Finding space for nature in cities: the considerable potential of redundant car parking

The study addresses how cities can deliver nature-based solutions (NBS) at scale within constrained urban environments by systematically reallocating land from on-street car parking to biodiverse green space. Urban areas face increasing heatwaves, flooding, biodiversity loss, and inequitable access to green space, while densification has reduced private greenery without equivalent public replacements. Despite bold municipal NBS targets, implementation at scale remains limited due to space constraints and political contestation. The authors propose a trade-off: consolidate on-street parking into nearby underutilised off-street garages and convert the freed kerbside spaces into multifunctional green infrastructure. The City of Melbourne, with extensive streetscape area, abundant garage capacity, and pressing canopy, stormwater, and biodiversity targets, serves as the case study. The research question is whether reallocation of redundant street parking near vacant garages (within short walking distances) can produce substantial ecosystem service benefits—tree canopy expansion, stormwater interception, and ecological connectivity—sufficient to materially advance city targets while minimising perceived costs to drivers by maintaining parking supply off-street.

The introduction synthesises evidence that well-designed NBS reduce heat, manage stormwater, and improve public health, yet urban densification has caused green space loss and biodiversity decline, exacerbating flood risk and heat islands. Many cities have adopted ambitious greening targets aligned with global policy drivers (SDGs, climate adaptation, biodiversity conservation, environmental justice, and post-COVID green recovery). However, scaling NBS requires significant land use change, often overlooked in optimistic discourse. Streets occupy a quarter to a third of central urban land, with a substantial share for on-street parking; meanwhile, planning rules have produced large volumes of underutilised off-street parking with documented vacancies across residential, private, and commercial garages. Literature highlights that consolidating parking via centralised facilities or shared/peer-to-peer systems is feasible but politically sensitive due to perceptions of convenience, cost, and impacts on commerce. Cities increasingly reconsider street space allocation, indicating potential for systematic reallocation to NBS if trade-offs are carefully managed (short walking distances, maintaining supply via garages).

Case study: City of Melbourne (37.7 km²) within metropolitan Melbourne, facing heat, flooding, and biodiversity challenges, with relevant open data and active parking reform interest. The analysis used a conservative ‘no net loss’ assumption: on-street parking would be moved off-street, not removed. Phase 1 – Identify redundant on-street parking: Using City of Melbourne open data on off-street parking capacity by type (residential, commercial, private) and on-street spaces plus street network, the team estimated available garage vacancy and applied location-allocation analysis in ArcGIS (Network Analyst, ‘maximise capacitated coverage’) to allocate nearby on-street spaces to off-street vacancies. Twelve scenarios varied by: (a) destination garage type used (commercial only; private and residential only; combined), (b) vacancy assumptions (lower: 30% commercial, 10% private/residential; higher: 70% commercial, 20% private/residential), and (c) maximum walking distance along the street network (100 m or 200 m) from on-street space to garage centroid. Distances to exact entries could not be resolved; building centroids were used. A 1:1 replacement of on-street with off-street capacity was assumed despite observed on-street spare capacity. Phase 2 – Design and model ecosystem services: A modular biodiverse green space design was created to replace parking bays, tailored for different contexts: (A) kerbside in commercial areas with decking/seating, tree, planter boxes, raingarden; (B) standard kerbside bays with tree, substantial understorey habitat, raingarden; (C) median bays with tree and understorey but no raingarden (median slope prevents inflow; seating omitted). Designs incorporated WSUD/BSUD principles; site constraints (overhead cables, high-speed roads, underground services, existing canopy, dining areas, medians) were addressed with flexible tree species selection, reinforced sleeves for barriers/furniture, and site-specific adaptability.

• Canopy modelling: Municipal inventory and canopy polygons were intersected to derive species-specific linear regressions relating DBH to crown area for nine native street-tree species already used locally (Allocasuarina verticillata, Angophora costata, Corymbia maculata, Eucalyptus camaldulensis, E. leucoxylon, E. polyanthemos, Melaleuca styphelioides, Syzygium smithii, Tristaniopsis laurina). An equal mix of species was assumed. For each viable converted space, one tree was planted unless existing canopy already overlapped the bay centroid (20–28% excluded). Canopy at maturity used the 95th percentile DBH; intermediate benefits used 25th and 50th percentile DBH. Growth reflects existing conditions; enhanced growth from passive irrigation was not modelled.
• Ecological connectivity: Functional connectivity was quantified using effective mesh size (meff) per Kirk et al. for two focal species (New Holland honeyeater and Blue-banded bee) with species-specific habitat definitions, barriers (roads/rails >15 m for honeyeater; >10 m for bee; tall buildings), and maximum gap-crossing distances (460 m honeyeater; 300 m bee). Each parking conversion added a habitat patch of bay extent; fragmentation layers were updated by scenario and changes in connected area, coherence, and connected area size were computed in R (sf package).
• Stormwater interception: A representative directly-connected hardstand catchment of 395 m² per raingarden was established from measurements across seven street typologies. Rooftop runoff was excluded; imperviousness assumed uniform. MUSIC v6.0 model inputs (filter area ≈14 m² = 3.5% of catchment, storage, media, inlets/outlets) followed Melbourne Water and industry guidelines. To avoid oversizing relative to catchment, only every fourth adjacent converted bay in a group was assumed to operate as a raingarden; single isolated bays could host a raingarden. Median conversions were excluded from stormwater modelling. Totals per scenario equalled per-unit MUSIC outputs multiplied by the number of viable raingarden sites. Outputs: Gross Pollutants (kg/yr), Total Suspended Solids (kg/yr), Total Phosphorus (kg/yr), Total Nitrogen (kg/yr).

• Redundant parking supply: Across 12 scenarios, 3,146 to 11,668 on-street spaces could be reallocated to nearby garages, depending on vacancy assumptions, garage types included, and 100 m vs 200 m walking thresholds. The maximum (11,668) equals 47% of the city’s 24,745 on-street spaces, covering about 50 ha of street area.
• Tree canopy: Converted spaces could deliver 31–59 ha of additional tree canopy at maturity, with 11–22 ha at intermediate stages, augmenting the existing 254 ha public-realm tree canopy. Species were selected primarily for habitat value, not canopy maximisation.
• Ecological connectivity: Connectivity increased for both focal species by adding stepping-stone habitats, with the Blue-banded Bee showing the greatest improvement. Gains were markedly higher under 200 m reallocation scenarios compared to 100 m.
• De-paving: The program could remove 6.6–24.5 ha of asphalt citywide, creating permeable, biodiverse green space equivalent to roughly 1.5–6 city blocks. Within the flood-prone Elizabeth Street Catchment, 2.7–7.7 ha could be de-paved.
• Stormwater interception: Raingardens could capture up to ~27 tonnes/yr of gross pollutants (litter) and ~202 tonnes/yr of sediment, plus hundreds of kg/yr of phosphorus and nitrogen.
• Policy benchmarks: Relative to City of Melbourne targets, scenarios deliver: canopy progress of ~9%–33% towards the 40% public-land canopy by 2040 target (437 ha); de-paving in the Elizabeth Street Catchment of 4%–12% of the 65 ha target; sediment interception of 33%–116% of the 175 t/yr target; phosphorus interception of 17%–118% of the 251 kg/yr target; nitrogen interception of 50%–59% of the 2,054 kg/yr target; litter interception of 6%–21% of the 130.5 t/yr target.

Systematically reallocating redundant kerbside parking into nearby underutilised garages can free substantial public land for multifunctional NBS, yielding integrated gains in canopy, stormwater treatment, ecological connectivity, and permeable surface area in dense urban cores. The City of Melbourne results demonstrate that thousands of bays—up to half the on-street supply—could be converted within short walking distances (≤200 m), delivering 31–59 ha of new canopy and meeting or exceeding citywide sediment and phosphorus interception targets in certain scenarios. Connectivity benefits are strong, especially for pollinators, and would be amplified at larger walking thresholds. This strategy is internationally relevant where streets occupy large shares of central land and off-street parking is abundant due to legacy minimum parking requirements. The interdisciplinary approach contrasts with single-function NBS programs by quantifying multiple ecosystem services, though it still undercounts broader benefits (cooling, air quality, mental and physical health, economic revitalisation). Politically, reallocating kerbside space is contentious; perceived losses in convenience and entrenched norms around street parking can spur opposition. Practically, enabling public access to private garages may require retrofits and pricing changes or subsidies. Nonetheless, modular designs and incremental roll-out make the transition feasible when aligned with policy goals and public support, as evidenced by large-scale greening efforts in cities like New York.

The study shows that converting redundant on-street parking to biodiverse green space—by consolidating parking into nearby vacant garages—can substantially advance multiple municipal sustainability targets in dense urban areas. In Melbourne, this single tactic could de-pave up to 24.5 ha, add 31–59 ha of canopy, improve ecological connectivity, and meet or exceed sediment and phosphorus interception goals. These findings support reframing street space allocation toward public benefit from NBS while maintaining overall parking supply off-street. Future work should: refine designs per site to unlock larger greening footprints; explore reduced replacement ratios (not strictly 1:1) and longer walking thresholds to scale impacts; quantify additional benefits (thermal comfort, air quality, local flood mitigation, public health, economic value, and jobs); and develop comprehensive tools for multifunctional NBS valuation to guide politically robust implementation.

Key limitations stem from conservative and simplifying assumptions: (1) 1:1 replacement of on-street spaces with garage capacity, despite known on-street spare capacity; (2) garage access modelled to building centroids rather than exact entry points; (3) limited walking thresholds (100–200 m) constrain potential consolidation; (4) off-street vacancy rates for commercial/private parking uncertain post-COVID; (5) existing canopy over bay centroids excluded from planting; (6) tree growth modelled from existing street conditions without passive irrigation benefits; (7) stormwater modelling assumed a fixed 395 m² hardstand catchment per raingarden, excluded rooftops, and limited to one raingarden per four adjacent bays; (8) median conversions excluded from stormwater due to camber; (9) multiple ecosystem co-benefits (cooling, air quality, health, socio-economic outcomes) were not quantified. Political and logistical challenges (garage retrofits, pricing, coordination, maintenance) may affect implementation and generalisability.