention enters an already differentiated thermal field in which dense, low-income and infrastructure-poor populations carry disproportionate exposure. The four cases therefore strengthen Metropolis Friction without requiring another conceptual layer. A method transferred between cities encounters different forms of resistance: the same canopy metric has different significance in humid Lagos and semi-arid Tunis; an accessibility protocol developed around Cape Town conservation areas would need to be reformulated where green spaces are community-managed neighbourhood fragments in Addis Ababa; a flood-storage reading derived from Tunisian sebkhas cannot be transferred as a formal wetland template to Lagos without testing different coastal, hydrological and settlement conditions. What travels is not the design solution but the relational question. Where does heat accumulate, and through which social conditions? Who can reach ecological space, and what blocks them? Who actually governs the landscape after the plan is approved? What environmental work is already being performed by land that a project represents as available? These questions are portable precisely because their answers are not. The sources also demonstrate why evidence itself must remain plural. Numerical simulation in Lagos, survey and GIS work in Cape Town, mixed governance analysis in Addis Ababa and design-research synthesis in Tunis reveal different aspects of metropolitan nature. None should be elevated into a universal hierarchy of rigour. Their validity depends on matching method to question and making limits explicit. A heat-risk model cannot explain why a resident avoids a reserve; a survey cannot calculate a humid heatwave; a governance model cannot substitute for hydrological analysis; a landscape studio cannot independently validate every ecological process it proposes to conserve. Comparative strength comes from making these forms of evidence interoperable without pretending they are interchangeable. That is the deeper contribution of this final African evidence set: it shows how locally grounded studies can support a shared metropolitan research architecture while retaining scientific irreducibility. Nature is not simply what remains outside development. It is a set of active relations through which urban form redistributes heat, mobility distributes access, institutions distribute responsibility and design distributes ecological futures. A metropolitan landscape becomes legible when those relations are traced together. The comparative project becomes credible when it standardises enough to ask the same question and preserves enough difference to accept that each city may change what the question means.
Anto Lloveras is an architect and urban researcher working across comparative urbanism, metropolitan nature, spatial epistemology and public knowledge infrastructures through LAPIEZA LAB and Socioplastics.
Ayele, B.Y., Megento, T.L. and Yeshitela Habetemariam, K. (2022) ‘The governance and management of green spaces in Addis Ababa, Ethiopia’, Heliyon, 8, e09413. https://doi.org/10.1016/j.heliyon.2022.e09413.
Mili, S. (2016) ‘Instant Cities on the Wet Coastal Zones—Tunisia’, Procedia Environmental Sciences, 34, pp. 525–538. https://doi.org/10.1016/j.proenv.2016.04.046.
Milliken, J. (2015) Accessing Green Space in Cape Town: A Case Study of Public Perceptions of Green Space and Barriers of Access in Eight Nature Reserves and Conservation Areas within Cape Town. MPhil dissertation. Cape Town: University of Cape Town.
Obe, O.B., Morakinyo, T.E. and Mills, G. (2023) ‘Assessing heat risk in a sub-Saharan African humid city, Lagos, Nigeria, using numerical modelling and open-source geospatial socio-demographic datasets’, City and Environment Interactions, 20, 100128. https://doi.org/10.1016/j.cacint.2023.100128.
