The Green Infrastructure Hiding in Plain Sight

How small leftover urban spaces can reduce urban heat, reconnect fragmented habitats and strengthen climate resilience

The false choice between development and green space

In dense cities, the standard answer to a shortage of greenery is to search for land for another park. That remains essential where land can be secured, but it is often the hardest intervention to deliver. Sites are expensive, ownership is fragmented and the same land is under pressure for housing, transport, jobs and public services. The result is a false policy choice: either reserve large areas for green space or accept increasingly hard, hot and ecologically fragmented neighbourhoods.

There is a third option. Almost every city contains a dispersed layer of small spaces that planning systems rarely treat as strategic assets: building setbacks, road verges, unused corners of public sites, school and religious compounds, residual strips between plots, underused forecourts, utility land and small vacant parcels. Some of these spaces must remain open for access, servicing, drainage, safety or future development. Others can perform an environmental function with little or no displacement of existing activity.

The central policy question is therefore not whether small spaces can replace parks. They cannot. It is whether a city can make thousands of small spaces work together as a second layer of green infrastructure, supporting the larger parks, wetlands, waterways and urban forests on which its environmental resilience depends.

Sources and notes:

[1] Rinku, M. A. (2021). Planning to increase collective urban green area with minimum intervention by utilising leftover spaces. MSc dissertation in Urban and Regional Planning, The University of Sheffield.

[2] Li, Y., Svenning, J.-C., Zhou, W. et al. (2024). Green spaces provide substantial but unequal urban cooling globally. Nature Communications, 15, 7108.

[3] World Bank (2025). An Unsustainable Life: The Impact of Heat on Health and the Economy of Bangladesh.

[4] Wu, Y., Patuano, A., Mashhoodi, B., Lenzholzer, S., Acred, A. & Narvaez Zertuche, L. (2025). How small green spaces cool urban neighbourhoods: Optimising distribution, size and shape. Landscape and Urban Planning, 253, 105224.

[5] Han, Y., Ren, Y., Wan, M. & Wan, M. (2026). Enhancing urban ecological connectivity through small stepping stones: A solution for inequitable geographical distribution of blue-green spaces. Journal of Environmental Management, 404, 129247.

[6] Perrelet, K., Moretti, M., Dietzel, A., Altermatt, F. & Cook, L.M. (2024). Engineering blue-green infrastructure for and with biodiversity in cities. npj Urban Sustainability, 4, 27.

[7] Croeser, T., Bekessy, S.A., Garrard, G.E. & Kirk, H. (2024). Nature-based solutions for urban biodiversity: Spatial targeting of retrofits can multiply ecological connectivity benefits. Landscape and Urban Planning, 251, 105169.

[8] National Parks Board Singapore (2026). Nature corridors and Nature Ways. Accessed August 2026.

Note on evidence: The Sylhet percentages are derived from selected study locations in the 2021 dissertation by the author. The article does not infer a specific temperature reduction for Sylhet from those percentages. Evidence on cooling effects is drawn from the separate international studies cited above, which use different thermal metrics and urban contexts.

Figures refer to selected Sylhet study locations and should not be interpreted as a citywide estimate.

Urban heat island is partly a land-use problem

The urban heat island effect occurs because the physical structure of cities changes how heat is absorbed, stored and released. Asphalt, concrete and dark roofs absorb solar energy during the day and release it slowly, while traffic, air-conditioning and other human activity add further heat. At the same time, the loss of vegetation removes shade and evapotranspiration, two of the most effective local cooling processes available to a city.

This matters because climate change and urban heat can compound one another. A global study of almost 500 major cities found that existing urban green infrastructure lowers warm-season daytime land-surface temperatures by about 2.9°C in the average city. It also found a marked inequality: cities in the Global South had average cooling capacity of about 2.5°C, compared with 3.6°C in the Global North.[2]

Bangladesh shows why that gap is a policy concern. The World Bank reports that maximum temperatures increased by 1.1°C nationally between 1980 and 2023, while Dhaka's maximum temperature rose by about 1.4°C. Between 1989 and 2020, Dhaka lost 47 per cent of its dense green space and vegetation as urban settlement expanded. In 2024, heat-related physical and mental health conditions were estimated to have caused 250 million lost workdays and economic losses of up to US$1.78 billion, around 0.4 per cent of GDP.[3]

Greening will not remove the urban heat island effect on its own. Building design, ventilation, reflective materials, labour protection, public-health preparedness and energy policy all matter. But vegetation is one of the few interventions that can simultaneously reduce radiant heat, improve thermal comfort, intercept rainfall and provide habitat. That makes the question of where additional greenery can fit especially important in land-constrained cities.

The finding is useful, but it needs to be interpreted carefully. The study did not model a resulting change in Sylhet's air temperature, and the 20.07 per cent figure cannot be applied mechanically to Dhaka, Chattogram or another city. What it demonstrates is spatial potential: meaningful additional green area can be hidden in the cumulative effect of many small sites. The thermal benefit of such a strategy has to be assessed using the wider evidence on urban greening and the particular morphology of each neighbourhood.

The Sylhet research also found that opportunity was unevenly distributed. Residential and internal areas contained much of the potential addable greenery, while government, public, educational and religious land offered spaces that may be easier to coordinate.[1] This has a direct policy implication: public land can provide early demonstration projects, but a citywide programme will eventually need incentives, standards and stewardship arrangements that reach private and institutional land as well.

Small spaces can reduce heat, if they are planned as a system

The cooling effect of a green space depends on more than its area. Shade, vegetation structure, soil moisture, surrounding building form, wind, street orientation, surface materials and the position of the space within the neighbourhood all influence performance. A small patch in the right place can improve pedestrian thermal comfort, while the same area scattered randomly may deliver much less benefit.

Recent modelling of green spaces smaller than one hectare reinforces this point. A 2025 study found that the spatial distribution of small green spaces strongly affected neighbourhood cooling. Grouping several small green areas produced the strongest cooling outside the spaces, reducing Physiological Equivalent Temperature by around 1.3°C in the modelled scenarios. Thermal conditions inside the spaces varied by as much as 4°C between different design configurations.[4] These figures are not universal temperature reductions, but they show that design and location matter as much as simply increasing the amount of green on a map.

For urban heat policy, this suggests a shift from a planting target to a cooling network. Residual spaces should be prioritised where several factors coincide: high daytime or night-time heat, heavy pedestrian use, low existing canopy, vulnerable populations, extensive hard surfacing and opportunities to link with existing vegetation. Street-level shade is particularly valuable because it changes the thermal conditions people actually experience when walking, waiting for public transport or working outdoors.

The most effective interventions will vary. Some sites may support canopy trees; others may be better suited to lower planting, rain gardens, permeable surfaces, climbers or a combination of vegetation and shade structures. In narrow streets, poor species choice or inappropriate tree form can interfere with utilities or ventilation. A heat-mitigation programme therefore needs design standards based on local urban form, not a single planting template.

The second dividend: an ecological habitat network

The same small spaces can serve another function that is often missing from urban greening programmes. Cities rarely lose biodiversity only because they lack total green area. They also lose it because habitats become fragmented. A park, wetland or mature institutional ground may remain ecologically valuable, but roads, buildings and large areas of hard surfacing can isolate it from the next habitat patch.

The original Sylhet research recognised that even small urban green spaces can support plants and animals, and that their biodiversity value is influenced by connectedness as well as size.[1] More recent work provides a clearer planning framework. Blue-green infrastructure can support species dispersal through continuous corridors, but it can also use smaller 'stepping-stone' habitats, such as ponds, green roofs and planted patches, where continuous corridors are impossible.[6]

A 2026 study in Xiangyang, China, illustrates the potential. Researchers identified 24 small stepping-stone sites at breaks and barriers in the city's ecological network. In the model, adding those sites increased the number of optimal ecological corridors from 73 to 121.[5] The precise numbers belong to Xiangyang, but the principle is widely relevant: the ecological value of a small site depends partly on what it connects.

Evidence from Melbourne points in the same direction. Researchers modelled the conversion of street parking spaces into small green spaces and compared untargeted investment with schemes located at important ecological barriers. Targeted greening produced up to roughly twice the connectivity benefit per unit of intervention for the species modelled.[7] In other words, cities can often obtain more ecological value from the same quantity of greening by improving site selection.

This creates a powerful opportunity to combine heat adaptation and biodiversity policy. A residual plot that sits on a hot walking route and also closes a gap between two habitat areas can deliver a double dividend. The planning task is to identify those overlaps before money is committed.

From leftover land to urban infrastructure: seven policy moves

1. Map heat, habitat and residual land together
Create a citywide inventory of small unbuilt or underused spaces using current imagery, cadastral data, public-land records and field surveys. Overlay this with heat maps, tree canopy, flood risk, pedestrian activity, existing parks, wetlands, watercourses and other ecological assets. The aim is not to produce a list of empty plots. It is to understand what each space could do in relation to the systems around it.

2. Screen out land that is not genuinely available
A space may look unused while performing an essential role for fire access, servicing, drainage, utilities, visibility, pedestrian movement, informal livelihoods or future development. A transparent suitability test is necessary before any greening proposal is made. This protects the credibility of the programme and avoids treating every unbuilt area as environmental surplus.

3. Prioritise the heat-connectivity overlap
Rank sites by the public benefits they can deliver. A small space on a heat-exposed pedestrian route that also bridges an ecological gap and absorbs runoff may justify investment ahead of a larger but isolated site. This is especially important where budgets are limited.

4. Design for cooling and habitat quality
Greenness alone is not enough. Cooling depends on shade, evapotranspiration, soil and spatial form; habitat value depends on species choice, vegetation structure, food, shelter, water, lighting and management. Native or locally appropriate planting, multiple vegetation layers and reduced pesticide use can improve ecological value, while canopy placement should respond to solar exposure and pedestrian need.

5. Use public land first, then widen participation
Municipal land, schools, hospitals and other public sites can establish visible early projects because ownership and maintenance responsibilities are clearer. Private and institutional land will require a different toolkit, including technical support, small grants, stewardship agreements, development incentives and proportionate planning requirements for new development.

6. Lock in maintenance before planting
Urban greening often fails after the launch event. Every intervention should identify who waters, prunes, replaces and monitors planting, how these costs are funded, and what happens when a site changes owner or occupier. A smaller programme with credible long-term maintenance will deliver more cooling and habitat value than a larger programme with poor survival.

7. Measure outcomes, not planting numbers
Cities should monitor canopy survival and growth, shaded walking area, local thermal comfort, permeability, habitat use and connectivity. Tree counts are an input. The policy outcome is whether neighbourhoods become cooler, more permeable and more ecologically connected, especially in the places where heat exposure and environmental inequality are greatest.

→ All Topics

Bangladesh can test the model at different urban scales

Bangladesh is well placed to test this approach because its cities present very different combinations of density, climate, water, ecology and development pressure. A national framework should standardise the method, not the planting design. Mapping, suitability testing, heat and habitat prioritisation, stewardship and monitoring can be common requirements, while interventions remain locally specific.

Sylhet is an obvious pilot because the 2021 study provides a baseline method and a set of testable propositions. A new audit could update the mapping with current imagery and add heat, drainage and ecological connectivity layers. It could ask not only how much green area is still available, but which spaces would reduce heat exposure and strengthen links between watercourses, wetlands, institutional grounds and neighbourhood vegetation.

Dhaka would require a more constrained strategy. The World Bank's evidence on rising heat and vegetation loss makes the need clear.[3] The most realistic opportunities may lie in public and institutional grounds, school sites, road verges, utility land, courtyards, rooftops and spaces released through street redesign. In the densest areas, thermal benefit to pedestrians may be the first priority, with ecological mapping used to identify where small patches can also reinforce the remaining green and blue network.

Chattogram should tie the approach to hill vegetation, slope stability, drainage and watercourses. Khulna would require salinity- and waterlogging-tolerant planting. Rajshahi would place greater weight on drought resilience and continuous shade. Industrial areas in Gazipur and Narayanganj offer another type of opportunity through factory setbacks, boundary buffers and worker routes. Smaller cities may have the greatest preventative advantage because they can protect and connect residual spaces before rising land values make intervention much harder.

International practice shows that connectivity can be built incrementally

Singapore demonstrates the principle at a larger and more deliberate scale. Its Nature Ways use multi-tier planting along roads and other linear routes to connect parks and nature reserves, supporting the movement of birds and butterflies while also providing shade and greenery for residents. As of 2026, Singapore reported 57 Nature Ways extending about 260 km.[8]

The lesson is not that other cities should reproduce Singapore's spending or governance model. It is that an ecological network can be assembled from different types of urban land. Streets, parks, waterways, institutional sites, building plots and small residual spaces can all contribute if they are planned as parts of the same system.

Protect the core, then connect it

There is an important limit to the argument. Small green spaces cannot compensate for the destruction of wetlands, mature urban forests, major parks or established tree canopy. Nor should a city use a programme of pocket greening to justify the loss of larger ecological assets. The correct hierarchy is the opposite: protect and restore the core habitats first, secure continuous corridors where possible, then use smaller spaces to fill gaps, extend shade and create stepping stones where fragmentation cannot otherwise be overcome.

This hierarchy also helps avoid a common weakness in urban climate policy. The goal should not be to make development look greener after the fact. It should be to preserve the environmental systems that already work, then use targeted small-scale interventions to improve performance where the city has become too hot, too impermeable or too fragmented.

A different way to value scarce urban land

The deeper policy issue is how cities value small pieces of land. Conventional appraisal tends to ask what a site can build, service or accommodate. Residual-space greening adds another question: what public function could this land perform if development is not its best or immediate use?

In a warming city, a strip of land can have value because it shades a footpath, cools a school edge, absorbs stormwater or helps a pollinator move between habitat patches. Those functions do not appear as floor space, but they affect public health, infrastructure performance and ecological resilience. As heat intensifies, their value is likely to rise.

The distributional question matters just as much. Greening should not become a premium amenity concentrated around affluent neighbourhoods, flagship developments or prestige streets. Heat exposure, lack of existing green space, pedestrian need and ecological fragmentation should influence where public investment goes. The same spatial analysis that identifies leftover land can be used to direct environmental benefits towards neighbourhoods that currently receive the least.

Conclusion

The Sylhet research began with a modest question: how much additional green area could be created without major land acquisition or redevelopment? In the selected study locations, the answer was substantial. But the more important policy lesson is the change in perspective that produced it.[1]

Dense cities contain a dispersed stock of small spaces that planning systems routinely overlook. Some should remain exactly as they are. Others can provide shade, reduce heat exposure, absorb rainfall, support biodiversity or bridge the gaps between larger habitats. The strongest sites can do several of these things at once.

A roadside strip will never replace a mature park, and a planted setback will never replace a wetland. Cities do not have to choose between large habitats and small spaces. The larger assets provide the ecological and climatic core; targeted smaller interventions can become the connective tissue that carries cooling and habitat benefits through the urban fabric.

For fast-growing cities, particularly in the Global South, this is a practical form of climate adaptation. Before concluding that there is no room left for nature, city authorities should first map the room that is already there, identify where urban heat and ecological fragmentation overlap, and make those spaces work as part of a coherent network.

The policy proposition

• Treat suitable residual land as part of urban green infrastructure, rather than as isolated leftover plots.

• Target greening where heat exposure and ecological fragmentation overlap, so the same investment can cool people and reconnect habitat.

• Protect large parks, wetlands, watercourses and mature tree groups first. Small-space greening should connect and extend these assets, not compensate for their loss.

• Measure results through canopy survival, shade, thermal comfort, permeability and habitat connectivity, not simply the number of trees planted.

What the Sylhet case tells us, and what it does not

A 2021 MSc study at the University of Sheffield tested the potential of this approach in Sylhet, a rapidly developing regional city in Bangladesh. Using aerial photomapping, street-level observation, land-use classification and professional interviews, the research identified existing greenery and mapped leftover spaces that could potentially be planted without major redevelopment.[1]

Across the selected study locations, about 23 per cent of land was already green. Suitable leftover spaces were estimated to be capable of adding another five percentage points, increasing green cover in those locations to about 28 per cent. In relative terms, that was a 20.07 per cent increase in the amount of existing green area.[1]

Murshed Ahmed is a Senior Urban Planner based in the UK and Chartered Member of the Royal Town Planning Institute.