02GreenRule - 3-30-300 Green Equity Lab

User manual, version 0.1.1 (2026-11-16). QGIS 3.28 to 4.x. GPL-2.0-or-later. Yusuf Eminoğlu (Geophilo).

1. What the plugin does

The 3-30-300 rule (Konijnendijk 2023) asks that every resident sees at least 3 trees from home, lives in a neighbourhood with at least 30 percent tree canopy, and has a public green space within 300 m. Measuring it needs three GIS techniques at once: line of sight from homes to individual trees, canopy mapping from imagery or canopy-height data, and network distance to qualifying parks (Browning et al. 2024).

02GreenRule carries such a study from open data to a neighbourhood scorecard. It fetches OpenStreetMap buildings, trees, parks and paths and an open canopy source for the study area; derives a canopy mask (Meta/WRI 1 m canopy height, ESA WorldCover 10 m, a Sentinel-2 NDVI median with an Otsu or fixed threshold, or your own raster); detects tree tops; counts the trees in line of sight from the facade viewpoints of every home; computes the canopy share per neighbourhood, home buffer or grid cell; computes the pedestrian-network distance to qualifying parks; and reports population-weighted compliance, inequality (Gini, concentration index, quintile gaps, Theil decomposition), Gaussian E2SFCA park supply, planting need, and a self-contained HTML report that typesets every formula and lists every reference.

The same chain is available in the lab dock (page by page, with charts, cards and styled layers) and in the Processing provider (one algorithm per step plus a full-run algorithm, usable in models and batch runs). Only qgis.core, GDAL, NumPy and PyQt are required; SciPy is used when present, with identical results without it.

Why it exists: the one QGIS plugin built for the rule, Análise 3-30-300 para QGIS, replaces "visible" with "mapped trees within a radius", measures straight-line distance and needs a ready-made canopy layer; it has no visibility model, no canopy detection, no network and no equity analysis. Other Hub plugins cover single pieces (street-level green view from paid Street View imagery, NDVI coverage per census sector, green space factor), and the R packages GreenExp_R and greenR live outside QGIS.

2. Install

  1. QGIS: Plugins > Manage and Install Plugins, search 02GreenRule, install. Or Install from ZIP with a release zip.
  2. A toolbar button and the menu Plugins > 02GreenRule open the lab; the Processing toolbox lists the provider 02GreenRule.
  3. Nothing else needs to be installed. Online sources (Overpass, Meta/WRI canopy height, ESA WorldCover, Earth Search Sentinel-2) are read anonymously; GDAL /vsicurl/ takes the QGIS proxy when one is set. Everything fetched is cached in the QGIS profile folder zero2greenrule/cache, so a repeated run works offline.

3. Units and working CRS

All metric work happens in a projected CRS in metres. If the buildings are geographic, not in metres, or Web Mercator (EPSG:3857 family, whose metres are stretched by 1/cos⁡φ1/\cos\varphi), every input is reprojected to the UTM zone of the area centre (EPSG:326xx north, 327xx south); otherwise the input CRS is kept. Every raster (your raster, Meta heights in EPSG:3857, WorldCover in EPSG:4326, Sentinel-2 in its own UTM zone) is warped onto one grid in the working CRS: masks and classes with nearest neighbour, heights and NDVI with bilinear resampling. The default cell is 1 m for sources of 2 m or finer, else the source resolution (10 m). When the canopy cell is coarser than 1 m, line of sight runs on its own 1 m building surface. Hard limits: study area at most 50 km² (OpenStreetMap fetches warn above 16 km²) and at most 64 million cells per grid, with a coarser cell proposed. The run log and the report state every reprojection and resampling.

4. The lab dock, page by page

The dock has a dark icon rail on the left; each page runs only its own step in a QGIS background task (progress bar and Cancel button; the dock stays responsive). Results go to a 02GreenRule layer group. Every control has a tooltip with its unit and reference.

1. Area

Study area from the map extent, a polygon layer or the neighbourhood layer. Figures: area and padded area (km²), the cell size and the working CRS with the reason it was chosen. Load demo district adds the synthetic district of section 8.

2. Data

Buildings (height and storeys fields), homes (optional address points, or a residential filter expression on the buildings, and a residents field), paths and streets, parks (access field), trees (height field), park entrances, neighbourhoods (id, name and residents fields) and an optional terrain model. Fetch missing layers downloads the ticked layers from OpenStreetMap for the padded area (the attribution line is shown). When Overpass is busy, Fetch with 02AgentOSM runs that plugin's Processing algorithm if it is installed.

3. Canopy

Source chips: Meta 1 m height, WorldCover 10 m, Sentinel-2 NDVI, or My raster (height, 0/1 mask or NDVI). Per source: the minimum tree height; the season and maximum cloud cover; the NDVI threshold (Otsu or fixed) with a histogram that marks tt; the cell size. Build canopy and detect tree tops adds the canopy raster (two-class style, water in blue) and the tree tops. Figures: canopy share of land, tree tops, cell, source note.

4. 3 · Trees in view

Mode: line of sight or within distance; trees required (3), radius RR (50 m), viewpoint floor (ground, middle, top), eye height (1.6 m), viewpoint spacing (10 m), storey height and default building height; optional visible greenness (VGVI). Count trees in view adds the homes layer. Figures: residents passing, median trees, the change if trees within RR counted instead of visible trees, homes. Show view fan of the selected home draws the sight lines of one home (visible green, blocked red) on the map.

5. 30 · Canopy

Neighbourhood: zones, a home buffer (250 m) or a square grid (500 m); target (30 percent) and minimum valid share (50 percent). Compute canopy share reuses the Canopy page settings. Figures: residents passing, median canopy share, area canopy, zones below target; a histogram and a gauge per zone.

6. 300 · Park access

Minimum merged park size (0.5 ha, 1 ha or custom), park kinds, distance (300 m), snap limit (100 m), network or straight line. Measure park access runs the part. Figures: residents within, median distance, unreachable homes, qualifying parks, mean detour; a distance histogram.

7. Equity

Population basis (floor area, footprint, equal), the equity variable and its direction (higher is more deprived, or better off), the park supply catchment d0d_0. Run the rule and equity runs every part. Charts: Lorenz curves with Gini, the concentration index with a plain reading, pass shares per quintile, the Theil split and Style homes by park supply (a map of E2SFCA park area per resident).

8. Scorecard

Run all (canopy, 3, 30, 300, equity, cards). A card for the study area and one per neighbourhood: grade, three gauges, residents, trees needed, homes beyond 300 m, the weakest part. A ranked table; Homes by score (0 to 3) and Zones by grade style the map. Export: Write HTML report, GeoPackage + GeoTIFF (every layer and the canopy raster), CSV tables.

Help

A short guide, the method references, the data licences, and links to this manual and the repository. The visible greenness here is geometric (a viewshed over the canopy mask); street-level photo indices are a different measure.

5. Method and equations

The equations carry the same numbers in the HTML report. Symbols: NN trees required, CC canopy target, DD distance threshold, RR view radius, nin_i trees visible from home ii, cic_i canopy share of its neighbourhood, did_i its park distance.

Canopy (the 30 part, and the tree source of the 3 part)

Meta/WRI 1 m canopy height (Tolan et al. 2024): cells at or above the minimum height (3 m) are canopy; tiles are the zoom-9 quadkeys of the area; 255 is treated as no data. ESA WorldCover 10 m 2021 v200 (Zanaga et al. 2022): class 10 is canopy, class 80 water. Sentinel-2 L2A: reflectance from the item's own scale and offset, SCL classes 0, 1, 3, 8, 9 and 10 masked, the per-pixel median NDVI of up to six leaf-on scenes (Earth Search collection 1; the earlier collection is the fallback).

Eq. (1) Normalised difference vegetation index Tucker 1979

NDVI=ρNIR−ρRedρNIR+ρRed,ρ=DN⋅s+o(1)\mathrm{NDVI} = \frac{\rho_{\mathrm{NIR}} - \rho_{\mathrm{Red}}}{\rho_{\mathrm{NIR}} + \rho_{\mathrm{Red}}},\qquad \rho = \mathrm{DN} \cdot s + o \tag{1}

Eq. (2) Otsu threshold Otsu 1979

t∗=arg⁡max⁡t  σB2(t)=ω0(t) ω1(t)[μ0(t)−μ1(t)]2(2)t^{*} = \arg\max_{t} \; \sigma_B^{2}(t) = \omega_0(t)\, \omega_1(t) \left[ \mu_0(t) - \mu_1(t) \right]^{2} \tag{2}

The mask codes are 0 land without canopy, 1 canopy, 254 water (valid data, outside the land share), 255 no data. For a neighbourhood uu with cell area aa:

Eq. (3) Canopy share of a neighbourhood Konijnendijk 2023

cu=∑k∈u1[canopyk] a∑k∈u1[landk] a,vu=#{k∈u:validk}#{k∈u}(3)c_u = \frac{\sum_{k \in u} \mathbb{1}[\text{canopy}_k]\, a}{\sum_{k \in u} \mathbb{1}[\text{land}_k]\, a},\qquad v_u = \frac{\#\{k \in u : \text{valid}_k\}}{\#\{k \in u\}} \tag{3}

cuc_u is NULL when vuv_u is below the minimum valid share. Ten-metre NDVI does not separate trees from grass, so this source is called "vegetation (NDVI)" in the dock and the report.

Tree tops

Local maxima of the canopy height model with a window that grows with height (Popescu & Wynne 2004), heights at least the minimum, tops closer than 2 m thinned; merged with inventory or OpenStreetMap trees (duplicates within 2 m, the point wins). Without tree points or heights, canopy patches give tree equivalents (patch area / crown area) and the 3 part runs in radius mode and is labelled an estimate.

Eq. (4) Variable search window for tree tops Popescu & Wynne 2004

w(h)=2.51503+0.00901 h2[m],3≤w≤15(4)w(h) = 2.51503 + 0.00901\, h^{2} \quad [\mathrm{m}], \qquad 3 \le w \le 15 \tag{4}

Trees in view (the 3 part)

Facade viewpoints lie every spacing metres along each home's outline, 0.5 m outside; viewpoints inside another building (shared walls) are dropped. With LL storeys of height ss and eye height ee:

Eq. (5) Viewpoint height Franklin & Ray 1994

zv=zterrain+f s+e,f∈{0, ⌊L/2⌋, L−1}(5)z_v = z_{\mathrm{terrain}} + f\, s + e, \qquad f \in \{0,\ \lfloor L/2 \rfloor,\ L - 1\} \tag{5}

The building surface is the terrain (or flat ground) plus building heights (height field, else storeys times storey height, else the default height). A tree tt at horizontal distance D≤RD \le R is visible when the surface stays below the sight line at every sample ss (step half a cell, from 1 m on), Franklin & Ray (1994):

Eq. (6) Line of sight to a tree Franklin & Ray 1994

visible(v,t)  ⟺  max⁡s: 1≤ds<DDSM(s)−zvds<zt−zvD,D≤R(6)\text{visible}(v, t) \iff \max_{s:\ 1 \le d_s < D} \frac{\mathrm{DSM}(s) - z_v}{d_s} < \frac{z_t - z_v}{D},\qquad D \le R \tag{6}

ztz_t is the crown top; the sight line to the top lies above the line to mid-crown, so "top or mid-crown visible" equals "top visible". Vegetation is not an obstacle. trees_vis counts distinct trees visible from any viewpoint of the home; trees_near counts trees within RR (the radius-only proxy). Optional visible greenness (Labib et al. 2021) uses a radial viewshed with a running maximum elevation angle; VV is the set of visible cells, gcg_c the canopy mask:

Eq. (7) Visible greenness Labib et al. 2021

VGVI=∑c∈Vgc w(dc)∑c∈Vw(dc),w(d)=1 or e−d/R(7)\mathrm{VGVI} = \frac{\sum_{c \in V} g_c\, w(d_c)}{\sum_{c \in V} w(d_c)}, \qquad w(d) = 1 \ \text{or}\ e^{-d/R} \tag{7}

Park access (the 300 part)

Qualifying parks: polygons of the chosen kinds whose access is not private or no, merged when they touch or lie within 5 m, with a merged area of at least 0.5 ha (the lower bound in Konijnendijk 2023; the WHO indicator uses 1 ha, Annerstedt van den Bosch et al. 2016). Access points: your entrances, else boundary points every 20 m within 25 m of a path, plus network nodes inside the park (distance 0); a park with none gets its nearest boundary point (kind "fallback", logged). The pedestrian graph keeps every highway except motorway, trunk and their links and access=no; lines connect where they share a vertex (within 0.5 m); edge weight is planar length. Crossings of lines without a shared vertex are counted and named in the run log: a street layer that is not split at its intersections falls apart into corridors (Processing: Split with lines); in OpenStreetMap such crossings are bridges and tunnels. One multi-source Dijkstra (1959) run from all access points gives d(⋅)d(\cdot) at every node; a home snapped at offset sis_i to edge (u,v)(u, v) of length LL at fraction tit_i:

Eq. (8) Network distance of a snapped home Dijkstra 1959

di=si+min⁡(d(u)+tiL,  d(v)+(1−ti)L)(8)d_i = s_i + \min\left( d(u) + t_i L,\; d(v) + (1 - t_i) L \right) \tag{8}

Homes beyond the snap limit or on a component without a park are NULL. d300_euc is the straight-line distance to the nearest qualifying park polygon (0 inside) and detour = d300_net / d300_euc.

Combined

Eq. (9) Rule score Konijnendijk 2023

Si=1[ni≥N]+1[ci≥C]+1[di≤D](9)S_i = \mathbb{1}[n_i \ge N] + \mathbb{1}[c_i \ge C] + \mathbb{1}[d_i \le D] \tag{9}

Eq. (10) Green Rule Index (plugin composite) Konijnendijk 2023

GRIi=13[min⁡ ⁣(niN,1)+min⁡ ⁣(ciC,1)+clip⁡ ⁣(1−di−DD,0,1)](10)\mathrm{GRI}_i = \tfrac{1}{3}\left[ \min\!\left(\tfrac{n_i}{N}, 1\right) + \min\!\left(\tfrac{c_i}{C}, 1\right) + \operatorname{clip}\!\left(1 - \tfrac{d_i - D}{D}, 0, 1\right) \right] \tag{10}

all_three is 1 when every part passes and 0 when any known part fails; SiS_i and the Green Rule Index (a plugin composite, not part of the published rule) are NULL when a part is NULL. Population weights: the residents field; else the zone residents split over its homes by floor area (footprint times storeys), footprint or equally; else 1 per home.

Equity

Indicators: trees_vis, canopy_pct, access 1/(1+d/D)1/(1 + d/D) and gri. With PiP_i and LiL_i the cumulative population and value shares of the Lorenz curve (Nesbitt et al. 2019 use Lorenz curves and Gini for green equity):

Eq. (11) Population-weighted Gini from the Lorenz curve Nesbitt et al. 2019

G=1−∑i=1n(Pi−Pi−1)(Li+Li−1)(11)G = 1 - \sum_{i=1}^{n} (P_i - P_{i-1})(L_i + L_{i-1}) \tag{11}

The concentration index (Kakwani, Wagstaff & van Doorslaer 1997) ranks residents by the equity variable from most deprived (r=0r = 0) to least deprived (r=1r = 1), ties at their mid-rank; C>0C > 0 means greenness is concentrated among the advantaged:

Eq. (12) Concentration index Kakwani et al. 1997

C=2 cov⁡w(h,r)μh(12)C = \frac{2\, \operatorname{cov}_w(h, r)}{\mu_h} \tag{12}

Theil T splits into a within-zone and a between-zone part (Theil 1967; Shorrocks 1980), sgs_g being the group's share of the total:

Eq. (13) Theil decomposition Theil 1967; Shorrocks 1980

T=∑gsgTg+∑gsgln⁡μgμ(13)T = \sum_g s_g T_g + \sum_g s_g \ln\frac{\mu_g}{\mu} \tag{13}

Quintiles of the population by the equity variable give pass shares per part and the Q5 - Q1 difference and Q5/Q1 ratio. Park supply is the Gaussian two-step floating catchment area (Luo & Wang 2003; Dai 2011) with network distances from bounded Dijkstra runs per park:

Eq. (14) Gaussian E2SFCA park supply Luo & Wang 2003; Dai 2011

Rj=Aj∑kPk G(dkj),Ai=∑jRj G(dij),G(d)=e−12(d/d0)2−e−121−e−12 (d≤d0)(14)R_j = \frac{A_j}{\sum_k P_k\, G(d_{kj})}, \quad A_i = \sum_j R_j\, G(d_{ij}), \quad G(d) = \frac{e^{-\frac{1}{2}(d/d_0)^2} - e^{-\frac{1}{2}}}{1 - e^{-\frac{1}{2}}}\ (d \le d_0) \tag{14}

Scorecard

Per neighbourhood: residents, homes, pass shares, mean visible trees, canopy percent, median distance, share meeting all three, a grade (A at least 80 percent, B 60, C 40, D 20, else E; a plugin convention), homes beyond the distance, the weakest part, and the planting need with land area AulandA^{\mathrm{land}}_u, canopy area AucanopyA^{\mathrm{canopy}}_u and crown area acrown=50 m2a_{\mathrm{crown}} = 50\ \mathrm{m}^2:

Eq. (15) Planting need Konijnendijk 2023

Pu=max⁡(0, C Auland−Aucanopy),nutrees=⌈Pu/acrown⌉(15)P_u = \max\left(0,\ C\, A^{\mathrm{land}}_u - A^{\mathrm{canopy}}_u\right), \qquad n^{\mathrm{trees}}_u = \left\lceil P_u / a_{\mathrm{crown}} \right\rceil \tag{15}

Residents by number of parts met

The report also gives the population-weighted share of residents meeting none, one, two and all three parts, over the homes whose three parts are known (the score field). A city where most residents meet two parts needs a different programme from one where many meet none.

Threshold sensitivity

The thresholds 3, 30 percent and 300 m are a rule of thumb (Konijnendijk 2023). The report adds the population-weighted share of residents passing each part at neighbouring thresholds: 1 to 6 trees, 20 to 40 percent canopy, and 200, 300, 400, 500 and 800 m, with the run's own threshold in bold. The study-area card and its planting need count only the cells inside the study area; the working grid itself is padded by the view or buffer radius.

6. Processing algorithms

Provider zero2greenrule, groups Data and canopy, 3-30-300 rule and Equity and scorecard. Every algorithm copies its inputs, reprojects them to the working CRS and writes its outputs with real field types (integers for ids and flags, doubles for measures, NULL where a value is undefined). Parameters marked advanced are under Advanced parameters. From Python: processing.run("zero2greenrule:<id>", {...}). The equation numbers in the help refer to section 5.

Fetch OpenStreetMap layers for an extent Data and canopy

Id: zero2greenrule:fetch_osm

Downloads buildings, trees (natural=tree nodes), public green spaces (leisure park, garden, nature_reserve, common; landuse recreation_ground, village_green) and paths/streets (every highway way) from the Overpass API for the extent and writes them in the working CRS. Building height comes from height (m or ft) or building:levels x 3 m and is NULL when neither is tagged. Answers are cached by request in the cache folder, so a repeated run works offline. Keep the extent below about 16 km2 (Overpass fair use); the request carries a User-Agent naming the plugin.

Output fields: buildings osm_id, height, levels, building; trees osm_id, height, source; parks osm_id, kind, name, access; paths osm_id, highway, access.

Data licences: Map data (c) OpenStreetMap contributors, ODbL 1.0. Meta and World Resources Institute High Resolution Canopy Height Maps, CC BY 4.0. ESA WorldCover 10 m 2021 v200, CC BY 4.0. Contains modified Copernicus Sentinel data (free and open).

ParameterDescriptionTypeDefaultNotes
EXTENTExtent to fetchExtent-
LAYERSLayersChoicebuildings, trees, parks, pathsOptions: buildings; trees; parks; paths. Several allowed.
CACHECache folder (empty: the profile cache) optionalFolder-
URLOverpass interpreter URL optional advancedStringhttps://overpass-api.de/api/interpreter
BUILDINGSOSM buildingsFeature sink (output)-
TREESOSM treesFeature sink (output)-
PARKSOSM parksFeature sink (output)-
PATHSOSM pathsFeature sink (output)-
N_BUILDINGSNumber of osm buildingsOutput number-Returned in the result dictionary.
N_TREESNumber of osm treesOutput number-Returned in the result dictionary.
N_PARKSNumber of osm parksOutput number-Returned in the result dictionary.
N_PATHSNumber of osm pathsOutput number-Returned in the result dictionary.

Canopy mask Data and canopy

Id: zero2greenrule:canopy_mask

Builds the canopy mask of the 30 part on a grid in the working CRS: 0 land without canopy, 1 canopy, 254 water (left out of the land share), 255 no data. Sources: Meta/WRI 1 m canopy height (cells at or above the minimum height), ESA WorldCover 10 m (class 10 tree cover, class 80 water), a Sentinel-2 L2A NDVI median of up to six leaf-on scenes (SCL classes 0, 1, 3, 8, 9, 10 masked) or your own raster. Online sources stream through GDAL /vsicurl/. NDVI is thresholded at a fixed value or by Otsu; 10 m NDVI does not separate trees from grass, so it is labelled 'vegetation (NDVI)'.

NDVI = (NIR - Red) / (NIR + Red), reflectance = DN * scale + offset (manual Eq. 1)

t* = argmax_t omega0(t) omega1(t) [mu0(t) - mu1(t)]^2 (manual Eq. 2)

Cell size 0 takes the source resolution (1 m for Meta, 10 m for WorldCover and Sentinel-2, the pixel size of a projected user raster).

Data licences: Map data (c) OpenStreetMap contributors, ODbL 1.0. Meta and World Resources Institute High Resolution Canopy Height Maps, CC BY 4.0. ESA WorldCover 10 m 2021 v200, CC BY 4.0. Contains modified Copernicus Sentinel data (free and open).

ParameterDescriptionTypeDefaultNotes
EXTENTStudy area extentExtent-
SOURCECanopy sourceChoiceMy raster: 0/1 maskOptions: Meta/WRI 1 m canopy height (online); ESA WorldCover 10 m (online); Sentinel-2 NDVI median (online); My raster: canopy height; My raster: 0/1 mask; My raster: NDVI.
RASTERMy raster (for the 'My raster' sources) optionalRaster layer-
MIN_HEIGHTMinimum canopy height (m)Number3.0Min 0.
NDVI_THRESHOLDNDVI threshold (empty: Otsu) optionalNumber-Min -1. Max 1.
CELLCell size (m, 0: from the source)Number0.0Min 0.
MAX_CLOUDSentinel-2 maximum cloud cover (percent) advancedNumber20.0Min 0. Max 100.
SEASONSentinel-2 season 'YYYY-MM-DD/YYYY-MM-DD' (empty: leaf-on of the last full year) optionalString-
OUTPUTCanopy maskRaster (output)-
CANOPY_SHARECanopy share of land (0..1)Output number-Returned in the result dictionary.
THRESHOLDThreshold used (m or NDVI)Output number-Returned in the result dictionary.

Detect tree tops Data and canopy

Id: zero2greenrule:detect_tree_tops

Tree tops from a canopy height model by local maxima with a window whose diameter grows with the height (pooled deciduous/pine fit of Popescu & Wynne 2004), clamped to 3-15 m. Cells below the minimum height are ignored; a plateau of equal tops keeps the cell nearest its centre; tops closer than the minimum spacing are thinned, the taller kept. SciPy is used when present, with an identical numpy fallback.

w(h) = 2.51503 + 0.00901 h^2 [m], 3 <= w <= 15 (manual Eq. 4)

Output fields: tree_id, height (m), source = detected. A geographic or Web Mercator CHM is warped (bilinear) to the UTM zone of its centre first.

ParameterDescriptionTypeDefaultNotes
CHMCanopy height model (m)Raster layer-
EXTENTExtent (empty: the whole raster) optionalExtent-
MIN_HEIGHTMinimum tree height (m)Number3.0Min 0.
MIN_SPACINGMinimum spacing between tops (m)Number2.0Min 0.
CELLCell size (m, 0: the raster's own) advancedNumber0.0Min 0.
OUTPUTTree topsFeature sink (output)-
TREESNumber of tree topsOutput number-Returned in the result dictionary.

3 - Trees in view from every home 3-30-300 rule

Id: zero2greenrule:trees_in_view

The 3 part: how many distinct trees each home sees. Facade viewpoints are placed every spacing metres along each home's outline, 0.5 m outside (viewpoints inside another building are dropped). A tree within the radius is visible when no cell of the building surface (terrain plus building heights) rises above the sight line from the viewpoint to the crown top. Trees come from the tree points, from tops detected on the canopy height model, or both (duplicates within 2 m merged, points win).

z_v = z_terrain + f s + e, f in {0, floor(L/2), L - 1} (manual Eq. 5)

visible(v, t) iff max over samples (DSM(s) - z_v) / d_s < (z_t - z_v) / D, D <= R (manual Eq. 6)

Output fields: trees_vis (distinct trees visible from any viewpoint), trees_near (trees within the radius, the radius-only proxy), pass_3 = trees_vis >= required, pop_w; the other rule fields stay NULL. Mode 'radius' counts trees within the radius without the line-of-sight test.

Geographic or Web Mercator input is reprojected to the UTM zone of its centre; every distance and area is planar metres there. Values that cannot be computed are NULL, never a sentinel.

ParameterDescriptionTypeDefaultNotes
BUILDINGSBuilding footprintsVector features-
RESIDENTIALResidential filter (expression on buildings; empty: every building is a home) optionalExpression-
HEIGHT_FIELDBuilding height field (m) optionalField-Field of BUILDINGS.
LEVELS_FIELDBuilding storeys field optionalField-Field of BUILDINGS.
POP_FIELDResidents per building field (empty: zone population or 1 per home) optionalField-Field of BUILDINGS.
STOREYStorey height (m) advancedNumber3.0Min 1. Max 10.
DEFAULT_HEIGHTHeight of buildings without height or storeys (m) advancedNumber9.0Min 1.
TREESTree points (optional) optionalVector features-
TREE_HEIGHT_FIELDTree height field (m) optionalField-Field of TREES.
CHMCanopy height model for tree tops (optional) optionalRaster layer-
EXTENTStudy area extent (empty: the extent of the buildings) optionalExtent-
TREES_REQUIREDTrees requiredNumber3Min 1.
MODECounting modeChoiceLine of sightOptions: Line of sight; Within the radius.
RADIUSView radius (m)Number50.0Min 1.
FLOORViewpoint floorChoiceGroundOptions: Ground; Middle; Top.
EYE_HEIGHTEye height above the floor (m)Number1.6Min 0.
SPACINGViewpoint spacing along facades (m)Number10.0Min 1.
TREE_HEIGHTHeight of trees without a height (m) advancedNumber8.0Min 0.5.
MIN_HEIGHTMinimum tree height on the CHM (m) advancedNumber3.0Min 0.
OUTPUTHomes - trees in viewFeature sink (output)-
SHARE_PASSPopulation share passing the 3 partOutput number-Returned in the result dictionary.
MEDIAN_VISIBLEMedian visible trees per homeOutput number-Returned in the result dictionary.

30 - Canopy share of each neighbourhood 3-30-300 rule

Id: zero2greenrule:canopy_share

The 30 part: the canopy share of the neighbourhood of every home. The neighbourhood is the zone holding the home, a disc of the buffer radius around it, or a generated square grid. Land is every valid cell that is not water; a unit whose valid share is below the minimum gets a NULL share. The canopy raster is a 0/1 mask (for instance the output of 'Canopy mask'), a canopy height model (threshold) or NDVI (fixed threshold or Otsu).

c_u = sum canopy cells * a / sum land cells * a, v_u = valid cells / all cells (manual Eq. 3)

Output: homes with canopy_pct, canopy_valid, pass_30 = canopy_pct >= target; neighbourhoods with canopy_pct, valid_share, canopy_m2, land_m2, pass_30 (in buffer mode the neighbourhood table uses the zones, or the grid when none are given).

Geographic or Web Mercator input is reprojected to the UTM zone of its centre; every distance and area is planar metres there. Values that cannot be computed are NULL, never a sentinel.

ParameterDescriptionTypeDefaultNotes
BUILDINGSBuilding footprintsVector features-
RESIDENTIALResidential filter (expression on buildings; empty: every building is a home) optionalExpression-
HEIGHT_FIELDBuilding height field (m) optionalField-Field of BUILDINGS.
LEVELS_FIELDBuilding storeys field optionalField-Field of BUILDINGS.
POP_FIELDResidents per building field (empty: zone population or 1 per home) optionalField-Field of BUILDINGS.
STOREYStorey height (m) advancedNumber3.0Min 1. Max 10.
DEFAULT_HEIGHTHeight of buildings without height or storeys (m) advancedNumber9.0Min 1.
CANOPYCanopy rasterRaster layer-
KINDCanopy raster holdsChoice0/1 canopy mask (254 = water)Options: 0/1 canopy mask (254 = water); Canopy height (CHM, metres); NDVI (-1..1).
MIN_HEIGHTMinimum canopy height for a CHM (m)Number3.0Min 0.
NDVI_THRESHOLDNDVI threshold (empty: Otsu) optionalNumber-Min -1. Max 1.
ZONESNeighbourhood zones (optional) optionalVector features-
ZONE_ID_FIELDZone id field optionalField-Field of ZONES.
ZONE_NAME_FIELDZone name field optionalField-Field of ZONES.
ZONE_POP_FIELDZone residents field (split over its homes) optionalField-Field of ZONES.
NEIGHBOURHOODNeighbourhoodChoiceZonesOptions: Zones; Home buffer; Square grid.
BUFFERBuffer radius (m)Number250.0Min 10.
GRID_SIZEGrid cell size (m)Number500.0Min 50.
TARGETCanopy target (percent)Number30.0Min 0. Max 100.
MIN_VALIDMinimum valid share of a unit advancedNumber0.5Min 0. Max 1.
EXTENTStudy area extent (empty: the extent of the buildings) optionalExtent-
OUTPUTHomes - canopy shareFeature sink (output)-
ZONES_OUTNeighbourhood canopyFeature sink (output)-
SHARE_PASSPopulation share passing the 30 partOutput number-Returned in the result dictionary.
CANOPY_SHARECanopy share of land in the grid (0..1)Output number-Returned in the result dictionary.

300 - Park access from every home 3-30-300 rule

Id: zero2greenrule:park_access

The 300 part: distance from every home to the nearest qualifying public green space. Parks of the chosen kinds whose access is not private/no are merged when they touch or lie within the merge gap; a merged park qualifies from the minimum size (0.5 ha by default, the lower bound in Konijnendijk 2023; 1 ha is the WHO indicator). Access points are your entrances, else boundary points every 20 m within 25 m of a path, plus network nodes inside the park. One multi-source Dijkstra run on the pedestrian network (motorway, trunk and access=no excluded; lines connect where they share a vertex) gives the network distance; homes beyond the snap distance or on a component without a park are NULL.

d_i = s_i + min(d(u) + t_i L, d(v) + (1 - t_i) L) (manual Eq. 8)

Output: homes with d300_net, d300_euc (straight line to the park polygon, 0 inside), detour = d300_net / d300_euc, park_id, pass_300; qualifying parks (park_id, area_ha, kind, name, n_access); access points (park_id, kind = entrance | edge | inside | fallback). Without a path layer the straight-line distance is used.

Geographic or Web Mercator input is reprojected to the UTM zone of its centre; every distance and area is planar metres there. Values that cannot be computed are NULL, never a sentinel.

ParameterDescriptionTypeDefaultNotes
BUILDINGSBuilding footprintsVector features-
RESIDENTIALResidential filter (expression on buildings; empty: every building is a home) optionalExpression-
HEIGHT_FIELDBuilding height field (m) optionalField-Field of BUILDINGS.
LEVELS_FIELDBuilding storeys field optionalField-Field of BUILDINGS.
POP_FIELDResidents per building field (empty: zone population or 1 per home) optionalField-Field of BUILDINGS.
STOREYStorey height (m) advancedNumber3.0Min 1. Max 10.
DEFAULT_HEIGHTHeight of buildings without height or storeys (m) advancedNumber9.0Min 1.
PARKSGreen spacesVector features-
KIND_FIELDGreen space kind field optionalFieldkindField of PARKS.
KINDSQualifying kinds (comma separated; empty: all) optionalStringpark,garden,nature_reserve,recreation_ground,village_green,common
ACCESS_FIELDAccess field (private/no excluded) optionalFieldaccessField of PARKS.
STREETSPaths and streets (optional) optionalVector features-
ENTRANCESPark entrances (optional) optionalVector features-
MIN_HAMinimum park size (ha)Number0.5Min 0.
DISTANCEDistance threshold (m)Number300.0Min 1.
MODEDistanceChoiceNetworkOptions: Network; Straight line.
SNAP_MAXMaximum snap distance to the network (m) advancedNumber100.0Min 1.
MERGE_GAPMerge parks closer than (m) advancedNumber5.0Min 0.
EXTENTStudy area extent (empty: the extent of the buildings) optionalExtent-
OUTPUTHomes - park accessFeature sink (output)-
PARKS_OUTQualifying parksFeature sink (output)-
ACCESS_OUTPark access pointsFeature sink (output)-
SHARE_PASSPopulation share passing the 300 partOutput number-Returned in the result dictionary.
PARKS_QUALIFYINGQualifying parksOutput number-Returned in the result dictionary.

3-30-300 rule (full run) 3-30-300 rule

Id: zero2greenrule:green_rule_3_30_300

The whole 3-30-300 study in one run: canopy mask, tree tops, trees in view (3), canopy share (30), park access (300), the combined score and Green Rule Index, population-weighted equity indicators and a card per neighbourhood. The output folder receives greenrule.gpkg (every layer), canopy_mask.tif, homes.csv, neighbourhoods.csv and greenrule_report.html, a self-contained report that typesets every formula and lists every reference and data licence.

S_i = 1[n_i >= N] + 1[c_i >= C] + 1[d_i <= D] (manual Eq. 9)

GRI_i = (min(n_i/N, 1) + min(c_i/C, 1) + clip(1 - (d_i - D)/D, 0, 1)) / 3 (manual Eq. 10)

The Green Rule Index is a plugin composite, not part of the published rule. all_three is 1 when every part passes and 0 when any known part fails; score and gri are NULL when a part is NULL. Population weights: the residents field, else the zone residents split over its homes by floor area, else 1 per home. Without tree points or heights the 3 part runs in radius mode on canopy-patch tree equivalents and is reported as an estimate.

Geographic or Web Mercator input is reprojected to the UTM zone of its centre; every distance and area is planar metres there. Values that cannot be computed are NULL, never a sentinel.

Data licences: Map data (c) OpenStreetMap contributors, ODbL 1.0. Meta and World Resources Institute High Resolution Canopy Height Maps, CC BY 4.0. ESA WorldCover 10 m 2021 v200, CC BY 4.0. Contains modified Copernicus Sentinel data (free and open).

ParameterDescriptionTypeDefaultNotes
BUILDINGSBuilding footprintsVector features-
RESIDENTIALResidential filter (expression on buildings; empty: every building is a home) optionalExpression-
HEIGHT_FIELDBuilding height field (m) optionalField-Field of BUILDINGS.
LEVELS_FIELDBuilding storeys field optionalField-Field of BUILDINGS.
POP_FIELDResidents per building field (empty: zone population or 1 per home) optionalField-Field of BUILDINGS.
STOREYStorey height (m) advancedNumber3.0Min 1. Max 10.
DEFAULT_HEIGHTHeight of buildings without height or storeys (m) advancedNumber9.0Min 1.
STREETSPaths and streets optionalVector features-
PARKSGreen spacesVector features-
KIND_FIELDGreen space kind field optionalFieldkindField of PARKS.
KINDSQualifying kinds (comma separated; empty: all) optionalStringpark,garden,nature_reserve,recreation_ground,village_green,common
ACCESS_FIELDAccess field (private/no excluded) optionalFieldaccessField of PARKS.
ENTRANCESPark entrances (optional) optionalVector features-
TREESTree points (optional) optionalVector features-
TREE_HEIGHT_FIELDTree height field (m) optionalField-Field of TREES.
ZONESNeighbourhood zones (optional) optionalVector features-
ZONE_ID_FIELDZone id field optionalField-Field of ZONES.
ZONE_NAME_FIELDZone name field optionalField-Field of ZONES.
ZONE_POP_FIELDZone residents field optionalField-Field of ZONES.
SOURCECanopy sourceChoiceMeta/WRI 1 m canopy height (online)Options: Meta/WRI 1 m canopy height (online); ESA WorldCover 10 m (online); Sentinel-2 NDVI median (online); My raster: canopy height; My raster: 0/1 mask; My raster: NDVI.
RASTERMy raster (for the 'My raster' sources) optionalRaster layer-
DEMTerrain model (optional) optionalRaster layer-
EXTENTStudy area extent (empty: the extent of the buildings) optionalExtent-
TREES_REQUIREDTrees requiredNumber3Min 1.
TARGETCanopy target (percent)Number30.0Min 0. Max 100.
DISTANCEPark distance threshold (m)Number300.0Min 1.
MODECounting modeChoiceLine of sightOptions: Line of sight; Within the radius.
RADIUSView radius (m)Number50.0Min 1.
FLOORViewpoint floorChoiceGroundOptions: Ground; Middle; Top.
NEIGHBOURHOODNeighbourhoodChoiceZonesOptions: Zones; Home buffer; Square grid.
BUFFERBuffer radius (m)Number250.0Min 10.
GRID_SIZEGrid cell size (m)Number500.0Min 50.
DIST_MODEPark distanceChoiceNetworkOptions: Network; Straight line.
MIN_HAMinimum park size (ha)Number0.5Min 0.
MIN_HEIGHTMinimum canopy height (m) advancedNumber3.0Min 0.
NDVI_THRESHOLDNDVI threshold (empty: Otsu) optional advancedNumber-Min -1. Max 1.
EYE_HEIGHTEye height above the floor (m) advancedNumber1.6Min 0.
SPACINGViewpoint spacing along facades (m) advancedNumber10.0Min 1.
TREE_HEIGHTHeight of trees without a height (m) advancedNumber8.0Min 0.5.
MIN_VALIDMinimum valid share of a unit advancedNumber0.5Min 0. Max 1.
SNAP_MAXMaximum snap distance to the network (m) advancedNumber100.0Min 1.
MERGE_GAPMerge parks closer than (m) advancedNumber5.0Min 0.
D0Park supply catchment d0 (m) advancedNumber1000.0Min 50.
CROWN_AREACrown area per planted tree (m2) advancedNumber50.0Min 1.
EQUITY_FIELDEquity variable (field on buildings or zones) optionalField-Field of BUILDINGS.
HIGHER_DEPRIVEDHigher equity values mean more deprivedBooleanyes
VGVICompute visible greenness (VGVI) advancedBooleanno
FOLDEROutput folder (report, GeoPackage, CSV, canopy GeoTIFF) optionalFolder (output)-
OUTPUT3-30-300 homesFeature sink (output)-
ZONES_OUT3-30-300 neighbourhoodsFeature sink (output)-
TREES_OUTTrees usedFeature sink (output)-
PARKS_OUTQualifying parksFeature sink (output)-
ACCESS_OUTPark access pointsFeature sink (output)-
REPORT3-30-300 reportOutput HTML file-Returned in the result dictionary.
SHARE_3Population share passing 3Output number-Returned in the result dictionary.
SHARE_30Population share passing 30Output number-Returned in the result dictionary.
SHARE_300Population share passing 300Output number-Returned in the result dictionary.
SHARE_ALLPopulation share passing all threeOutput number-Returned in the result dictionary.

Equity indicators Equity and scorecard

Id: zero2greenrule:equity_indicators

Population-weighted inequality of the 3-30-300 results, read from the homes output of the full run (or of a layer with the same fields). Indicators: trees_vis, canopy_pct, access = 1 / (1 + d / D) with d the distance the run tested pass_300 on (network, or straight line for a straight-line run), and gri. For each: the Gini from the Lorenz curve, the Theil T split into between-zone and within-zone parts, and, with an equity variable, the concentration index ranked from most deprived to least deprived (C > 0: greenness is concentrated among the advantaged). Population quintiles of the equity variable give pass shares per part and the Q5 - Q1 gap.

G = 1 - sum (P_i - P_(i-1)) (L_i + L_(i-1)) (manual Eq. 11)

C = 2 cov_w(h, r) / mu_h (manual Eq. 12)

T = sum_g s_g T_g + sum_g s_g ln(mu_g / mu) (manual Eq. 13)

The equity variable is a field of the homes layer, or of a polygon layer (census areas, buildings) holding the homes. Zones for the Theil split are optional (one group without them).

ParameterDescriptionTypeDefaultNotes
HOMES3-30-300 homesVector features-
POP_FIELDPopulation weight field optionalFieldpop_wField of HOMES.
EQUITY_FIELDEquity variable on the homes optionalField-Field of HOMES.
EQUITY_LAYERPolygons holding the equity variable (optional) optionalVector features-
EQUITY_LAYER_FIELDEquity variable on the polygons optionalField-Field of EQUITY_LAYER.
HIGHER_DEPRIVEDHigher equity values mean more deprivedBooleanyes
ZONESZones for the Theil split (optional) optionalVector features-
DISTANCEPark distance threshold D (m)Number300.0Min 1.
OUTPUTEquity indicatorsFeature sink (output)-
QUINTILESEquity quintilesFeature sink (output)-
GINI_GRIGini of the Green Rule IndexOutput number-Returned in the result dictionary.
CONC_GRIConcentration index of the Green Rule IndexOutput number-Returned in the result dictionary.
GAP_ALLQ5 - Q1 difference of all threeOutput number-Returned in the result dictionary.

Neighbourhood scorecard Equity and scorecard

Id: zero2greenrule:neighbourhood_scorecard

One card per neighbourhood from the homes output of the full run: residents, homes, the population share passing each part and all three, mean visible trees, canopy percent, median park distance, a grade (A at least 80 percent of residents meet all three, B 60, C 40, D 20, else E; a plugin convention), homes beyond the distance threshold, the weakest part and, with a canopy mask, the planting need and the trees it takes.

P_u = max(0, C A_land - A_canopy), trees = ceil(P_u / a_crown) (manual Eq. 15)

Without a canopy mask the canopy percent is the population-weighted mean of the homes' canopy_pct and the planting need is NULL. Output fields: zone_id, name, population, homes, share_3, mean_trees, canopy_pct, share_30, share_300, median_d300, share_all, grade, planting_need_m2, trees_needed, homes_beyond_300, weakest.

ParameterDescriptionTypeDefaultNotes
HOMES3-30-300 homesVector features-
POP_FIELDPopulation weight field optionalFieldpop_wField of HOMES.
ZONESNeighbourhoodsVector features-
ZONE_ID_FIELDZone id field optionalField-Field of ZONES.
ZONE_NAME_FIELDZone name field optionalField-Field of ZONES.
CANOPYCanopy mask (optional, from 'Canopy mask') optionalRaster layer-
TARGETCanopy target (percent)Number30.0Min 0. Max 100.
DISTANCEPark distance threshold (m)Number300.0Min 1.
CROWN_AREACrown area per planted tree (m2)Number50.0Min 1.
MIN_VALIDMinimum valid share of a zone advancedNumber0.5Min 0. Max 1.
OUTPUTNeighbourhood scorecardFeature sink (output)-
SHARE_ALLPopulation share meeting all threeOutput number-Returned in the result dictionary.
TREES_NEEDEDTrees needed in all zonesOutput number-Returned in the result dictionary.

7. Output layers, fields and files

Homes (points, one per home)

FieldMeaning
home_id, bld_idHome number (1..n) and the 0-based position of its building (negative for a home point outside every footprint).
pop_wResidents (population weight).
trees_vis, trees_near, vgvi, pass_3Distinct trees visible, trees within RR, visible greenness (0..1, NULL unless computed), 3 part passed (1/0).
canopy_pct, canopy_valid, pass_30Canopy percent of the neighbourhood, its valid share (0..1), 30 part passed.
d300_net, d300_euc, detour, park_id, pass_300Network and straight-line distance (m), their ratio, nearest qualifying park, 300 part passed.
score, all_three, gri, equity_qParts met (0..3), all three met, Green Rule Index (0..1), equity quintile (1 = most deprived).
trees_short, d300_overGap to the thresholds: max⁡(0,N−ni)\max(0, N - n_i) trees short of the 3 part and max⁡(0,di−D)\max(0, d_i - D) metres beyond the 300 part (NULL when unknown); a planting or park-access programme can start with the homes that miss by little.

Neighbourhoods (polygons, one per zone or grid cell)

zone_id, name, population, homes, share_3, mean_trees, canopy_pct, share_30, share_300, median_d300, share_all, grade, planting_need_m2, trees_needed, homes_beyond_300, weakest. Shares are fractions 0..1, canopy_pct is a percentage.

Other layers

Trees used (tree_id, height, source = inventory, osm, detected or canopy patch), qualifying parks (park_id, area_ha, kind, name, n_access), park access points (park_id, kind = entrance, edge, inside or fallback). The data attribution lines are written into every output layer's metadata.

Files

An output folder holds greenrule.gpkg (tables homes, neighbourhoods, trees, parks, park_access), canopy_mask.tif (codes 0/1/254/255), homes.csv, neighbourhoods.csv and greenrule_report.html: a self-contained page (no network needed) with the summary, the cards, Lorenz and quintile charts, every equation of section 5 typeset with KaTeX, the parameters, the run log, the references and the data licences.

8. Worked example: the demo district

Load demo district (Area page) or core.demo.make_demo_district() builds a district in EPSG:32635 whose answers are known by construction. Streets run every 150 m (x 0 to 900 m, y 0 to 600 m) plus one isolated footway; a motorway crosses the grid diagonally and must be dropped by the pedestrian filter. Four blocks hold green spaces: two parks of 1.96 ha with entrances on the surrounding streets, a 0.2 ha garden (too small) and a private park. Every other block holds one 10 m x 10 m home of 9 m; one more home sits by the isolated footway. Home kk gets k mod 6k \bmod 6 trees 15 m north of its facade in open view, a tree behind a 30 m wall when k mod 3=0k \bmod 3 = 0 and a tree just beyond R=50R = 50 m when k mod 2=0k \bmod 2 = 0. The West zone has 40.0 percent canopy, the East zone 10.0 percent.

With the residential filter "resid" = 1, the residents field pop, the trees, the zones, the canopy mask, the entrances and the equity variable dep_idx (higher is more deprived), the full run gives:

MeasureStudy areaWestEast
Homes / residents21 / 67211 / 37810 / 294
Share passing 3 (residents)38.7 %39.4 %37.8 %
Canopy / share passing 3025.0 % / 56.3 %40.0 % / 100 %10.0 % / 0 %
Share passing 300 / median distance35.1 % / 330 m30.6 % / 330 m40.5 % / 255 m
Share meeting all three / grade3.4 % / E6.1 % / E0 % / E
Planting need / trees needed36 000 m² / 7200 / 072 000 m² / 1 440
Weakest part30030030

Every home's visible trees equal its planted open-view trees (0 to 5): the walled tree is never seen, the far tree never counted; in radius mode the walled tree is counted, which is the difference the 3 page reports. The network distances are the planted grid distances (30 to 480 m); the isolated home is NULL, not a large number. The East planting need is 0.30×360 000−36 000=72 000 m20.30 \times 360\,000 - 36\,000 = 72\,000\ \mathrm{m}^2, or 1 440 trees of 50 m². Equity: the Gini of visible trees is 0.473, of canopy percent 0.275; the concentration index of the Green Rule Index is 0.017; the canopy inequality lies entirely between the two zones (Theil within part 0), and the Q5 - Q1 gap of residents meeting all three is 12.9 percentage points.

The smoke tests of the plugin check these values: per-home results against the planted truth, zone canopy, hand-computed straight-line distances, an independent pairwise Gini and concentration index, translation invariance (+137 km, +211 km) and identical results through the dock, the core and Processing.

9. Open data sources and licences

Overpass API (overpass-api.de) answers only requests with a User-Agent, which the plugin sends; keep fetches below about 16 km². Meta/WRI canopy height tiles are stripped GeoTIFFs: for areas taller than about 1 km the whole quadkey tile is downloaded to the cache once (its size is shown first). ESA WorldCover and Sentinel-2 are cloud-optimised and read by window.

10. Limitations

11. References

  1. Annerstedt van den Bosch, M., Mudu, P., Uscila, V., et al. (2016). Development of an urban green space indicator and the public health rationale. Scandinavian Journal of Public Health, 44(2), 159-167. https://doi.org/10.1177/1403494815615444
  2. Browning, M. H. E. M., Locke, D. H., Konijnendijk, C., et al. (2024). Measuring the 3-30-300 rule to help cities meet nature access thresholds. Science of the Total Environment, 907, 167739. https://doi.org/10.1016/j.scitotenv.2023.167739
  3. Dai, D. (2011). Racial/ethnic and socioeconomic disparities in urban green space accessibility: Where to intervene? Landscape and Urban Planning, 102(4), 234-244. https://doi.org/10.1016/j.landurbplan.2011.05.002
  4. Dijkstra, E. W. (1959). A note on two problems in connexion with graphs. Numerische Mathematik, 1(1), 269-271. https://doi.org/10.1007/BF01386390
  5. Franklin, W. R., & Ray, C. K. (1994). Higher isn't necessarily better: Visibility algorithms and experiments. In T. C. Waugh & R. G. Healey (Eds.), Advances in GIS Research: Sixth International Symposium on Spatial Data Handling (pp. 751-770). Taylor & Francis.
  6. Kakwani, N., Wagstaff, A., & van Doorslaer, E. (1997). Socioeconomic inequalities in health: Measurement, computation, and statistical inference. Journal of Econometrics, 77(1), 87-103. https://doi.org/10.1016/S0304-4076(96)01807-6
  7. Konijnendijk, C. C. (2023). Evidence-based guidelines for greener, healthier, more resilient neighbourhoods: Introducing the 3-30-300 rule. Journal of Forestry Research, 34(3), 821-830. https://doi.org/10.1007/s11676-022-01523-z
  8. Labib, S. M., Huck, J. J., & Lindley, S. (2021). Modelling and mapping eye-level greenness visibility exposure using multi-source data at high spatial resolutions. Science of the Total Environment, 755, 143050. https://doi.org/10.1016/j.scitotenv.2020.143050
  9. Luo, W., & Wang, F. (2003). Measures of spatial accessibility to health care in a GIS environment: Synthesis and a case study in the Chicago region. Environment and Planning B: Planning and Design, 30(6), 865-884. https://doi.org/10.1068/b29120
  10. Nesbitt, L., Meitner, M. J., Girling, C., Sheppard, S. R. J., & Lu, Y. (2019). Who has access to urban vegetation? A spatial analysis of distributional green equity in 10 US cities. Landscape and Urban Planning, 181, 51-79. https://doi.org/10.1016/j.landurbplan.2018.08.007
  11. Nieuwenhuijsen, M. J., Dadvand, P., Márquez, S., et al. (2022). The evaluation of the 3-30-300 green space rule and mental health. Environmental Research, 215, 114387. https://doi.org/10.1016/j.envres.2022.114387
  12. Otsu, N. (1979). A threshold selection method from gray-level histograms. IEEE Transactions on Systems, Man, and Cybernetics, 9(1), 62-66. https://doi.org/10.1109/TSMC.1979.4310076
  13. Popescu, S. C., & Wynne, R. H. (2004). Seeing the trees in the forest: Using lidar and multispectral data fusion with local filtering and variable window size for estimating tree height. Photogrammetric Engineering & Remote Sensing, 70(5), 589-604. https://doi.org/10.14358/PERS.70.5.589
  14. Shorrocks, A. F. (1980). The class of additively decomposable inequality measures. Econometrica, 48(3), 613-625. https://doi.org/10.2307/1913126
  15. Theil, H. (1967). Economics and Information Theory. Amsterdam: North-Holland.
  16. Tolan, J., Yang, H.-I., Nosarzewski, B., et al. (2024). Very high resolution canopy height maps from RGB imagery using self-supervised vision transformer and convolutional decoder trained on aerial lidar. Remote Sensing of Environment, 300, 113888. https://doi.org/10.1016/j.rse.2023.113888
  17. Tucker, C. J. (1979). Red and photographic infrared linear combinations for monitoring vegetation. Remote Sensing of Environment, 8(2), 127-150. https://doi.org/10.1016/0034-4257(79)90013-0
  18. WHO Regional Office for Europe (2016). Urban green spaces and health: A review of evidence. Copenhagen: WHO Regional Office for Europe.
  19. Zanaga, D., Van De Kerchove, R., Daems, D., et al. (2022). ESA WorldCover 10 m 2021 v200. Zenodo. https://doi.org/10.5281/zenodo.7254221

12. Citation, licence, links

Eminoğlu, Y. (2026). 02GreenRule - 3-30-300 Green Equity Lab (version 0.1.1) [QGIS plugin]. https://gitlab.com/geophilo1/zero2greenrule. https://doi.org/10.5281/zenodo.24263658. The DOI 10.5281/zenodo.24263658 always resolves to the latest version; version 0.1.1 has DOI https://doi.org/10.5281/zenodo.24263659.

Licence: GNU General Public License v2.0 or later. Source and issues: gitlab.com/geophilo1/zero2greenrule (tracker work items). Homepage: geophilo.com/zero2greenrule.