New method maps where and how cities should turn brownfields into green spaces

Abandoned factories, disused rail yards and derelict industrial plots are often dismissed as scars of a city’s industrial past. A new study argues they may be its greatest untapped asset. Researchers have developed a cross-scale planning approach that uses ecosystem service assessments to answer two deceptively simple questions about urban greening: where should a city green its brownfields, and how should each individual site be greened? Tested in Tangshan, a coal-mining city in northern China with more than 140 years of industrial history, the method links strategic urban-scale priorities to concrete, site-level actions, offering a blueprint for shrinking and resource-exhausted cities worldwide.

The work, published in the journal One Ecosystem, addresses a persistent gap in nature-based solution planning. Transforming brownfields into green spaces is widely recognised as a cost-effective way to enhance urban resilience, and countries such as the United States, Canada, Germany and France have long promoted the practice. In France, a national brownfield database and dedicated funding were established in 2022 to accelerate restoration. Yet because brownfields rarely exist as isolated parcels but instead form clusters across the urban landscape, greening them piecemeal risks wasting resources. Previous studies either ranked brownfields from a supply-side perspective focused on a single service, such as landscape connectivity, or measured post-greening benefits at the site level without connecting back to city-wide strategies. The result is a fragmented approach that offers little guidance to coherent planning.

The new framework closes that silo by integrating spatially explicit assessments of ecosystem service supply, demand and the balance between them. It distinguishes between strategic planning, which engages stakeholders to develop comprehensive land-management strategies, and detailed planning, which determines operational feasibility and what type of nature-based solution should be implemented where on a specific site. The approach unfolds in three steps: mapping supply and demand for selected services, synthesising the two into a diagnostic ratio that reveals deficits and surpluses, and then defining strategies and actions through multi-criteria analysis and cluster analysis.

Five ecosystem services were selected for the Tangshan case study, each anchored in the city’s own planning goals. Runoff mitigation and urban heat island mitigation were chosen to address climate adaptation, carbon storage and habitat quality to align with carbon neutrality and ecological restoration targets, and recreation service to meet the demand for accessible public green space in densely populated areas. Supply was modelled primarily with the InVEST suite of models, while demand was estimated from population density, statistical records, planning standards and meteorological data.

The technical details reveal the rigour behind the mapping. Runoff mitigation supply was quantified with the InVEST Urban Flood Risk Mitigation model, which estimates pluvial runoff reduction through a runoff retention index based on land cover, soil hydrological groups and curve numbers. Demand was expressed as a flood risk index, the product of hazard and vulnerability, combining urban population and buildings at risk on a grid scale, with weights of 0.6 and 0.4 attributed to population and buildings respectively. Urban heat mitigation was assessed with the InVEST Urban Cooling Model, which calculates temperature reduction from shading, evapotranspiration, albedo and the cooling effects of large green areas, using an air blending distance of 550 metres and a maximum cooling distance of 450 metres. Heat demand was defined by hazard and exposure, with overheated patches identified from Land Surface Temperature values classified into five categories ranging from 28.7 to above 41.4 degrees Celsius.

Carbon storage supply was estimated from above- and below-ground vegetation and soil organic carbon pools, excluding dead organic carbon because of its negligible content. Demand was derived from per capita emissions: dividing Tangshan’s 310 million tonnes of CO₂ among its 7.7 million residents yields 40.16 tonnes of CO₂ per person, equivalent to roughly 10.84 tonnes of carbon, which was then multiplied by population density to create a spatially explicit demand map. Habitat quality was measured with the InVEST habitat quality model, treating agricultural, built-up and barren land as threat sources, while demand was benchmarked against the regional median habitat quality. Recreation supply was mapped with the InVEST Urban Nature Access model using a 500-metre search radius, and demand was set at the 18 square metres of green space per person specified in the city’s Greenspace System Planning.

To synthesise supply and demand, the researchers calculated the Ecosystem Service Supply and Demand Ratio, a diagnostic indicator in which zero indicates balance, negative values signal ecological deficits requiring urgent intervention and positive values represent surplus. The five ratio maps were normalised to the range from minus one to one for comparability. At the urban scale, average standardised values were computed for each brownfield using zonal statistics, then classified and scored from 0 to 2 per service and summed into a combined regeneration priority map. At the site scale, principal component analysis reduced the data to three components explaining over 85 percent of total variance, and k-means clustering revealed the internal heterogeneity of supply-demand mismatches within each site.

The results are striking. Using multi-source data and visual interpretation of Point of Interest records and 2022 satellite imagery, calibrated against urban zoning maps and municipal archives, the team identified 1,473 brownfields covering 317.1 square kilometres, about 2 percent of Tangshan’s land area, ranging from 955 square metres to 17.9 square kilometres. Around two-thirds of these sites lie in areas with high demand and low supply for carbon storage and habitat provision, roughly 25 percent could contribute to runoff mitigation and recreation, and only about 5 percent fall in areas with unmet demand for microclimate regulation. The combined priority map shows that 431 plots, 29.2 percent of the total, score in the very high or high tiers, predominantly in urban centres and peripheries, while 619 parcels, 42 percent, rank low or very low, mostly in suburban and remote northern mountainous regions.

Cluster analysis identified four distinct supply-demand profiles that translate directly into prescriptions. Cluster 1, covering about 80.6 square kilometres, shows deficits in all services except recreation, calling for comprehensive interventions such as parks, urban forests, bio-swales and green roofs. Cluster 2, about 69.7 square kilometres, is characterised by deficits in carbon storage, runoff mitigation and especially habitat quality, pointing to artificial wetlands and retention ponds. Clusters 3 and 4, roughly 72.8 and 88.5 square kilometres, show surpluses due to low population density; Cluster 3 requires only maintenance of existing green space, while Cluster 4, with notably higher heat mitigation ratios, would benefit most from natural succession rather than intentional greening. Illustrative examples show how the two scales work together: two former steel factories in the city centre, one scoring a maximum 10 for very high priority, fall mostly into Cluster 1, suggesting de-sealing and revegetation through urban forests, while smaller portions in Cluster 2 call for bioswales, constructed wetlands or rain gardens, and areas in Cluster 3 need only spontaneous greening.

The implications reach far beyond Tangshan. The authors note that the framework measures current urgency rather than predicting future ecological impact, and that its linear aggregation assumes services are substitutable, a necessary simplification for strategic screening. Data limitations, such as reliance on a global soil database and empirical cooling parameters, mean on-the-ground validation is still needed, and future work should add the temporal dimension, since sequential greening projects dynamically alter supply and demand, particularly when sites are close together. The method also extends naturally to managing abandoned housing in shrinking cities in Germany, Japan and the United States, echoing right-sizing strategies. Because land in China is primarily state or collectively owned, the priority map can serve as a direct reference for municipal resource allocation, though the researchers caution that findings should support rather than dictate decisions, that green gentrification risks require active stakeholder participation, and that hidden soil contamination demands strict environmental impact assessment and long-term monitoring before recreation-focused greening proceeds.

Subject of Research: A cross-scale ecosystem service approach linking urban brownfield greening priorities with site-specific nature-based solution actions, tested in Tangshan, China

Article Title: From urban scale priorities to local actions: an approach to develop brownfield greening strategies, based on ecosystem services supply and demand

Article References: Chen, B., Cortinovis, C., Hashimoto, S., & Zheng, X. (2025). From urban scale priorities to local actions: an approach to develop brownfield greening strategies, based on ecosystem services supply and demand. One Ecosystem, 10, Article e175875. https://doi.org/10.3897/oneeco.10.e175875

Image Credits: AI Generated

DOI: 10.3897/oneeco.10.e175875

Keywords: brownfield greening, ecosystem services, nature-based solutions, urban planning, Tangshan, InVEST, shrinking cities, supply and demand, urban, scale, priorities, local